Socioeconomic Impact of HIV/AIDS in Ukraine THE WORLD BANK Copyright © 2006 1818 H Street, NW Washington, DC 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: May 2006 Library of Congress Cataloging-in-Publication Data has been requested. Socioeconomic Impact of HIV/AIDS in Ukraine THE WORLD BANK ii Table of Contents Acknowledgements .............................................................................................................................. ix Executive Summary ............................................................................................................................ x CHAPTER 1 Introduction .............................................................................................................. 1 CHAPTER 2 The HIV/AIDS Epidemic in Ukraine: Status and Trends .................................. 4 The Changing Pattern of Transmission of the HIV/AIDS Epidemic .................................... 4 HIV Sentinel Surveillance Data ............................................................................................ 5 Most-at-Risk Populations (MARPs) ...................................................................................... 5 Ukraine's Response to HIV/AIDS .......................................................................................... 9 CHAPTER 3 Demographic Forecast under the HIV/AIDS Epidemic ...................................... 10 Analysis at the National Level .............................................................................................. 10 Analysis at the Regional Level .............................................................................................. 15 CHAPTER 4 Impact of the Epidemic on the Labor Force and Government Revenues .......... 19 Analysis at the National Level .............................................................................................. 19 Analysis at the Regional Level .............................................................................................. 20 Impact of Epidemic on Government Budget Position and Special Social Protection Funds............................................................................................ 20 CHAPTER 5 Estimating the Macroeconomic Costs of the HIV/AIDS Epidemic .................... 23 Simple Growth Model ............................................................................................................ 24 Macroeconomic Model............................................................................................................ 27 Multisector CGE Mode .......................................................................................................... 28 CHAPTER 6 Policy Implications and Conclusions .................................................................... 32 ANNEX 1 HIV/AIDS in Ukraine: Official Data ................................................................................ 35 ANNEX 2 Demographic Forecast: Methodology and Assumptions.................................................... 40 ANNEX 3 Methodology for Estimating Labor Force and Employment ............................................ 48 ANNEX 4 Methodology and Assumptions for Estimating the Impact of HIV/AIDS on Government Budget and Social Insurance Funds ...................................................... 63 ANNEX 5 Growth Model: Methodology and Assumptions ................................................................ 76 ANNEX 6 Measuring the Burden of HIV/AIDS in Ukraine: Methodology ........................................ 80 ANNEX 7 Macroeconometric Model: Methodology, Assumptions, and Estimation .......................... 86 ANNEX 8 CGE Model: Methodology and Assumptions ...................................................................... 98 Bibliography ........................................................................................................................................ 108 iii List of Figures Figure 2-1. Cumulative Reported Cases of HIV Infection, 1987-2004 .............................................. 4 Figure 2-2. Total Numbers of HIV Tests and Officially Registered HIV Cases, 1987-2005 .............. 5 Figure 2-3. Leading Modes of HIV Transmission, 1987-2004 .......................................................... 6 Figure 2-4. Total Number of HIV Tests Conducted, by Category, 1994-2004.................................... 8 Figure 3-1. Forecasted Number of Those Infected with HIV, 1994-2014 .......................................... 11 Figure 3-2. Forecasted Share of AIDS Deaths in Total Number of Deaths, 1994-2004 .................... 12 Figure 3-3. Forecasted Share of AIDS Deaths in Total Adult (15-49) Deaths, 2004-2014................ 12 Figure 3-4. Forecasted Accumulated AIDS Deaths, 1994-2014 ........................................................ 13 Figure 3-5. Forecasted Life Expectancy at Birth, 1994-2014, in Years ............................................ 13 Figure 3-6. Forecasted Total Population, 1994-2014 ........................................................................ 14 Figure 3-7. Age-Gender Composition of Population, "No-AIDS" and "AIDS Medium" Scenarios, 2014 .................................................................................................. 14 Figure 3-8. Forecasted Total AIDS Orphans, 1994-2014 .................................................................. 15 Figure 3-9. Estimated AIDS-Related Mortality in the Working-Age (15-59) Population in Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa Oblasts and Ukraine, 2014.............................. 17 Figure 3-10. Forecasted Life Expectancy in "No-AIDS," "AIDS Optimistic," and "AIDS Pessimistic" Scenarios, Odesa, Mykolayiv, Donetsk, and Dnipropetrovsk Oblasts, 1994-2014, in Life Years.................................................................................................. 18 Figure 4-1. Regional Comparison of Estimated Labor Force Reduction from the Epidemic, Compared to the "No-AIDS" Scenario, 2014 (Percentage of Reduction) ...................................... 20 Figure 5-1. Number of HIV Infections Averted by Realizing "AIDS Optimistic" Rather Than "AIDS Pessimistic" Scenario, 2004-14 .................................................................... 26 Figure 5-2. Number of DALYs Averted by Realizing "AIDS Optimistic" Rather Than "AIDS Pessimistic" Scenario, 2004-14 .................................................................... 26 iv List of Tables Table 4-1. Estimated Reduction in Selected Labor Market Indicators in the "No-AIDS" Scenario, 2004-14 (in Thousands) and Percentage of Reduction .............................. 19 Table 4-2. Estimated Reductions in Selected Labor Market Indicators due to the Epidemic, Compared to the "No-AIDS" Scenario, 2014 (in Thousands and as Percentage of Reduction) .......................................................................................................... 20 Table 4-3. Regional Comparison of Estimated Losses from the Epidemic in Selected Labor Market Indicators, Compared to the "No-AIDS" Scenario, 2014 (Percentage of Reduction).............................................................................................................. 21 Table 4-4. Total Additional Annual Non-medical Budgetary Losses/Costs Associated with HIV/AIDS, 2014...................................................................................................................... 22 Table 5-1. Estimated Annual Medical Expenditure Associated with HIV/AIDS Prevention and Treatment in 2014................................................................................................ 25 Table 5-2. Projected Epidemic Outcomes, 2014, Scenario Analysis.................................................. 25 Table 5-3. Macroeconometric Model: Estimated Difference in Sectoral Output in Two Epidemic Scenarios, 2005-14 ................................................................................................ 28 Table 5-4. CGE Model: Macroeconomic Implications of the Epidemic, Scenario Analysis .............. 29 Table 5-5. CGE Model: Sectoral Implications of HIV/AIDS Epidemic, Scenario Analysis................ 30 v List of Abbreviations ANC....................................................Antenatal Clinic AR ......................................................Autonomous Republic ART....................................................Antiretroviral Therapy ARV....................................................Antiretroviral CGE....................................................Computable General Equilibrium CIS......................................................Commonwealth of Independent States CPI .....................................................Consumer Price Index DALY..................................................Disability Adjusted Life Year EGW...................................................Electricity, Gas, and Water FDIs ...................................................Foreign Direct Investments GBD ...................................................Global Burden of Disease GDP....................................................Gross Domestic Product GFATM ..............................................Global Fund for AIDS, Tuberculosis and Malaria HIV/AIDS...........................................Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome IDUs...................................................Injecting Drug Users ILO .....................................................International Labor Organization M3 ......................................................Broad Money Aggregate MARP.................................................Most-at-risk Population MOH...................................................Ministry of Health MTCT.................................................Mother-to-Child Transmission NAS....................................................National Academy of Sciences NBU ..................................................National Bank of Ukraine OLG....................................................Overlapping Generations PPI......................................................Producer Price Index SPF.....................................................Social Protection Fund STI......................................................Sexually Transmitted Infection TB.......................................................Tuberculosis TFP ....................................................Total Factor Productivity UAH ...................................................Hryvnia UNAIDS.............................................Joint United Nations Programme on HIV/AIDS UNDP.................................................United Nations Development Program VAT.....................................................Value Added Tax WHO ..................................................World Health Organization YLD ....................................................Years Lost to Disability YLL.....................................................Years of Life Lost Currency is Hryvnia or UAH: Exchange rate 1US$ = 5.3 UAH. vi Acknowledgements T his study was jointly conducted by the World Institute of Social Research; Pavlo Smyrnov, Anna Bank and a group of Ukrainian experts in col- Dovbakh, Lyudmyla Husak, Andrey Klepikov, laboration with the Joint United Nations International HIV/AIDS Alliance in Ukraine; and Program on HIV/AIDS (UNAIDS) and the Valeriy Khmarskiy, Ministry of Health of Ukraine. International HIV/AIDS Alliance in Ukraine through the support from the Global Fund for AIDS, Alexandra Sidorenko provided methodological sup- Tuberculosis and Malaria. The World Bank team is port to the study. She and Shiyan Chao prepared the led by Shiyan Chao, Senior Health Economist, and final report. Vinay Saldanha from UNAIDS con- includes Alexandra Sidorenko, Consultant, tributed to the study. Olena Bekh from the World Australian National University. The Ukraine team Bank, Kiev office, provided coordination between consists of Olga Balakireva, Social Monitoring the Bank and the Ukraine research team. Beth Center of Ukraine; Nina Baranova, Center for Goodrich edited the report. Perspective Social Studies, Ministry of Labor and Social Policies; Igor Burakovsky, Institute for The study team is thankful to the study peer review- Economic Research and Policy Consultations; Irina ers, Martha Ainsworth, Mattias Lundberg, and Demchenko, Socioconsulting Analytical Center; Edmundo Murrugarra from the World Bank and Yuriy Kruglov, Ukrainian National AIDS Center, Marc Suhrcke from World Health Organization Ministry of Health, Ukraine; Nataliya Levchuk, (WHO), for their valuable comments. Institute of Demography and Social Research, National Academy of Sciences (NAS) of Ukraine; The team is grateful to World Bank managers Zoryana Medvid, Center for Perspective Social Paul Bermingham, Country Director of Ukraine, Studies, Ministry of Labor and Social Policies and Belarus, and Moldova, and Armin Fidler, Health NAS of Ukraine; Oleg Nivyevskiy, Institute for Sector Manager, for their guidance and support to Economic Research and Policy Consultations; the study. The team is also grateful for comments Ferdinand Pavel, Institute for Economic Research and suggestions received from Toomas Palu, Jack and Policy Consultations; Nataliya Polyak, Center for Langenbrunner, and Merrell Tuck-Primdahl, the World Perspective Social Studies, Ministry of Labor and Bank; Alla Scherbinskaya, Ukrainian National AIDS Social Policies and NAS of Ukraine; Mariya Center; Arkadiusz Majszyk, Anna Shakarishvili, and Skrypnichenko, Institute for Economic Forecasting, Lidiya Andruschak, UNAIDS; Veena Lakhumalani, the NAS of Ukraine; Maryna Varban, Socioconsulting British Council; and Yuriy Kobyscha and Ariele Braye, Analytical Center; Olexandr Yaremenko, Ukrainian WHO, Ukraine Office. Financial assistance from UNAIDS Trust Fund is gratefully acknowledged. vii Executive Summary T his study of the socioeconomic impact of Ukraine's HIV/AIDS Epidemic HIV/AIDS in Ukraine was prompted by the Ukraine's rate of HIV infection is growing need to understand the potential impact of the fast. HIV/AIDS is a relatively new phenomenon for country's rapidly growing HIV/AIDS epidemic. Ukraine, with rapid spread of the virus only since Ukraine was classified by the World Health 1994. Consequently, the overall prevalence rate is Organization as a low HIV prevalence country in still relatively low, but the rate of infection increase 1995, but only a decade later, Ukraine suffers the is alarming: an average 33 percent increase per year worst HIV/AIDS epidemic in Europe (DeBell and since 1994. UNAIDS estimates that Ukraine had Carter 2005). Clearly, it is collective failure that pre- 360,000 infected adults as of the end of 2003 and an vents Ukraine from controlling the epidemic. Lack of adult prevalence rate of 1.4 percent. (The exact num- understanding of the epidemic and its potentially ber of infections is unknown due to the high degree devastating impact contribute to stigma, denial, and of uncertainty associated with the size of the most- inadequate responses. at-risk populations.) This study projects that nearly a half million people (477,000) were infected with HIV The World Bank and the Ministry of Health of in 2004, a 32 percent increase over 2003. It also proj- Ukraine jointly conducted this study in collaboration ects that, by 2014, the total number of HIV-positive with the Joint United Nations Programme on people will range from 478,500 under the optimistic HIV/AIDS (UNAIDS) and the International HIV/AIDS scenario to 820,400 under the pessimistic one. The Alliance in Ukraine. adult prevalence rate will be between 1.9 and 3.5 percent, depending on projection assumptions. The study assesses the short- and medium-term (2004-14) socioeconomic impact of the HIV/AIDS The pattern of transmission is changing. epidemic and provides evidence for policy making. Until now, the HIV epidemic in Ukraine was concen- Using data available in January 2005, it evaluates the trated in sub-populations, mainly injecting drug users epidemic's impact on population growth, life expectancy, employment, and health care and social service costs in Ukraine and projects the potential benefits of disease prevention and treatment. Data 1 Epidemic scenarios differ in their assumptions about the size and dynamics of the most-at-risk populations, yielding different esti- from the Ukrainian AIDS Center and socioeconomic mates of adult prevalence rates. In our optimistic scenario, the information from the government and other agencies adult HIV prevalence rate peaks at 2% in 2010, peaks at 2.48% in 2009-10 in the medium scenario, and rises continuously, reaching were used to construct a baseline "no-AIDS" demo- 3.5% in 2014 in the pessimistic scenario. Reduction in the vertical graphic projection of the Ukrainian population and transmission rate (15.9% in 2003) is faster in the optimistic sce- three epidemic scenarios--optimistic, medium, and nario (to 10% in 2004 and then to 5% in 2014) than in medium (gradual reduction to 5% by 2014) and pessimistic scenarios (grad- pessimistic.1 These projections were used to apply ual reduction to 10% in 2014). Availability of antiretroviral (ARV) several macroeconomic models to estimate the therapy to those who need it increases from 1% in 2004, to 30% in 2010, and further to 50% in 2014 in the optimistic scenario; to 5% in impact of the HIV/AIDS epidemic on various sub- 2005, further to 10% by 2010, and remaining at 10% until 2014 in the populations, regions, and sectors. medium one; and to 5% in 2005 and remaining at 5% until 2014 in the pessimistic scenario. The study also constructed three cost scenarios for antiretroviral therapy (ART). viii (IDUs), with the prevalence rate among pregnant Socioeconomic Impact of HIV/AIDS women in urban areas still below 1 percent. Barnett The epidemic's impact on demographics and et al. (2001) pointed out that the epidemic was shift- health status could be devastating. The ing from high-risk groups to the general population largest demographic impact of the HIV/AIDS epidem- through heterosexual transmission. Feshbach and ic in Ukraine is through its effect on population mor- Galvin (2005) reinforced this conclusion in their bidity and mortality rates. The majority of all HIV recent article. The epidemic's tendency to spill into infections are among those in the most active repro- the general population is reflected in official ductive age (20-34). The disease certainly affects Ukrainian AIDS Center data indicating that the share their capacity for childbearing, and Ukraine has per- of infections caused by intravenous drug use sistently declining birth rates. Over 1991-2003, the decreased from 83.6 percent in 1997 to 46.5 percent Ukrainian population declined by almost 4 million, in 2004, while the percentage of heterosexually an average of 300,000 per year. Given the shrinking transmitted infections grew from 11.3 percent to size of the young adult groups and the persistent 32.4 percent. This change in the transmission pattern demographic decline, even modest increases in adult calls for more aggressive measures to curb the epi- prevalence rates could result in a strong long-term demic's spread in the general population. demographic impact. The young and women are hit hardest. The The study estimates that the number of new AIDS estimated HIV incidence rate for adults aged 15-49 in cases reached 13,700 in 2004, and annual AIDS 2004 was 0.25 percent, with the highest incidence deaths were approaching 10,000 even in the opti- rate of 0.69 percent in the 20-24 age group. Two- mistic scenario. By 2014, AIDS-related deaths will thirds of all new HIV infections are among young account for almost a third of all male deaths in the people aged 20-34, and 39 percent of the newly 15-49 age group and 60 percent of female deaths in infected are women, according to the 2004 medium that age group. In 2014, AIDS is projected to reduce scenario. Young women are more vulnerable than male life expectancy by 2-4 years: from 65.6 in the young men: the incidence rate for women 20-24 is hypothetical "no-AIDS" scenario to 63.4 (optimistic) 0.88 percent and 0.5 percent for men of the same and 61.6 (pessimistic) scenario. Similarly, a female age. By 2014, it is estimated that the 20-34 age group born in 2014 will be expected to live three years less will account for three-quarters of all new HIV infec- (to age 72.9) in the optimistic scenario and almost tions, half of which will be among women. five years less (to age 71.0) in the pessimistic one, instead of an expected 75.8 years in the "no-AIDS" HIV/AIDS is unevenly distributed across the scenario. A potentially catastrophic increase in country. Among the worst-affected regions are HIV/AIDS morbidity and mortality is expected in the those in the southeast oblasts of Donetsk, medium term if prevention measures fail. Also, sev- Dnipropetrovsk, Odesa, and Mykolaiv. Accounting eral complicating factors exacerbate the situation: for only a quarter of the total population of Ukraine, the demographic decline, the high prevalence of these regions will bear an estimated 36-43 percent of tuberculosis (TB) and sexually transmitted infec- accumulated HIV cases by 2014 and 31-38 percent of tions (STIs), and a generally weak health system. annual AIDS-related deaths. Donetsk Oblast will account for 13-19 percent of Ukraine's HIV infec- HIV/AIDS has become one of the major tions, followed by Odesa Oblast at 10-14 percent. obstacles to economic growth in Ukraine. The epidemic in these oblasts is unfolding against a AIDS affects all agents in an economy: households, backdrop of natural population decline, which is businesses, and the government, and its effects faster than the national average. By 2014, AIDS-relat- impact many of the economy's aspects: greater ed death rates in these two oblasts will exceed the mortality and morbidity; reduced labor supply, labor national average by a factor of 1.5-2.1. efficiency, and labor productivity; loss of investment ix in human capital and diminished returns to such estimated to be 41 million-629 million hryvnia (UAH) investment; increased health care spending and the by 2014. (This wide range is due to the high degree of loss of tax revenues; and decreases in public and pri- uncertainty about the future costs for both antiretro- vate savings and investment, among others. Reduced viral [ARV] and non-ARV medical treatment.) fertility among women infected with HIV amplifies the demographic decline and is responsible for longer- At the business level, the negative impact of term effects. Based on various plausible AIDS HIV/AIDS usually includes greater direct expendi- scenarios during 2004-14, the study found an expected tures for medical treatment, larger contributions to 1-2 percent reduction in the labor force due to the sickness/disability/death benefits, and the loss of epidemic. In addition, since the younger groups are investment in recruiting and training employees. most affected, the labor force losses will be felt for a long time. Furthermore, HIV/AIDS in Ukraine has a For the health sector, health budgets take a direct pronounced gender differential: the sharpest decline blow from the increased demand for hospital and in labor force participation is for females in the 15-19 outpatient services, with bed occupancy by age group. This decline is in addition to the labor HIV/AIDS patients stretching available resources. force reduction due to the underlying demographic Furthermore, the medical workforce itself is likely to trend, estimated to be a 10.4 percent fall by 2014 from be decimated by the epidemic, causing a clash the 2004 rate. In the worst-affected oblasts, the contri- between growing demand for professional care and bution of HIV/AIDS to labor force shrinkage is more a shrinking pool of medical professionals. pronounced: an additional estimated 2.7 to 3.6 percent for Donetsk and 2.2 to 4.2 percent for Odesa Oblasts. In the public sector, HIV/AIDS impacts both rev- enue and expenditures. The loss of productive time The phenomenon of children being orphaned to for income-generating activities lowers the tax base, HIV/AIDS is already taking a toll on both society and shifting more of the tax burden to the healthy households in Ukraine. According to the medium remainder, who may in turn respond by reducing scenario, Ukraine will have 42,000 dual orphans due their labor supply. Like the medical sector, the public to AIDS-related deaths of both parents by 2014. The and business sectors are also likely to lose their number of children who have lost at least one parent employees to the epidemic. Direct budget revenue to AIDS is projected to reach 105,000-169,000 by losses through the fall in employment due to 2014, depending on the scenario. Those children are HIV/AIDS, forgone income taxes, and unpaid pension at risk of impeded access to quality education, health and social security (temporary disability and unem- care, and even basic needs, which in turn puts them ployment) levies are estimated to reach 263-418 mil- at higher risk for unemployment, diseases, and lion UAH (in optimistic-pessimistic scenarios). At the poverty. Without adequate social assistance from the same time, the projected additional budget expendi- government and society at large for these children, a ture in 2014 will add 109-200 million UAH for perma- vicious cycle results. nent disability pensions due to HIV/AIDS, 20-35 mil- lion UAH in pensions from the Social Protection Medical expenses associated with treating HIV/AIDS Fund, 7-12 million in temporary HIV disability pay- and opportunistic infections can become catastrophic ments, and 3-8 million in AIDS orphan pensions. The at the household level, driving poor households total HIV/AIDS-related, government-funded addition- below the poverty line. This is particularly true in al benefits are estimated to be 139-255 million UAH countries such as Ukraine, with weak social and per year by 2014. private insurance systems. This study makes explicit assumptions about availability and price of antiretro- Other negative effects include an increase in the coun- viral therapy (ART) to devise cost scenarios for drugs try risk premium and possible effects on trade (both and hospitalization. Depending on the cost scenario in goods and services) and balance of payments. selected, total annual AIDS care expenditure is x A comparison of the non-AIDS scenario with opti- demic has serious consequences for Ukraine's socie- mistic and pessimistic outcomes shows that Ukraine ty and jeopardizes future development. In the medi- could experience a 1-6 percent reduction in the level um term, the study shows a significant impact on of output (gross domestic product, or GDP, in con- economic growth, investment and social welfare, life stant prices), a 2-8 percent reduction in total welfare, expectancy, and population growth. If current trends and a 1-9 percent reduction in investment. continue without effective control of the epidemic, the longer-term impact could be much more devas- Sectoral analysis suggests that labor-intensive sec- tating. The cost of inaction or ineffective action tors whose labor inputs suffer from the epidemic would be prohibitive. will be among the worst affected. In the sectoral analysis based on the computable general equilibri- The epidemic's distribution as shown in this study um (CGE) model, sectors such as those producing calls for attention to and effective targeting of youth, non-energy materials and processing metallurgy and females, and the worst-infected oblasts. Prevention metal were found to be the most affected, with and treatment programs must reach these target output falling by up to a third in the worst-case sce- groups, and messages and services must fit their nario. Given the relative share of these sectors in the needs. country's trade structure, the pessimistic scenario's fall of 40 percent in exports of these sectors trans- The current transmission pattern signals a need for a lates into a 5.5 percent fall in GDP, an 8 percent fall prevention strategy focused on harm-reduction pro- in total welfare, and a 9 percent fall in investment. grams as well as sex education for youth. Even though the mode of transmission is evolving towards The availability of ART provides hope for extending heterosexuals, IDUs still constitute the majority of life expectancy and healthy years to infected people those infected; special effort must be made to reach with treatment access. The study estimates that the them. cost of ART ranges from 353 million UAH under the optimistic case (with 50 percent of AIDS patients Donetsk, Dnipropetrovsk, Odesa, and Mykolaiv in need treated) and 52 million UAH under the pes- Oblasts are among the worst-affected regions. Given simistic one (with only 5 percent of patients getting the important role they play in Ukraine's economy, treatment). However, without adequate treatment, these regions should be treated with priority in such as in the pessimistic case, the cost for caring implementing HIV prevention, education, and treat- for AIDS patients is more than 80 percent higher ment measures. than that cost under the optimistic ART case. The analysis confirms that providing ART is a cost-effec- Due to data limitations, this study could only model tive measure under a range of scenarios with respect the impact of ARV treatment as one of the possible to the cost of treatment. In the low-cost scenario, as interventions. The study demonstrates that providing little as 419 UAH per person per year on average is ARV treatment can be cost-effective and that scaled required to prevent a new HIV infection. Even in the up treatment is the key to avoiding escalating health high-cost scenario, it costs an average of 762 UAH costs. ARV treatment needs to be complemented with per person per year to prevent one new infection, preventive education to curb the epidemic's spread. net of avoided hospitalization costs. The epidemic is still at the early stage in Ukraine, which means that timely, effective interventions, Policy Implications including the availability of ARV treatment, could In line with other international studies, the results halt and reverse the epidemic and reduce its impact from this study demonstrate that the HIV/AIDS epi- on socioeconomic development. xi CHAPTER 1 Introduction T his study was prompted by the rapid growth in percent, and among commercial sex workers (CSWs) Ukraine of the HIV/AIDS epidemic and its it is 22.2 percent. At the same time, mounting evi- threat to the general population and economy. dence shows that the wider population is increasing- According to estimates from the Joint United Nations ly at risk, mostly through heterosexual contacts. Programme on HIV/AIDS (UNAIDS), 360,000 (range: Potential catastrophic increases in HIV/AIDS morbid- 180,000 to 590,000) people were living with HIV/AIDS ity and mortality are likely in the medium term if in the country as of late 2003, with an adult (age 15- measures to curb the epidemic fail. Several factors 49) prevalence rate of 1.4 percent (range: 0.7 to 2.3 exacerbate the situation: persistent demographic percent). The exact number of infections is unknown, decline, a high prevalence of tuberculosis (TB) and but the officially registered new HIV cases reported sexually transmitted infections (STIs), and a health each year doubled over four years: from 6,216 in 2000 system needing reform. to 12,491 in 2004. In addition, 2000-2004 witnessed a four-fold increase in the annual official numbers of Access to antiretroviral therapy (ART) has been new AIDS cases and AIDS deaths. Between 1994 and very limited but is expanding. A Global Fund for 2005, Ukraine's epidemic was concentrated in sub- AIDS, Tuberculosis and Malaria (GFATM) pilot proj- populations, mainly injecting drug users (IDUs), with ect for 200 patients is being extended to 2,000, and the prevalence rate among pregnant women in urban on July 1, 2005, 1,950 persons were receiving ART areas still below 1 percent. However, a Barnett et al. (Ukrainian AIDS Center data). By the following (2001) study's warning to take very seriously the October, 2,866 AIDS patients were undergoing treat- probability of a "generalized"2 heterosexually trans- ment funded by GFATM. In April 2005, ART was mitted epidemic was reinforced in Feshbach and commenced in six of the country's worst affected Galvin (2005), who deduce that "all the evidence, areas: the oblasts of Donetsk, Dnipropetrovsk, however incomplete, suggests that a heterosexual Odesa, and Mykolaiv; the Autonomous Republic epidemic has certainly begun." The epidemic's ten- (AR) of Crimea; and Kiev city. dency to spill into the general population is reflected in official Ukrainian AIDS Center data: the share of Although HIV/AIDS is becoming a major obstacle to infections caused by intravenous drug use decreased economic growth in Ukraine, recognition of the need over 1997-2004 from 83.6 percent to 46.5 percent, to re-assess priorities and implement an effective, while the percentage of heterosexually transmitted anti-HIV/AIDS national strategy is growing. Losing infections grew from 11.3 percent to 32.4 percent. To the momentum of the recent economic recovery prevent the epidemic from becoming self-sustaining would be tragic after Ukraine's painful economic outside of the risk groups, such as IDUs, commercial sex workers (CSWs), and people with sexually trans- 2 "Generalized" and "concentrated" epidemics are defined as fol- mitted infections (STIs), effective prevention and lows: a concentrated epidemic has HIV prevalence in most-at-risk education are called for. subpopulations at 5 percent or higher and among pregnant women in urban areas below 1 percent. In a generalized epidemic, social networking in the general population is sufficient to sustain the High-risk groups remain the worst affected by the epidemic outside the most-at-risk sub-populations, and HIV preva- epidemic: 2004 sentinel surveillance data indicate lence among pregnant women is consistently above 1 percent. See http://data.unaids.org/Topics/Epidemiology/Manuals/EPP_ that the average prevalence rate among IDUs is 37.2 GeneralizedEpidemic_05_en.pdf. 1 transition since independence in 1991. Prior to investment in the economy through the increased recent positive developments, Ukraine experienced a consumption of health care and an increased coun- decade of severe political and economic instability try risk premium. In addition to the quantifiable and decline (World Bank 2004). An economic adjust- economic costs of disease, there are also intangible ment phase included extreme macroeconomic insta- losses from pain and suffering. bility and hyperinflation in 1993. By 1998, the offi- cially reported gross domestic product (GDP) had HIV/AIDS affects all agents in the economy: house- fallen to 40 percent of its 1990 level. Even if the holds, businesses, and the government. On both the degree of the actual fall is overestimated due to the household and business levels, its direct effects large size of the informal sector, there was a severe are due to the increased mortality and morbidity economic decline and genuine hardship for many (loss of years of healthy life, reduced labor supply, Ukrainians in the 1990s. The difficulties of the transi- and reduced efficiency of labor due to illness). tion stage are reflected in Ukrainian demographic HIV/AIDS leads to changes in labor force composi- statistics, with life expectancy falling for males and tion due to its heavier effect on the productive-age females from 66 and 75 years to 62 and 73 years, population. AIDS-related mortality disproportionate- respectively, between 1989 and 1997 (World Bank ly affects people during their productive years; it 2004). Furthermore, massive depopulation (by also affects women more than men. Morbidity almost 4 million during 1991-2003) through reduced reduces healthy life years, causing increased expen- fertility and out-migration has accelerated a growing diture on medical care with a negative effect on the share of the elderly population. Superimposed on income available for other purposes, including sav- these demographic trends, the HIV epidemic sup- ing and household investment in human capital. A presses already-low fertility even further, both deep- sick employee supplies fewer hours in the labor mar- ening and extending population decline. ket, and sickness makes anyone less efficient. When other household members must leave the labor force This study evaluates the broad economic effects of to care for a sick family member, labor supply drops the epidemic, delving beyond the costs of prevention again. Lower fertility ultimately produces a longer- and treatment. Both near- and medium-term (2004- term negative demographic effect and fewer people 2014) cost estimates were developed to inform poli- to contribute to the economy. Also, the number of cy makers of the potential costs of the epidemic dur- orphans rises with AIDS deaths, increasing the eco- ing this decade. To inform decision making on pre- nomic burden on the state and surviving family vention and treatment programs, the study highlights members. Last, medical expenses associated with the channels through which HIV/AIDS affects the the treatment of HIV/AIDS and opportunistic infec- national economy as well as households. tions may become catastrophic at the household level, driving marginally poor households below the HIV/AIDS has a direct impact on human health, an poverty line. This is particularly so in economies input in economic development and an indispensable with underdeveloped social and private insurance component of human capital.3 Infectious diseases markets. As a result, income inequality may worsen. influence economic activities and economic growth both directly and indirectly. At the first instance, dis- In the private sector HIV/AIDS affects employers ease has a negative impact on healthy life expectan- through the loss of investment in recruiting and cy. Early death and chronic disability result in the loss of future income and in medical care expendi- tures. The second effect includes reduced invest- 3 A positive correlation between health and economic growth has been established in Bloom and Sachs (1998), Bhargava et al. ment in one's own and one's children's education and (2001), Cuddington, Hancock, and Rogers (1994), Cuddington and health, especially in societies with high infant/child Hancock (1994), Robalino, Voetberg, and Picazo (2002), and Robalino, Jenkins, and Maroufi (2002) and analyzed in detail in mortality and high fertility (a behavioral quality- WHO Commission on Macroeconomics and Health (2001) and quantity trade-off). Third is a negative impact on Haacker (2004b). 2 training an employee who becomes disabled by Other negative effects include likely effects on AIDS. Loss of productive labor shifts the burden of trade (both in goods and services) and on balance of contributing to benefits, including the pension sys- payments. tem, to fewer healthy workers. This in turn may reduce benefits or healthy workers' labor This study uses several methods to detail the likely supply. impacts and costs of Ukraine's HIV/AIDS epidemic. Chapter 2 draws together available data to describe The public sector can also encounter losses through the current AIDS epidemic. Chapter 3 presents dif- its investment in recruiting and training of its labor fering demographic impacts based on various projec- force when its employees become sick. Public rev- tions with and without AIDS. Chapter 4 shows enues drop when workers reduce their contribution, specifically where the impacts on the labor force and either due to illness or to give care to family mem- government revenue will be greatest, and Chapter 5 bers, which in turn means fewer people paying estimates the cost of the epidemic and implications income taxes. of providing ART. Chapter 6 presents policy implica- tions. The methodology, assumptions, and models The health sector is likely to take a direct blow from and their results are detailed in the annexes. increasing demand for medical care and reduced numbers of health workers due to the epidemic. 3 CHAPTER 2 The HIV/AIDS Epidemic in Ukraine: Status and Trends U kraine's HIV/AIDS epidemic has been spread- number of new AIDS cases registered during March- ing at an alarming rate for the past few years, October 2005 is almost as high as the official number with officially registered new HIV cases per of all AIDS cases registered in 2004 (2,745). A similar year reaching an all-time high of 12,595 in 2004, a 25 trend is seen for AIDS deaths: 1,205 deaths over percent increase from 2003. Each day brings 32 new March-October 2005 compared to 1,775 for all of HIV diagnoses and 8 AIDS deaths among Ukrainians. 2004. Figure 2-1 illustrates the cumulative growth of Data from the Ukrainian AIDS Center (March 1, 2005) officially registered HIV cases over 1987-2004. These indicate that the cumulative number of registered data suggest that the epidemic is accelerating. HIV cases was 76,875 Ukrainian nationals, including 6,055 children as well as 314 foreigners. Those data The recent increase in the number of new registered also showed that 9,065 adults and 329 children devel- HIV infections was not driven by improved testing or oped AIDS, and 5,504 adults and 156 children died. more tests (Figure 2-2), an issue discussed more Official HIV prevalence based on the registered cases extensively below under "Non-uniformity across is 115.4 per 100,000. Newer data (October 1, 2005) Regions and Sub-populations." It is important to note place the cumulative number of officially registered that the number of tests conducted has not increased HIV-positive persons at 84,437, total AIDS cases at appreciably since 1997. Annex 1 Table A1-1 reports 11,757, and AIDS deaths at 6,865. This marks an addi- officially registered new HIV and AIDS cases and tional 7,248 newly registered HIV cases, 2,363 AIDS AIDS deaths for 1987-2004, and Table A1-2 presents cases, and 1,205 AIDS deaths in seven months. The national HIV serosurveillance data by major category. The Changing Pattern Figure 2-1. Cumulative Reported Cases of HIV Infection, 1987-2004 of Transmission of the HIV/AIDS Epidemic 80,000 74,856 The leading modes of HIV 70,000 62,672 transmission in Ukraine are 60,000 through intravenous drug use, 52,356 followed by heterosexual 50,000 43,600 transmission. However, official 40,000 data suggest that the share of 36,600 30,000 transmission mode is shifting 24,561 30,388 and that the epidemic has 20,000 started to spread outside the 15,986 10,000 high-risk groups through het- 1,673 183 7,073 erosexual transmission. HIV 0 entered the population of 1987-94 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 intravenous drug users and Source: Ukrainian AIDS Center. expanded among them during 4 1995-98 through the sharing Figure 2-2. Total Numbers of HIV Tests and Officially Registered HIV Cases, of contaminated needles and 1987-2005 equipment. It is spreading 8,000 14,000 increasingly through hetero- 7,214 12,491 sexual contact. The share of 7,000 6,781 12,000 IDUs among all new HIV vic- tims dropped from 63 percent 6,000 5,518 10,009 10,000 in 2000 to 46.3 percent in 5,481 8,934 5,000 8,590 8,761 2004, while the share of 4,881 Persons 8,000 infection through heterosexu- 4,000 4,027 7,009 al contact increased from 23 Thousand 3,515 5,830 6,216 6,000 5,422 3,000 3,096 percent to 32 percent. During 2,638 2,465 2,210 2,303 4,000 2000-2004, the epidemic 2,553 2,507 2,000 2,087 2,119 2,151 broadened, with the percent- 2,000 age of cases growing on aver- 1,000 1,499 579 age by 30 percent per year 81 55 48 40 34 45 51 44 0 0 from heterosexual contact 1985 1990 1995 2000 2005 and by 32 percent per year from vertical transmission Tested Official HIV+ (mother-to-child transmission Source: Ukrainian AIDS Center. or MTCT). By comparison, the cases of HIV infection among the IDUs grew on average by than females. The increased share of females from 10 percent per year in 2000-2004. Male-to-male 36.5 percent to 42 percent of total HIV infections over sexual contacts account for an insignificant number 2000-2004 signals that the epidemic's generalization of reported cases (see Figure 2-3 on page 6 and may have commenced (Annex 1 Table A1-5). Annex 1 Tables A1-2 and A1-3). One can visualize Ukraine's epidemic as the superimposition of three Ukrainian AIDS Center data indicate that the waves: an explosive spread among the IDUs, a slow- HIV/AIDS epidemic is unfolding in all regions, albeit er but broader wave through heterosexual contacts, non-uniformly (see Annex 1 Table A1-5). The worst- and--as a consequence of both--a third component affected regions in terms of registered HIV preva- through MTCT. lence are the oblasts of Donetsk (16,161 cases), Dnipropetrovsk (13,868), Odesa (10,855), and Mykolayiv (4,986); the AR of Crimea (4,976 cases); HIV Sentinel Surveillance Data and Kiev city (3,144 cases). Most of those infected Serosurveillance data (Annex 1 Table A1-2) show that are aged 20-29. In terms of incidence rates, HIV seroprevalence among all tested increased from Dnipropetrovsk Oblast is the worst affected with 0.75 percent in 2003 to 0.92 percent in 2004. 59.81 new cases of per 100,000 in 2004, closely fol- Seroprevalence increases are also observed among lowed by the oblasts of Odesa (58.92), Mykolayiv pregnant women, reaching 0.34 percent in 2004, and (57.49), and Donetsk (52.86); Sevastopol city (50.21); donors, reaching 0.13 percent that same year. and the AR of Crimea (37.2). Females aged 15-30 are more likely to be infected through heterosexual contact than males of the same Most-at-Risk Populations (MARPs) age, and almost half (47 percent) of the reported HIV cases are among females of the most active reproduc- The epidemic so far in Ukraine is still concentrated tive age, 20-29. Above the age of 30, males are more among certain population groups that are at higher likely to be infected through heterosexual contact risk of infection: IDUs, commercial sex workers 5 Figure 2-3. Leading Modes of HIV Transmission, 1987-2004 8,000 7,448 7,000 6,516 6,000 5,778 5,000 4,815 4,360 4,587 4,041 4,000 3,771 3,881 3,964 3,000 3,043 2,273 2,499 2,000 1,385 1,323 1,427 1,885 1,371 1,830 1,024 1,007 914 1,000 709 727 433 527 389 161 260 378 295 314 92 236 196 202 173 231 19 0 294 1987-95 1996 1997 1998 1999 2000 2001 2002 2003 2004 IDU Heterosexual MTCT Unknown Source: Ukrainian AIDS Center. (CSWs), men who have sex with men (MSM), people The highest level of HIV infection among IDUs (59 with sexually transmitted infections (STIs), and percent) was recorded in Simferopol, more than dou- prisoners. Surveillance data on these groups enables ble that found in earlier surveys (see Annex Table forecasting the epidemic's future spread among them A1-7). Seroprevalence remains consistently high and extrapolations to estimate prevalence in the gen- (58.3 percent) among IDUs in Odesa Oblast. Among eral population. The shape of epidemic profile will IDUs in Donetsk it remained relatively unchanged depend on the size of the risk groups and interac- over 2000-2004, at about 41.6 percent, but this does tions between them and sexual partners from not indicate stabilization of the epidemic: Donetsk outside these groups. has the highest rate (55.6 percent) of new cases among very young (15-19) IDUs, perhaps indicating As part of the HIV monitoring and surveillance pro- the rapid spread of infection among teenage IDUs in gram, HIV sentinel surveillance has been conducted this region. Volyn Oblast posts a stable yet high HIV since 1999 among the most-at-risk populations. The seroprevalence rate among IDUs of 32.8 percent. results among these groups suggest that the number of reported HIV cases is grossly underestimated The Poltava region has seen a reduction in surveyed based on the estimated HIV prevalence among the HIV prevalence rates of IDUs, but more than a third MARPs (Artyukh et al. 2005a). of new cases (36 percent) occur in the 15-19 age group. The corresponding number for Odesa Oblast Injecting Drug Users is 26.1 percent despite the generally higher sero- prevalence rate among the IDUs. In Kharkiv the HIV The survey of HIV seroprevalence among IDUs indi- prevalence among IDUs declined somewhat lately cates that in eight regions studied, seroprevalence and now stands at 14 percent, while the Sumy Oblast ranged from 10-59 percent, confirming that IDUs, fol- IDU seroprevalence indicator (11.6 percent) is the lowed by CSWs, were a major driving force behind lowest in the eight regions. the epidemic in Ukraine. 6 Commercial Sex Workers and means to monitor HIV's spread in the general popula- STI Patients tion. Mandatory screening of blood donors for HIV has been in place since the 1989 Ministerial Order by HIV seroprevalence surveys were conducted among the Ministry of Health. Donors in Ukraine receive a CSWs in five Ukrainian regions, where seropreva- payment for donating blood (apart from donating to lence ranges from 11 percent in Kherson city to relatives) and may be motivated to donate by the 31 percent in Odesa city (Artyukh et al. 2005b; see payment. During 1998-2004, the share of infected Annex 1 Table A1-8). The rising share of HIV trans- donors rose from 0.07 percent to 0.13 percent. In missions through heterosexual intercourse has been 2004 the highest levels of HIV seroprevalence among driven in part by infections among CSW partners, donors were observed in Mykolayiv, Odesa, many of whom inject drugs. The average HIV preva- Dnipropetrovsk, Chernigiv, Donetsk, and Kiev lence rate among all tested CSWs was 18.7 percent Oblasts, all of which are the regions with the highest in 2002, 22.2 percent in 2004, and 8-32 percent in HIV prevalence apart from Chernigiv, which is below 2005 (Ukrainian AIDS Center, 2005). the national average incidence level (see Annex 1 Table A1-6). HIV cases are also being reported increasingly in STI patients. Serosurveillance of this group indicates Seroprevalence among pregnant women also indi- widely varying rates by region and year. cates epidemic trends in the adult population. According to Ukrainian AIDS Center data, the sero- Men Who Have Sex with Men and Other prevalence rate among pregnant women rose from Sexual Transmission 0.12 percent in 1998 to 0.34 percent in 2004, with the The lack of data for the group of men who have sex highest rates observed in Mykolayiv, Dnipropetrovsk, with men results from the limited accessibility to Donetsk, Odesa, Chernigiv, and Kiev Oblasts. The this group by survey programs. However, serosur- first four oblasts in this list have the highest IDU veillance among the MSM groups conducted on a rates. Since 2000 measures have been implemented small sample in two cities reveals high prevalence to prevent MTCT, targeting all pregnant women who rates (7 out of 25 tested in Odesa and 3 out of 22 agree to undertake voluntary HIV testing. In 2003, tested in Simferopol: Amdzhadin et al. forthcoming). 15.9 percent of infants born to HIV-positive mothers The sentinel surveillance report shows the preva- tested positive, a 43 percent reduction compared lence rate among MSMs was between 10 and 30 per- to 2001. cent in 2004 (Ukrainian AIDS Center). Non-uniformity across Regions and Data reported in Annex 1 Table A1-9 also suggest Sub-populations that sexual transmission of HIV infection is rising. The above indicates that the epidemic spreads non- uniformly across regions and population groups, Prisoners with the major explanatory factor being the geo- Increasingly, HIV cases are also being reported among graphic distribution of intravenous drug use. At the prisoners. From 1987 to the end of 2005, 14,998 new same time, the total number of HIV tests conducted cases of HIV were diagnosed in the penitentiary sys- among high-risk groups and other categories tem in Ukraine, and 999 prisoners had developed depends on the local administration and policies that AIDS. Among the prisoners tested for HIV, 5.5 percent also differ from oblast to oblast. The Ukrainian AIDS were HIV positive in 2002, and that rate increased to Center undertook epidemiological studies in 2004 to 9.4 percent in 2005 (State Department of Prisons). better understand the role of IDUs in shaping the national epidemic. The study concluded that official Low-Risk Groups data on reported HIV cases are closely linked to HIV-testing practices. The targeted testing of IDUs Official data on low-risk groups such as blood had been relatively stable and somewhat reduced in donors and pregnant women provide an important 7 Figure 2-4. Total Number of HIV Tests Conducted, by Category, 1994-2004 In thousands 1,800 1,600 IDUs STIs Pregnant Donors 1,400 1,200 1,000 800 600 400 200 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: Ukrainian AIDS Center. recent years, while prenatal HIV testing had Galvin (2005) discuss the debate on whether general- increased (see Figure 2-4). As a result, the increased ization has started in Ukraine and deduce that a het- share of females among new HIV cases and the erosexual epidemic has certainly begun. reduced share of IDUs may be partially attributed to the change in the composition of the tested popula- In sum, HIV infection is spreading at an increasing tion.4 Examination of data from Donetsk and Odesa rate in Ukraine, with injecting drug use as the lead- Oblasts on those infected with HIV who passed the ing mode of transmission, but the share of the sexual virus to their sexual partners revealed that in 50-60 transmissions is increasing, as is the share of MTCT: percent of such cases, primary exposure of index both signs that HIV infection is starting to penetrate patients to the virus occurred through sharing con- the lower-risk population. While still low at the taminated drug injection equipment. Hence, the het- national level, the officially reported HIV incidence erosexual transmission to the sex partners can be among pregnant women is 0.6 percent in Donestk traced back to the equipment use by the index and Odesa, 0.7 percent in Dnipropterovsk, and 0.8 patient in more than half of cases. percent in Mykolaiv Oblasts (Annex 1 Table A1-6). If this trend continues and further spread is sustained Based on the above, the Ukrainian AIDS Center in the lower-risk group through heterosexual con- study concluded that the epidemic continued to be concentrated among IDUs and their sexual partners. However, analysis by national and international 4 Note that the registered new HIV cases among pregnant women experts (Barnett et al. 2001; DeBell and Carter 2005) have increased from 0.12 percent of those tested in 1998 to 0.34 percent in 2004. This increase cannot be attributed to any change indicates that the epidemic may be on the brink of in testing practice and therefore reflects the spread of infection spilling into the general population. Feshbach and among females. 8 tact, the epidemic may become generalized, at least National AIDS Committee was established in 1992 in the worst-affected oblasts. and replaced by the National AIDS Control Coordinating Council under the Cabinet in 1999. The As mentioned above, there are significant regional Ukrainian National Program on HIV/AIDS disparities in HIV prevalence in Ukraine. Among the Prevention for 2004-2008 was prepared, and vari- worst affected are the industrialized oblasts of ous prevention programs are being implemented. As Donetsk, Dnipropetrovsk, Odesa, and Mykolayiv; AR part of the public health system response, 35 region- Crimea; and Kyiv and Sevastopol cities. Sentinel sur- al AIDS centers are operating and provide preven- veillance studies of seroprevalence among the IDUs tive, diagnostic, medical, and counselling services confirms that the southern region (Simferopol and through activities coordinated by the Ukrainian AIDS Odesa cities) has the highest HIV prevalence rates Center. At the district level, similar services are pro- among IDUs. High seroprevalence rates were also vided in district hospitals, through infectious dis- confirmed in the western region (Volyn Oblast). eases departments and consultation clinics. As of Seroprevalence rate estimates based on sentinel sur- October 2005, 2,866 patients were receiving Highly veillance among STI patients vary from 1 percent in Active Anti-Retroviral Therapy, but that is only about Kharkiv in northeast Ukraine to 9 percent in Odesa. 15 percent of those needing it (Ukrainian AIDS Center, 2005). Ukraine's Response to HIV/AIDS The acceleration of the HIV/AIDS epidemic over Within the European region, Ukraine is the worst- 2002-04 requires re-estimation of the magnitude of affected country, with the highest adult HIV preva- the epidemic and its possible socioeconomic impact, lence rate. Public awareness of HIV/AIDS has been taking into account new data on availability of ARV increasing, and in recent years, the government, therapy, reduction in MTCT rates, and new estimates nongovernmental organizations, and international of the size of the IDU group as of January 2005. The agencies have improved the national response. The results of such analysis are presented in Chapter 3. 9 CHAPTER 3 Demographic Forecast under the HIV/AIDS Epidemic U nderlying economic and demographic condi- could counter negative demographic pressures and tions determine the impact of the HIV/AIDS reduce the epidemic's impact. epidemic. In Ukraine, the epidemic exacer- bates negative demographic trends with adverse Reductions in life expectancy reflect the hardships depopulation and eroding health. of transition. Difficult socioeconomic conditions, a decline in living standards, and a sharp reduction in The current demographic trend in Ukraine is charac- income in the last decade and a half all negatively terized by massive depopulation through reduced affected demographics. From 1990 to 1998 Ukraine fertility, increased mortality, and out-migration. The experienced a 60 percent fall in GDP. Additional current fertility rate of 1.1-1.2 births per woman is impacts of the HIV epidemic on Ukrainian demo- just a half of the replacement level rate of 2.2 (IDSS graphics are examined in the following sections. 2005). The number of births dropped over 1991-2003, from 630,800 to 408,600. With deaths exceeding Analysis at the National Level births by a factor of two, Ukraine's population decreases by more than 300,000 persons per year. This study constructed a baseline "no-AIDS" demo- Natural depopulation was 3.9 million in 1991-2003, a graphic projection of the Ukrainian population until trend projected to continue (Derzhkomstat 2005). 2014 and added three HIV/AIDS epidemic scenarios Ukraine's reduction of population size is not unique. (optimistic, medium, and pessimistic). The projec- Many European countries face similar birth rate tion period is 1994-2014. These scenarios differ in declines, but Ukrainian depopulation is among the their assumptions about the size and dynamics of the fastest in the world and, unlike in developed coun- most-at-risk populations, yielding different estimates tries, is accompanied by deterioration of health sta- of adult prevalence rates. In the optimistic scenario tus, increased mortality, and reduced life expectancy. the adult HIV prevalence rate peaks at 2 percent in Worsening health status is more common among 2010; it peaks at 2.48 percent in 2009-10 in the medi- males due to noncommunicable diseases, mental ill- um scenario; and it rises continuously reaching ness, stress, and alcohol-related accidents and 3.5 percent in 2014 in the pessimistic scenario. injuries (Brainerd and Culter 2005). Reduction in the MTCT rate (15.9 percent in 2003) is faster in the optimistic scenario (to 10 percent in Poor health status has resulted in increased mortali- 2004 and then to 5 percent in 2014) than in the medi- ty in practically all age groups except children. The um (gradual reduction to 5 percent by 2014) and pes- most significant losses are among those of working simistic (gradual reduction to 10 percent in 2014) age. High mortality in this group is the main reason scenarios. Availability of ART to those who need it for low life expectancy in Ukraine. On average, male increases from 1 percent in 2004 to 30 percent in (female) life expectancy is 11-12 (7-8) years less than 2010 and further to 50 percent in 2014 in the opti- in developed European countries. Life expectancy at mistic scenario; to 5 percent in 2005, to 10 percent birth in 2003 was 62.3 years for males and 73.5 for by 2010, and remaining there until 2014 in the medi- females. Improving health status and life expectancy um one; and to 5 percent in 2005 and remaining 10 there until 2014 in the pes- Figure 3-1. Forecasted Number of Those Infected with HIV, 1994-2014 simistic one. The study also constructed three cost scenar- 900 ios (referred to as A, B, and C) 800 Pessimistic for ART. Details of the model- 700 Medium Optimistic ing methodology and assump- 600 tions are in Annex 2.5 500 400 Thousand 300 HIV/AIDS Epidemic 200 Projections: 100 Major Findings 0 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 In the medium epidemic sce- nario, the study projects about Source: Authors' calculations. 477,000 Ukrainian adults were living with HIV/AIDS in 2004 (range: 448,000-491,000), which corresponds to an adult prevalence rate of 1.8 Second, despite the fact that most of those infected percent (range: 1.7-1.9 percent). In this scenario, the with HIV or dying from AIDS are males, the share of total number of infections will peak at 640,700 in females among the infected will grow under all three 2009; by 2014, the total will range from 478,500 to scenarios. In the medium scenario, the share of AIDS 820,400 (optimistic-pessimistic).6 Figure 3-1 presents deaths in all adults (15-49) will increase by a factor these data graphically while Annex 2 Tables A2-1 and of 2.7 for males and 3.5 for females during 2004-14. A2-2 provide more detailed forecast results. In 2014, about a third (32.3 percent) of all deaths in adult males and almost two-thirds of all deaths in Based on the forecast, the importance of AIDS as a adult females will be caused by AIDS (Figure 3-3). cause of death will increase, especially for younger These predictions reflect our modeling assumptions age groups. The share of AIDS deaths among total about rising female HIV transmission rates, based on deaths in Ukraine in 2004 was 2.3 percent. In the the infection pattern observed in 1995-2004. While optimistic, medium, and pessimistic forecasts, it is HIV/AIDS is contributing to extremely high mortality projected to grow by 2014 to 4.8 percent, 7.9 percent, rates in young and middle-age Ukrainian men, the and 8.6 percent, respectively (Figure 3-2 on page 12). relative importance of AIDS as a cause of death is more significant for females. Consequently, the epi- HIV's spread leads to growth in premature deaths, dis- ability, co-infection of opportunistic diseases and TB, 5 This study's methodology for projecting HIV/AIDS dynamics is and reduction in life expectancy. The risk of falling based on Schwartlander et al. (1999), UNAIDS (2002), etc. The sick and dying of AIDS varies by age and sex. First, internationally used method of back-projection (Becker, Watson, and Carlin [1991], Becker and Motika [1993], Becker and AIDS victims are mainly young people, so the most Marschner [1993], Becker and Egerton [1994], etc.) for estimating significant changes will occur in the structure of mor- unobserved past incidence of HIV infection and to predict future AIDS incidence is of limited applicability to countries with poor tality of the working- and childbearing-age population. AIDS incidence data. In particular, the share of AIDS-caused deaths among 6 total number of deaths in the 15-49 age group will Comparing our findings with those of Barnett et al. (BW): the BW adult prevalence rate estimate for 2005 was 1.47 percent in the increase from 13.2 percent in 2004 to 41.4 percent in optimistic and 2.92 percent in pessimistic scenarios. Our opti- 2014 (Figure 3-3 on page 12). AIDS will gradually mistic adult prevalence rate of 1.76 percent in 2005 is higher than BW's optimistic, while our pessimistic estimate of 2.09 percent is become the leading cause of death among younger lower than their pessimistic estimate: the band for our estimates is adults. This will deplete the young and productive narrower. BW's predicted adult prevalence rates for 2010 are 1.97 percent in the optimistic and 4.91 percent in pessimistic scenarios. population, with people aged 30-39 suffering the most. The difference in outcomes results from the difference in inputs into the Spectrum model as new evidence has become available. 11 Figure 3-2. Forecasted Share of AIDS Deaths in Total Number of Deaths, 24 age group. Two-thirds of all 1994-2004 new HIV infections are among young people aged 20-34, and 39 percent of the 10.0 newly infected are women, 9.0 Pessimistic according to the 2004 medium 8.0 Medium scenario. Young women are 7.0 Optimistic more vulnerable than young 6.0 men: the incidence rate for women 20-24 is 0.88 percent 5.0 Percent and 0.5 percent for men of the 4.0 same age. By 2014, the 20-34 3.0 age group is estimated to 2.0 account for three-quarters of 1.0 all new HIV infections, half of 0.0 which will be among women. 1994 1996 1998 2000 2002 2004 2006 2008 2010 2010 2014 This evidence mimics the find- ing on other STIs (Mavrov and Source: Authors' calculations. Bondarenko 2002), where the ratio of female to male infec- demic's impact is likely to be more striking among tions is 5:1 among those aged 15-17 and more than females. AIDS may become the leading cause of 2:1 for those 18-20. deaths in females aged 15-49 by 2010. The medium scenario indicates that the peak in The modelling results suggest that the hardest hit are annual AIDS deaths (59,000) will occur in 2014 young and female. The estimated HIV incidence rate because that is the final year of the model forecast for adults aged 15-49 in 2004 was 0.25 percent, with with accumulated AIDS deaths exceeding half a mil- the highest incidence rate of 0.69 percent in the 20- lion (510,900). Whether AIDS mortality will continue Figure 3-3. Forecasted Share of AIDS Deaths in Total Adult (15-49) Deaths, 2004-2014 70 60 Total Male Female 50 40 Percent30 20 10 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 12 to rise beyond the study forecast Figure 3-4. Forecasted Accumulated AIDS Deaths, 1994-2014 horizon will depend on whether the epidemic is curbed. In the 600 optimistic scenario, which assumes a slow progression from 500 Pessimistic HIV infection to AIDS, more Medium reduction of MTCT, and better 400 Optimistic access to ART, the cumulative number of AIDS deaths is well 300 below half a million: 301,300. In Thousand the pessimistic scenario, the num- 200 ber is 526,400 (Figure 3-4). The peak in the annual AIDS deaths 100 among children occurs in 2007 and equals 721 under the medium sce- 0 nario, and 525 under the opti- 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 mistic one. The pessimistic sce- Source: Authors' calculations. nario predicts AIDS childhood deaths will grow continuously, reaching 960 in 2014. to shrink even without HIV/AIDS. Under the assump- HIV/AIDS Impact on Life Expectancy tions of the "no-AIDS" scenario, the population The spread of HIV/AIDS and the related increase in would be 44.2 million in 2014, a reduction of 7.6 mil- mortality will have a negative impact on life lion from 51.8 million in 1994, the projection's base- expectancy in Ukraine. The maximum reduction in line. The AIDS epidemic will accelerate depopula- total life expectancy resulting from the epidemic will tion, likely causing an additional decrease of 0.3-0.5 be observed in the year of the highest AIDS mortali- million, leaving 43.7 million-43.9 million in total pop- ty, 2014. It is assumed that in the absence of AIDS, ulation (Figure 3-6 on page 14). life expectancy would increase by the end of fore- cast period to 65.6 years for Figure 3-5. Forecasted Life Expectancy at Birth, 1994-2014, in Years males and 75.8 years for females, but the epidemic 69 will likely bring life 69 expectancy down to 61.6-63.4 for males and to 68 71.0-72.9 for females, 68 depending on the scenario. This equals a reduction of 67 Years 3.2-4 years for males and 67 Pessimistic 2.9-4.8 years for females Medium (Figure 3-5). 66 Optimistic 66 The demographic forecast indicates that the Ukrainian 65 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 population would continue Source: Authors' calculations. 13 Figure 3-6. Forecasted Total Population, 1994-2014 The interaction between Ukraine's Millions epidemic and age dynamics and their combined effect are ambigu- 53 ous. On one hand, younger people 52 Pessimistic infected with HIV will withdraw 51 Medium from the labor force at some stage, Optimistic 50 increasing work load on the remaining working population. 49 The adults aged 15-59 are 48 estimated to total 28.4 million in 47 the "no-AIDS" scenario, but this 46 figure drops 300,000-500,000 45 depending on the scenario: the 44 total is 28.1 million in the opti- mistic and 27.9 million in the pes- 43 simistic scenarios. The largest 1994 1995 1996 1997 1998 1999 2000 2001 2002 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 losses are incurred by those 30-39 Source: Authors' calculations. and are expected to become par- ticularly acute in 2010 when nega- Figure 3-7. Age-Gender Composition of Population, "No-AIDS" and "AIDS Medium" tive demographic trends and popu- Scenarios, 2014 lation aging accelerate and the Thousands working-age group shrinks even further. In particular, in the "no- 80+ 80+ AIDS" scenario, the proportion of 75-79 75-79 people 60 years and older in the 70-74 70-74 total population will increase from 20.9 percent in 2004 to 21.3 per- 65-69 65-69 cent in 2014. With AIDS, it will 60-64 60-64 reach 21.5 percent at the end of 55-59 55-59 the forecast period (Figure 3-7). 50-54 50-54 45-49 45-49 On the other hand, the size of the 40-44 40-44 20-30 age group and this group's 35-39 35-39 relative share in the total popula- 30-34 tion are both predicted to fall after 30-34 2010, when the baby-boom effect 25-29 25-29 of the early 1980s runs out. The 20-24 20-24 number of people susceptible to 15-19 15-19 HIV--younger age groups--will 10-14 10-14 decline. HIV prevalence may 5-9 5-9 reduce or remain the same. 0-4 0-4 Further spread of the epidemic is possible if the infection generalizes 3000 2500 2000 1500 1000 500 0 500 1000 1500 2000 2500 3000 and is no longer contained within the higher-risk groups. In such case Males (No-AIDS) Females (No-AIDS) Males (AIDS Medium) Females (AIDS Medium) the entire population becomes at- risk, and the epidemic's potential Source: Authors' calculations. 14 devastation becomes Figure 3-8. Forecasted Total AIDS Orphans,* 1994-2014 worse as it penetrates families and causes 180 longer-term demographic 160 damage. How the epidem- 140 Pessimistic ic's impact will flow 120 Optimistic through different channels 100 is discussed next. 80 Thousand First, premature death 60 among many males of 40 reproductive age will have 20 a direct negative effect on 0 the number of male part- 1994 1999 2004 2009 2014 ners available to form * Children orphaned as a result of death due to AIDS of one or both parents. families. This reduction together with traditional Source: Authors' calculations. preferences for legal mar- riages when making fami- ly-planning decisions will slow the family formation ment as children from complete families. All of this process, reducing birth rates. contributes to social inequality and instability. Second, the economic burden may change, as the The HIV/AIDS epidemic affects not only those infect- dependency ratio (the ratio of the economically ed, but also their families, households, and society at dependent part of the population, either too young large. It impacts human resources not only in terms or too old to work, to the productive part of the pop- of quantity, but also quality. Not only does the size of ulation) increases. The economic burden on females the labor force decrease due to increased mortality will increase, as a gender role shift already observed among the younger age groups, but labor productivity in Ukraine during the period of economic downturn falls as well. Also, if the epidemic becomes general- with women taking over as main family breadwin- ized, it will hinder the process of human capital ners. In this context, AIDS will make the burden of accumulation by reducing considerably both the time responsibility for family survival on females even available to recoup investment in human capital and harder (SIFYA 2004; UISS 2002). rates of return to such investment. This study demon- strates that the HIV/AIDS epidemic in Ukraine leads Third, premature parent deaths will increase the to both quantitative and qualitative labor force losses. number of orphans. The study forecast suggests that in 2014 the number of AIDS orphans will reach Analysis at the Regional Level 105,100 in the optimistic and 169,300 in the pes- simistic scenarios (Figure 3-8). Dual orphans and The geographic distribution of HIV/AIDS in Ukraine semi-orphans may receive limited parental support is non-uniform, and regional demographic patterns or must be cared for by the state, and their access to vary. To study regional variation in the demographic quality education and human development is imped- impact of the HIV/AIDS epidemic, separate ed. For instance, young adults' access to higher edu- demographic and epidemic forecasts were built for cation in Ukraine correlates strongly with the finan- Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa cial and social status of their family. AIDS orphans Oblasts, the worst-affected regions. Separate epi- will live in financially disadvantaged households, demic forecasts are constructed using two scenarios often unable to achieve the same educational attain- (optimistic and pessimistic) for all these oblasts but 15 Dnipropetrovsk, where only a pessimistic scenario is The adverse demographic situation in these oblasts presented, due to limited surveillance data. is combined with both high prevalence of intra- venous drug use and HIV infection. HIV prevention Regional differentials in socioeconomic development programs among high-risk groups are actively imple- and social environment are linked to the demographic mented in Odesa and Mykolayiv Oblasts and to a trends and the HIV/AIDS profile. Dnipropetrovsk and much lesser degree in Dnipropetrovsk Oblast. Donetsk Oblasts are located in the southeast, boasting the highest economic potential, high levels of eco- The study predicts that most of the demographic nomic activity, and high population density. They are losses associated with the epidemic will accrue to highly industrialized and urbanized oblasts with envi- these four oblasts, with a subsequent negative ronmental degradation. At the same time, they are fac- impact on the regional economies. The number of ing the most unfavorable demographics in Ukraine, people infected with HIV in 2014 in these oblasts is characterized by a considerable loss of population in predicted to constitute 36-43 percent of Ukraine's the 1990s. In particular, over the last decade the natu- total HIV cases, while only a quarter of its population ral annual population decline in Dnipropetrovsk resides there. The predicted numbers of those infect- Oblast was 30,000 persons and in Donetsk Oblast, ed in 2014 is 32,500-44,200 in Mykolayiv; 48,900- over 45,000 (8.7 and 10.8 per 1,000 population, respec- 116,100 in Odesa; 85,300 in Dnipropetrovsk; and tively). Low life expectancy, high death rates among 92,200-105,600 in Donetsk (Annex 2 Table A2-5). The the working-age population (especially in males, contributions of Donetsk and Odesa Oblasts to the driven by high rates of accidents, poisonings, and total number of infections will increase from 11-12 trauma), and very low birth rates accompanied by the percent and 5-7 percent, respectively, of the national highest abortions rates in Ukraine, are all found in total in 2004 to 13-19 percent and 10-14 percent, Dnipropetrovsk and Donetsk Oblasts. respectively, in 2014. Odesa and Mykolayiv Oblasts are in the southern Annual AIDS deaths are predicted to reach 2,000- region, an industrial, agrarian, and recreational 3,000 in Mykolayiv Oblast and 6,000-9,000 in Donetsk region with an average level of socioeconomic devel- Oblast in 2014 (Annex 2 Table A2-6). The accumulat- opment and some degree of environmental degrada- ed AIDS deaths will increase considerably in the four tion. The population has a mixed ethnic composition oblasts and will account for 23-30 percent of the that is changing through migration. Low life national total in 2014 (Annex 2 Table A2-7). expectancy, high death rates from external causes and infectious diseases (TB, above all), and average While the greatest loss of lives to AIDS are in the birth rates characterize these oblasts' demographics. working-age groups, the absolute number of AIDS Over the last decade, annual natural decline in the deaths does not reflect the real gravity of AIDS' con- population constituted almost 7,000 persons in tribution to increased mortality: the latter depends Mykolayiv and 14,500 in Odesa Oblasts (or 7 per on the age-gender composition of a regional popula- 1,000 population in both). tion. Thus, to compare AIDS-related mortality in the working-age populations across oblasts, the study Factors with a negative effect on the social and used both the share of AIDS deaths in total number demographic situation in these oblasts include: of deaths in the 15-59 age group and mortality (deaths per 100,000 population in the relevant age A high share of employment in industrial sectors group) as indicators. The analysis shows that by with unsafe labor conditions and high risk of 2014, AIDS will account for about a third of all trauma; deaths in the working-age group (Table A2-8). Environmental degradation; and Mykolayiv and Odesa are hardest hit (discounting Dnipropetrovsk for which no optimistic scenario A high crime rate aggravated by the inflow of was drawn) with the highest AIDS death rates per refugees, migrants, and marginalized groups. 16 100,000. All four oblasts are sig- Figure 3-9. Estimated AIDS-Related Mortality in the Working-Age (15-59) nificantly above the national Population in Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa AIDS death rates among the Oblasts and Ukraine, 2014 working-age population by an Per 100,000 estimated factor of 1.4 in Dnipropetrovsk, 1.5-1.7 in 410 Donetsk, 1.7-1.8 in Odesa, and 373.2 370.8 1.7-2.1 in Mykolayiv Oblasts in 360 332.6 2014 (Figure 3-9). 313 Pessimistic 310 Optimistic The epidemic's impact on life 260 242.9 expectancy for four oblasts is 215.9 221.2 shown in Figure 3-10 on page 210 196.7 18. As the HIV/AIDS epidemic continues to spread, the maxi- 160 mum reduction in life expectan- 118.1 110 cy in the forecast period occurs in 2014, when AIDS-related 60 mortality peaks for this period. Dnipropetrovsk Donetsk Mykolaiv Odessa Ukraine Odesa and Mykolayiv Oblasts will suffer the most, with AIDS Source: Authors' calculations. shaving an estimated 3.4-4.1 years off male life expectancy and 4.3-5.2 year off females' under the optimistic sce- population of Dnipropetrovsk Oblast will decline by nario, and a corresponding 6.5-7.1 (males) and 7.7-7.8 an extra 40,000 due to the epidemic. Similarly, demo- years (females) under the pessimistic one (Annex 2 graphic losses in "with AIDS" compared to Table A2-9). This aggravates the already-unfavorable "no-AIDS" scenarios are 40,000-60,000 in Donetsk, regional demographic situation, with crude death 20,000-30,000 in Odesa, and 10,000-20,000 in rates higher and life expectancy lower than the Mykolayiv Oblasts. Clearly, AIDS accelerates the national average. already-rapid population decline in these oblasts (Annex 2 Table A2-10). The epidemic's effect on birth rates derives from the fact that HIV-infected females demonstrate relatively The analysis above suggests that the impact of lower fertility, so the absolute number of births dur- HIV/AIDS on regional demographic processes will ing the epidemic drops. Based on our calculations, be long lasting and continuing beyond this study's the cumulative number of unborn babies due to forecast horizon. The HIV/AIDS epidemic will aggra- the infection of potential mothers is 380-610 in vate the current negative trends in the regional Mykolayiv, 410-650 in Odesa, 2,000-3,500 in Donetsk, population dynamics. The epidemic's demographic and 2,700 in Dnipropetrovsk Oblasts. impact is likely to impose a heavy burden on all four oblasts, hindering their economic and social Both regional epidemic scenarios are superimposed development. on the baseline population decline. In 2014, the 17 Figure 3-10. Forecasted Life Expectancy* in "No-AIDS," "AIDS Optimistic," and "AIDS Pessimistic" Scenarios, Odesa, Mykolayiv, Donetsk, and Dnipropetrovsk Oblasts, 1994-2014, in Life Years Odesa Mykolayiv 70 70 68 68 66 66 64 64 Years Years 62 62 Without AIDS Optimistic Without AIDS Optimistic 60 Pessimistic 60 Pessimistic 58 58 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Dnipropetrovsk Donetsk 71 70 70 69 68 68 66 67 66 64 Years Years 65 62 64 Without AIDS Optimistic 63 60 Without AIDS Pessimistic Pessimistic 62 58 61 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 * For both males and females. Source: Authors' calculations. 18 CHAPTER 4 Impact of the Epidemic on the Labor Force and Government Revenues T he labor force forecast is Table 4-1. Estimated Reduction in Selected Labor Market Indicators in the based on the social budgeting "No-AIDS" Scenario, 2004-14 (in Thousands) and Percentage methodology developed by of Reduction Ukraine's Ministry of Labor and Indicator "No-AIDS" scenario Social Policies/National Academy of Reduction from Sciences, jointly with the United 2004 2014 Percentage 2004 to 2014 Nations Development Program Working-age population 36,173.9 33,751.9 2,422.0 6.7 (UNDP), International Labor Organization (ILO), and the World Labor force 22,490.4 20,154.8 2,335.6 10.4 Bank. Other methodological Employed 20,440.1 18,313.5 2,126.6 10.4 approaches included mid- and long- Unemployed 2,050.3 1,841.3 209.0 10.2 term projections of demand for Source: Authors' calculations. social services and unemployment benefits payable by the Special Fund (social insurance against unemployment). Labor pro- the labor force. The five-year age group and gender jections were constructed using State Statistics projections of the working-age population are based Committee data for 1998-2003 on the working-age on actual labor force participation data for 1998-2003 population, labor force participation (economic activ- and Chapter 3's demographic forecast. It is estimated ity), population not in the labor force (those on aged that the working-age population shrinks by 2.4 million or disability pension, full-time students, discouraged (6.7 percent) over the forecast period (2004-14). The workers, etc), and employed and unemployed popu- projected labor force would decline by 2.3 million lations. Data were disaggregated by five-year age (10.4 percent) over the same period. State Statistics groups and gender. The labor force forecast relied on Committee data indicate that the labor force declined the demographic projections in Chapter 3 (including over 1998-2003 by 3.3 million (1.2 million males and projections of the population aged 15-70) and macro- 2.1 million females). The sex ratio of labor force also economic projections for GDP, labor productivity, changed, with the share of males in the total labor and average monthly wages up to 2014. A hypotheti- force increasing from 49.2 percent in 1998 to 51.1 per- cal benchmark "no-AIDS" forecast was constructed, cent in 2003 (and a corresponding decline in the along with three "with-AIDS" projections based on female share from 50.8 percent to 48.9 percent). This three epidemic scenarios (medium, optimistic, and tendency is preserved over the ten-year forecast hori- pessimistic). Details are provided in Annex 3. zon, with the males' share reaching 52.3 percent. According to the forecast, total employment will Analysis at the National Level decline 10.4 percent by 2014, from 20.4 million to 18.3 million. The number of unemployed decreases As the first step, an assumption was made for the "no- 10.2 percent over the period from 2.1 million to AIDS" scenario that any person aged between 15 and 1.8 million. Details are in Table 4-1. 70 is a working-age person who is either in or out of 19 Table 4-2. Estimated Reductions in Selected Labor Market Indicators due to 2014, it is estimated that in 2014 the Epidemic, Compared to the "No-AIDS" Scenario, 2014 (in the HIV/AIDS epidemic yields an Thousands and as Percentage of Reduction) additional decline of 0.8-1.4 per- cent in the working-age popula- Indicator AIDS epidemic scenario tion, a 1.0-1.7 percent decline in MEDIUM OPTIMISTIC PESSIMISTIC the labor force and employment, Percentage Percentage Percentage and 1.4-3.0 percent reduction in Working-age population 441.1 1.3 268.1 0.8 472.8 1.4 unemployment (see Table 4-2). Labor force 323.4 1.6 193.2 1.0 351.0 1.7 Employed 286.4 1.6 170.4 1.0 301.7 1.7 Analysis at the Unemployed 66.1 3.6 24.9 1.4 55.2 3.0 Regional Level Source: Authors' calculations. To assess the epidemic's impact on the regional level, four oblasts were selected: Donetsk, Dnipropterovsk, Odesa, and Mykolayiv. The Labor Force in AIDS Epidemic Scenarios calculations were based on the hypothetical "no- The AIDS epidemic affects the size of a labor force AIDS" scenario and the three epidemic scenarios. significantly. Using Chapter 3's three epidemic sce- Projections were made for the working-age, economi- narios and applying the methodology outlined in cally active, employed, and unemployed populations Annex 3, three projections were constructed for the by gender using the same methodology as in the working-age population, labor force, employment, national level analysis. Results confirm the expected and unemployment. Comparing the endpoint of pro- reduction in the size of the working-age population jections in 2014 to the baseline "no-AIDS" value in and labor force, including a reduction in the employed population in all four oblasts. Figure 4-1. Regional Comparison of Estimated Labor Force Reduction from As Table 4-3 demonstrates, these the Epidemic, Compared to the "No-AIDS" Scenario, 2014 oblasts suffer from a far stronger (Percentage of Reduction) HIV/AIDS impact on the labor force and employment than the national average: by a factor of 2-2.5 in the pessimistic 5.0 4.7 4.6 scenario. Details of the estimations are 4.5 4.1 provided in Annex 3. Figure 4-1 illus- 4.0 trates the additional burden of 3.5 HIV/AIDS on the labor force in 2014 in 3.1 3.0 these oblasts compared to the national 2.5 2.5 average. 2.2 2.0 1.8 1.7 1.5 Impact of Epidemic on 1.0 1.0 Government Budget Position 0.5 and Special Social 0.0 0.0 Protection Funds OPTIMISTIC PESSIMISTIC HIV/AIDS epidemic scenario The study estimated forgone revenue to the state and special funds (including Dnipropetrovsk Mykolaiv Ukraine total pension and social insurance funds cov- Donetsk Odesa ering temporary disability, unemploy- ment, and social protection of those Source: Authors' calculations. with permanent disabilities) caused by 20 Table 4-3. Regional Comparison of Estimated Losses from the Epidemic in Selected Labor Market Indicators, Compared to the "No-AIDS" Scenario, 2014 (Percentage of Reduction) Percentage decline compared to "no-AIDS" 2014 baseline in: Working-age population Labor force Employed Unemployed Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Dnipropetrovsk NA 2.5 NA 4.6 NA 4.6 NA 4.6 Donetsk 1.3 2.2 3.1 4.1 3.1 4.1 3.1 4.1 Mykolaiv 1.2 1.2 2.2 4.7 1.1 3.6 9.2 11.6 Odesa 1.2 1.8 2.5 1.8 4.8 4.8 6.6 6.6 Ukraine total 0.8 1.4 1.0 1.7 1.0 1.7 1.4 3.0 Note: "NA" = not available: no optimistic scenario was constructed for Dnipropetrovsk due to lack of data. Source: Authors' calculations. the epidemic. Such forgone revenue results from a disability payments from the same fund for those reduction in the number of people employed and an infected with HIV and progressing to AIDS, and state increase in the number who cannot work due to ill- assistance to children with HIV/AIDS. ness. See Annex 4 for methodology, definitions, assumptions, and results from this section. This analysis assumes that everyone who develops AIDS becomes permanently disabled and eligible for Revenue forgone through unpaid taxes and levies a disability pension and related additional benefits. due to the reduction in employment was estimated Based on the projected number of AIDS cases for using two epidemic scenarios: optimistic and pes- 2004-14 and a range of assumptions for calculating simistic. Using the projected reduction in employ- the corresponding disability benefits (Annex 4), the ment, forgone state revenues are calculated as the estimated additional annual expenditure by the pen- amount of unpaid personal income tax; forgone pen- sion fund for permanent disability pensions to those sion fund contributions (unpaid fees for mandatory who develop AIDS will reach 109.2-200.0 million state pension insurance); forgone revenue to the dis- UHA (optimistic-pessimistic) by 2014. The corre- ability social insurance fund (unpaid premiums to sponding average growth rate of AIDS-related the fund); and forgone revenue to the unemployment outlays from the pension fund is 13-15 percent per social insurance fund (unpaid levies to this fund), year (optimistic-pessimistic scenarios). The total respectively. additional annual expenditure from the SPF related to permanent disability from AIDS is estimated to Using the estimates of reduction in employment, the reach 19.5-35.5 million UHA by 2014 (optimistic- average withholding rate, and the average monthly pessimistic), a 3-7 percent increase in total outlays wages, the study calculates the annual forgone rev- from these funds. Adding the guaranteed minimum enue to the state and special funds to be between pension to children with HIV/AIDS yields an addi- 263.8 million and 418.8 million UHA (optimistic-pes- tional 3.5-8.3 million UHA payment from the SPF. simistic scenario), or 0.13-0.21 percent of the total. Before developing full-blown AIDS, those infected On the expenditure side, direct budgetary costs take with HIV become progressively ill, requiring sick the form of permanent disability pensions from the leave from work, which is funded by a social protec- pension fund for those who progress to AIDS, addi- tion (temporary disability) fund. Overall, 524,000 tional financial assistance from the social protection persons in the optimistic and 721,100 persons in fund (SPF) for those disabled by AIDS, temporary the pessimistic scenarios will receive temporary 21 disability benefit over 2004-14. Table 4-4. Total Additional Annual Non-medical Budgetary This translates into an addi- Losses/Costs Associated with HIV/AIDS, 2014 tional annual outlay of UAH In 100,000 UAH 6.8-11.5 million by 2014 (Annex 4). Category Optimistic Pessimistic Forgone revenue 263.8 418.6 Table 4-4 lists these calcula- Expenditures tions, providing estimates of Pension fund: Permanent disability due to AIDS 109.2 200.0 total non-medical annual state Additional assistance to permanently disabled (SPF) 19.5 35.5 costs associated with social protection and pensions for Temporary HIV disability payments (SPF) 6.8 11.5 HIV/AIDS victims, plus the Assistance to children with HIV/AIDS 3.5 8.3 forgone revenue, at UHA Total additional expenditures 139.0 255.3 402.8-673.0 million, depending Total budgetary costs 402.8 673.9 on the epidemic scenario. Health expenditure associated Source: Authors' calculations. with HIV/AIDS is estimated in Chapter 5. 22 CHAPTER 5 Estimating the Macroeconomic Costs of the HIV/AIDS Epidemic T his study applied various macroeconomic Several studies of the economic costs of HIV/AIDS models to estimate the costs of HIV/AIDS in have been conducted in the Russian Federation. Ukraine, building on similar work done in Similar to Ukraine, Russia has experienced one of the Ukraine and Russia. The purpose is to use the most world's fastest growing epidemics over the past five recent available data and methodology to provide a years (according to the UNAIDS 2003 estimate, plausible range for the magnitude of the impact of 860,000 infected or 1.1 percent adult (15-49) preva- the epidemic. lence). Its apparent shift from an IDU-driven to a gen- eralized epidemic is similar to that of Ukraine. The literature on the macroeconomic costs of A World Bank team has developed a simple growth HIV/AIDS is large and continuously expanding.7 model (Ruehl, Pokrovsky, and Vinogradov [2002]). Among the analytical tools used for modelling are a UNDP developed a 35-sector CGE model based on neoclassical growth model8 (based on the aggregated Russian input-output tables (UNDP [2004] and Sharp variables, such a model necessarily misses microeco- [2004b]). The International Labor Organization (ILO) nomic effects on heterogeneous households), vari- has developed a model that combines an infection ous types of computable general equilibrium (CGE) probability profile with a partial equilibrium economic macroeconomic models,9 and macroeconometric model to assess the impact of HIV/AIDS on popula- models.10 As a rule, data requirements rise with the tion, labor force, sustainability of the pension fund, complexity of the model.11 Incorporating a mecha- costs of short-term disability benefit, health care nism for disease transmission and modelling its effects adds another layer of complexity. 8 A neoclassical growth model can be either aggregated (one-sector, two-factor) or disaggregated by type of labor (skilled/unskilled) and/or by sector. Open- and closed-economy assumptions yield dif- 7 International literature on modeling the economic impacts of ferent results. Haacker (2004a) demonstrates that a perfect capital HIV/AIDS has been thoroughly summarized and reviewed in mobility assumption yields more negative per capita effect and Haacker (2004b). Cross-sectional estimations are reported in Over highlights shortcomings of an aggregate approach in modelling the (2002), Cuddington (1993a and b), Cuddington, Hancock, and economic impact of AIDS. Rogers (1994), and Cuddington and Hancock (1994), among oth- 9 ers. A one-sector growth model with two types of labor, an exoge- CGE models contain behavioral equations for consumers/firms nous saving rate, and a closed-economy assumption is discussed in derived from the microeconomic optimization theory and can be Haacker (2004a). Open economy with perfect capital mobility is either comparative static or dynamic (the latter are based on the modelled in Freire (2004). Haacker (2002a) studies the effects of intertemporal optimization or are recursive dynamic). There are HIV/AIDS on the public sector and on economic growth in both single and multi-country models developed for various analytical closed- and open-economy settings. A model covering the informal purposes, all with varying degreea of sectoral disaggregation. They sector is proposed in Haacker (2002b). Intertemporally optimizing are formulated in either a representative consumer or an OLG consumers investing in human capital with the presence of AIDS is framework. Financial assets may or may not be included in the considered in Bell, Devarajan, and Gersbach (2004). Arndt and model. Lewis (2001) examine implications of the HIV/AIDS epidemic in 10 South Africa for sectoral employment and economic growth using Based on aggregate economic theory and a time series analysis a computable general equilibrium (CGE) model. An overlapping technique, macroeconometric models often lack theoretical struc- generations (OLG) framework with human capital is used. Studies ture. Econometric estimation often proves to be technically chal- of the impact of intervention strategies on the dynamics of the epi- lenging in transition and developing-economy settings. demic include Lewis (1989), IUSSP (1993), Kaplan and Brandeau 11 (1994), and FitzSimons, Hardy, and Tolley (1995). The experience Econometric estimates of core model parameters (such as elastici- of developing countries battling the epidemic, including prevention ties of substitution in consumption and production input bundles) policies, socioeconomic determinants of the epidemic, and AIDS' need to be obtained for CGE models. Multi-sectoral models require direct impact on the health sector and households, is discussed in input-output tables often unavailable in developing and transition Ainsworth, Fransen, and Over (1998). countries. 23 expenditures for employees, and changes of employ- of ART, the level of hospitalization of AIDS patients, ment on GDP (ILO [2004], reported in Sharp [2004a]). and the cost/coverage of hospitalization are consid- ered: scenario A is "ARV-low, HOSP(italization)-low"; Barnett et al. (2001), in work funded by the U.K. scenario B is "ARV-high, HOSP-low"; and scenario C Department for International Development and The is "ARV-high, HOSP-high." Detailed descriptions of British Council, evaluated the epidemic's social and the scenario assumptions, methodological approach, economic impact in Ukraine. It is the most widely data inputs, and model results are in Annex 5. In this cited study to date used successfully for policy advo- section, the optimistic scenario builds on the possible cacy. Since the epidemic situation has been changing policy intervention that extends availability of ART to rapidly since 2001, and new data have become avail- a greater number, compared to the pessimistic sce- able, the AIDS research and policy community need a nario. It also assumes that the measures of the re-evaluation of that study. Modeling also needed to Ukrainian National Program to Fight HIV/AIDS are be extended to capture the epidemic's macroeconom- successfully implemented, with a corresponding ic effects using the recently available methodology. reduction in the rates of MTCT, etc. By calculating the number of avoided new HIV cases and using the Given the degree of uncertainty about the magnitude Ukrainian cohort life expectancies, the Disability of the epidemic and its impacts on the factors of pro- Adjusted Life Years (DALYs) prevented through the duction and economic parameters, this study intervention are calculated and discussed. attempted to apply several models for the analyses. Comparing model implications establishes a plausible Model Results range for the magnitude of the effects. Lack of data Cost scenario A, ARV-low, HOSP-low, would result on the costing and effectiveness of many preventive in the following compared to the baseline in 2014: programs (harm reduction, sex education, and con- dom distribution) limited this study to a projection of Reduction in the level of output in constant the impact of prevention and treatment to ART while prices by 0.7 percent (optimistic scenario) and acknowledging that other measures are also impor- 1.3 percent (pessimistic), tant and can have significant socioeconomic impact. Per capita output unchanged, To ensure comparability of the results, all three mod- els use the same inputs generated by demographic Reduction in average GDP growth rate over and epidemiological forecasting module and the same 2004-14 of 0.06 percent (optimistic) and 0.11 scenarios with respect to the costs of treatment and percent (pessimistic), ART. A simple growth model, a macroeconometric Reduction in capital stock by 0.2 percent (opti- model, and a CGE model were all applied to evaluate mistic) and 0.3 percent (pessimistic), the macroeconomic costs of HIV/AIDS in Ukraine. Reduction in labor supply by 1 percent (opti- Both macroeconometric and CGE models are multi- mistic) and 1.5 percent (pessimistic), and sectoral, allowing us to study the differential effects of the epidemic on various sectors of the economy. Reduction in investment by 0.7 percent (opti- While different in theoretical structure, both models mistic) and 1.3 percent (pessimistic). demonstrate strong sectoral effects. 12 Epidemic scenarios differ in their assumptions about the size and Simple Growth Model dynamics of the most-at-risk populations, yielding different esti- mates of adult prevalence rates. In our optimistic scenario, adult HIV prevalence rate peaks at 2% in 2010 and in the pessimistic one This section analyzes the application of a simple at 3.5% in 2014. Reduction in the vertical transmission rate (15.9% growth model based on a hypothetical baseline in 2003) is faster in the optimistic scenario (to 10% in 2004 and then to 5% in 2014) than in the pessimistic one (gradual reduction "no-AIDS" scenario and two of the epidemic to 10% in 2014). Availability of ART to those who need it increases scenarios (optimistic and pessimistic) constructed in from 1% in 2004, to 30% in 2010, and further to 50% in 2014 in the Chapter 3.12 Three scenarios with respect to the cost optimistic scenario and to 5% in 2005 and holding at that level until 2014 in the pessimistic one. 24 The order of effects generated by the Table 5-1. Estimated Annual Medical Expenditure Associated with HIV/AIDS model is modest. Scenarios B and C Prevention and Treatment in 2014 with respect to the cost of ARV thera- 100,000 UAH py and hospital treatment have only Medical expenditure, including: Optimistic Pessimistic marginal effect on the macroeconomic variables (see Annex 5 Table A5-2). ART 353.0 51.9 AIDS care 275.8 504.5 Total AIDS-related medical expendi- Total medical expenditure 628.8 556.4 ture is higher in the optimistic sce- Medical expenditure as a percentage of MOH budget 4.19 percent 3.71 percent nario compared to the pessimistic one, reaching an annual amount of Source: Authors' calculations. UAH 628 million by 2014. More than 56 percent of the 2014 total medical expenditure in the optimistic scenario mistic scenario: in the former, the total number of is devoted to ART. This contrasts to a less than 10 infections, the adult prevalence rate, and the number percent share of ART in the pessimistic case, where of those needing ART exceed those in the optimistic most of the budget is allocated to hospital care. In scenario by a factor of 1.7-1.8, the number of new both cases, HIV/AIDS-related costs represent about 4 AIDS cases and annual AIDS deaths by a factor of percent of the MOH budget. At the same time, mortal- 1.8-1.9, and the number of new annual HIV infections ity and morbidity outcomes in the pessimistic sce- and annual births to HIV-positive mothers by a factor nario are significantly worse compared to the opti- of 3.2-3.7 (see Tables 5-1 and 5-2). Table 5-2. Projected Epidemic Outcomes, 2014, Scenario Analysis Ratio HIV/AIDS Summary: 2014 Optimistic Pessimistic (pessimistic to optimistic) Number infected with HIV, thousands 478.5 820.4 1.71 Adult prevalence rate, percentage 2.0 3.5 1.76 New annual HIV infections, thousands 29.0 94.0 3.24 Cumulative number needing ARV treatment, thousands 94.0 155.0 1.65 New annual AIDS cases, thousands 36.8 67.3 1.83 Annual HIV+ births, thousands 0.5 1.7 3.69 Annual AIDS deaths, thousands 34.8 64.9 1.86 Annual AIDS deaths per thousand 0.8 1.5 1.89 Cumulative AIDS deaths, thousands 301.3 526.4 1.75 AIDS orphans, thousands Dual 26.0 42.0 1.62 All 105.0 169.0 1.61 Life expectancy, years Total 68.5 66.7 0.97 Male 63.4 61.6 0.97 Female 72.9 71.0 0.97 Population, hundred thousands 43.9 43.7 0.99 Source: Authors' calculations. 25 Figure 5-1. Number of HIV Infections Averted by Realizing "AIDS Optimistic" et al. [2000]; and Murray and Rather Than "AIDS Pessimistic" Scenario, 2004-14 Lopez [2000]) and uses the Disability Adjusted Life Year 80 (DALY) measure. See Annex 6 70 HIV_avert for detailed discussion of the 60 GBD methodology. 50 The underlying pessimistic and 40 optimistic epidemiological sce- Thousand 30 narios rest on an assumption 20 about the availability of ART. The study assumes 100 percent 10 public financing of the therapy 0 and examines the cost-effective- 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 ness of such a policy interven- Source: Authors' calculations. tion measured in public expen- diture per DALY saved/avoided. Analysis of Optimistic versus Pessimistic Epidemic Scenarios as a To construct the DALY measure, the number of new Policy Intervention HIV cases avoided each year was predicted for 2004- 14 based on the Spectrum AIM projections. The The incidence-based methodology for the economic cases avoided were recorded by age-sex group for evaluation of the HIV intervention programs used in each year. As Figure 5-1 demonstrates, extension of this study follows the methodology of the Global ART to 50 percent of those in need (optimistic), Burden of Disease (GBD) study (see Murray, Lopez, compared to 10 percent (pessimistic) by 2014, would and WHO [1994]; Murray and Acharya [1997]; Murray prevent 50,000 new HIV infections per year on aver- age over 2004-14. Figure 5-2. Number of DALYs Averted by Realizing "AIDS Optimistic" Rather For each prevented case, the Than "AIDS Pessimistic" Scenario, 2004-14 Years of Life Lost (YLL) was 3000 calculated based on the cohort life expectancy, and the Years Lost to Disability (YLD) was 2500 calculated based on the assumptions about the dura- 2000 tion of the disease and its severity, presented in Annex 6 1500 Table A6-1. It is assumed that Thousand on average, a child develops 1000 full-blown AIDS in 6 years and dies in 7 years and 3.5 months. 500 An adult develops AIDS within 8 years and dies in 12.4 years. 0 Based on the disability weights 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 from Annex 6 Table A6-1, DALY (0,0) DALY (3,0) DALY (3,0)_d DALYs averted were calculat- ed for 2004-14, with three Source: Authors' calculations. Note: See footnote on page 27 for legend description. 26 measures constructed.13 Results are presented in 3.0 percent by 2014). GDP (expenditure) measured Figure 5-2. though the basic macroeconomic identity falls by 1.4-2.7 percent in the range of epidemic scenarios. Depending on the measure used, in excess of 21 mil- Gross investment falls by 1.5-2.8 percent compared lion undiscounted DALYs (13.8 million DALYs dis- to the benchmark, and budget revenue declines by counted at 3 percent) would be saved over 2004-14 if 1.3-2.6 percent by 2014, depending on the epidemic the optimistic scenario is followed instead of the pes- scenario. The estimations show only slight decreases simistic one. In low-cost scenario A, this would be in savings, imports, and average labor productivity achieved at an average cost of 11-37 UHA per DALY (output per employee). (depending on whether an undiscounted or a dis- counted DALY measure is used), which translates Sectoral Analysis into a total average health expenditure of 9-29 UHA The model allowed us to estimate value-added in per DALY after taking into account the corresponding separate sectors of economic activity (in constant reduction in AIDS hospitalization costs. The average 1996 prices) for the following sectors: agriculture, annual cost per averted HIV infection is 419 UHA (or hunting, forestry and fishing industry; mining; manu- 328 UHA net of avoided hospitalization costs). The facturing; production and distribution of electricity, high-cost scenario C generates an average cost of 54- gas, and water (EGW); construction; wholesale and 184 UHA per DALY (depending on definition), which retail trade, trade in transport facilities and repair translates into the total average health expenditure of services; transport and communications; financial 17-57 UHA per DALY (after accounting for avoided services; real estate operations, leasing and services hospitalization costs). The average annual cost, net of to legal entities; government services; health and avoided hospitalization costs, per averted HIV infec- social services; education; and other services. The tion in the high-cost scenario is 762 UHA. Provision epidemic's estimated impacts on these sectors are of ART appears to be a highly cost-effective interven- presented in Table 5-3 on page 28. tion in this hypothetical comparative analysis of opti- mistic and pessimistic scenarios. Details of the As follows from Table 5-3, the agriculture, hunting, results are in Annex 6 Table A6-3. forestry, and fishing industry is the most affected (a decline of 1.2-2.3 percent compared to the "no-AIDS" Macroeconomic Model scenario in 2014), followed by transport and commu- nications (a reduction of 1.2-2.2 percent), and con- In this section, a macroeconometric model is used to struction (minus 1-1.8 percent). The gap between evaluate the economic costs of Ukraine's epidemic, sectoral outputs in "no-AIDS" and "with-AIDS" sce- using the labor force projections from our epidemic narios widens as the epidemic unfolds. scenarios. The impact of HIV/AIDS on sectoral employment during 2004-14 is estimated and used as The results for agriculture are explained by the fact an input into the macroeconometric model. The that the regions worst affected by HIV/AIDS produce methodology for model estimation and application is almost 80 percent of Ukraine's total agricultural out- discussed in Annex 7. put. Losses in the labor force and a high estimated labor intensity in the agricultural production func- Predicted decline in the level of GDP depends on the tion lead to an overall strong sectoral effect. Annex 7 measure of GDP used (product or expenditure). GDP Table A7-3 ranks the oblasts with regard to major (production) shortfall is larger due to the direct agricultural producers in terms of the HIV preva- impact of reduced employment (a reduction of 1.6- lence in 2004, with 1 corresponding to the lowest prevalence rate and 5 the highest. Most agricultural employment is located in the areas with a relatively 13 Three measures are DALY (0,0), undiscounted; DALY (3,0), dis- high HIV prevalence. Due to the unavailability of counted at 3 percent; and DALY (3,0)_d (discounted), with YLL rural/urban prevalence data, this was the best feasi- discounted back from the time of death to the time of infection. 27 Table 5-3. Macroeconometric Model: Estimated Difference in Sectoral on the general methodological approaches Output in Two Epidemic Scenarios, 2005-14 discussed above (e.g., Sharp [2002]; Ruehl, Pokrovsky, and Vinogradov [2002], and Percentage difference from Haacker [2004 a and b]), the impact of "no-AIDS" scenario 2005 2014 2005 2014 HIV/AIDS on the economy was modeled as OPTIMISTIC PESSMISTIC three distinct shocks, as follows: All sectors -0.35 -0.90 -0.31 -1.73 Agriculture, hunting, forestry, & fishing -0.29 -1.20 -0.26 -2.31 Reduction in labor supply. A 1.5 per- Mining -0.03 -0.10 -0.03 -0.19 cent, 2 percent, and 4 percent decline in labor force endowment is assumed for Manufacturing -0.09 -0.23 -0.08 -0.43 optimistic, medium, and pessimistic sce- EGW -0.07 -0.20 -0.06 -0.38 narios, respectively, based on the labor Construction -0.26 -0.95 -0.23 -1.83 force projections of Chapter 4. The most Wholesale & retail trade -0.05 0.16 -0.05 -0.31 pessimistic scenario in terms of labor sup- Transport & communications -0.27 -1.17 -0.23 -2.24 ply shock is based on the projected magni- Financial services -0.05 -0.23 -0.05 -0.44 tude of labor force decline in the most- Real estate -0.03 -0.08 -0.03 -0.16 affected regions. See Annex 8 for details on the distribution of shocks by type of labor. Government services -0.09 -0.41 -0.08 -0.79 Health & social services -0.03 -0.11 -0.03 -0.20 Reduction in labor productivity. The Education -0.19 -0.67 -0.16 -1.29 study assumed a 1.5 percent, 4 percent, and Other services -0.11 -0.22 -0.10 -0.43 7 percent reduction in labor productivity in optimistic, medium, and pessimistic sce- Source: Authors' calculations. narios, respectively. Changes in labor pro- ductivity were modeled through the Total Factor ble way to estimate the effect of HIV/AIDS on agri- Productivity (TFP) score, distributed according to cultural labor. The magnitude of the effect is the factor shares by type of labor (using Annex 8 explained by the particular form of the estimated Tables A8-1 and A8-2). production function. Increase in public spending. Based on the epi- State Statistics Committee of Ukraine (Derzhkomstat demic and cost scenario projections provided earlier, 2004a) data indicate that the worst-affected oblasts the public expenditure related to HIV/AIDS will in terms of the HIV prevalence are the ones with the increase within the range 0.2 percent, 2 percent, and highest output per worker and the highest agricultur- 3.5 percent of the government budget, for the al wages that drive inward migration. Table A7-4 optimistic, medium, and pessimistic scenarios, reports nominal wages by oblast, confirming that respectively. Dnipropetrovsk, Donetsk, Zaporizhya, Kyiv, Lugansk, Odesa, and Kharkiv Oblasts report wages significant- Based on these types of shocks, three scenarios were ly above the national average. All of these oblasts are applied. One scenario (Scenario 2 or "Medium") has at or above the national HIV prevalence rate. three subscenarios within it, referred to here as sce- narios 4-6 and listed in the tables under "Medium sub- scenarios." Multisector CGE Model A 20-sector computable general equilibrium (CGE) Scenario 1: Optimistic: 1) drop in labor supply model was also developed and applied to study the of 1.5 percent; 2) decrease in labor productivity epidemic's macroeconomic effects. Model descrip- of 1.5 percent; and 3) increase in public spending tion, methodology, and results are in Annex 8. Based by 0.2 percent; 28 Scenario 2: Medium: 1) drop in labor supply of Scenario 6: Increase in public spending of 2.0 2.0 percent; 2) decrease in labor productivity of percent. 4.0 percent; and 3) increase in public spending by Scenarios 4-6 were used to evaluate the relative 2.0 percent; importance of the underlying shocks in generating Scenario 3: Pessimistic: 1) drop in labor supply the overall effect. of 4.0 percent; 2) decrease in labor productivity of 7.0 percent; and 3) increase in public spending Estimated Macroeconomic Implications by 3.5 percent; Table 5-4 documents the negative impact of Scenario 4: Drop in labor supply of 2.0 percent; HIV/AIDS across all scenarios. Total welfare and Scenario 5: Decrease in labor productivity of GDP substantially decrease under all six scenarios, 4.0 percent; including those with single shocks, and the gap Table 5-4. CGE Model: Macroeconomic Implications of the Epidemic, Scenario Analysis Scenarios Medium subscenarios Higher MACRO Reduced Lower public INDICATORS Benchmark Pess-c Med-m Opt-c labor TFP spending Welfare (equivalent variation, change in percentage) - -8.3 -4.6 -2.2 -2.6 -3.3 0.0 GDP Index (change in percentage) - -5.5 -3.1 -1.6 -1.8 -2.3 0.2 Private investment (change in percentage) -9.0 -5.0 -2.4 -2.8 -3.6 0.0 Real factor return (change in percentage) ­ Return to capital - -7.03 -3.87 -1.90 -2.22 -2.55 -0.11 ­ Wage rate for unskilled labor - -7.46 -4.17 -1.78 -1.93 -3.55 -0.03 ­ Wage rate for skilled labor - -2.58 -1.70 0.07 0.55 -3.56 0.05 ­ Wage rate for highly skilled labor - -1.42 -1.05 0.37 0.89 -3.18 0.14 Aggregate exports (UAH billion) 113.24 102.54 107.12 110.40 110.05 108.52 113.18 Aggregate imports (UAH billion) 109.92 99.09 103.74 107.05 106.69 105.18 109.84 Total exports (change in percentage) - -9.46 -5.41 -2.51 -2.82 -4.17 -0.06 Total imports (change in percentage) - -9.86 -5.63 -2.61 -2.94 -4.31 -0.08 Tariff revenue (share of public budget) 10 % 9 % 9 % 10 % 10 % 10 % 10% Indirect tax revenue (share of public budget) 49 % 55 % 52 % 51 % 51 % 51 % 50% Indirect tax rate (weighted average) 12 % 15 % 13 % 12 % 12 % 12 % 12% Consumer Price Index (change in percentage) - -0.75 -0.42 -0.18 -0.20 -0.34 -0.01 Producer Price Index (change in percentage) - -3.01 -1.59 -0.66 -0.77 -0.90 -0.29 Real exchange rate (change in percentage) - -1.60 -0.77 -0.34 -0.43 -0.25 -0.19 Source: Authors' calculations. 29 widens from optimistic to pessimistic scenarios. Increased public spending yields a smaller effect. Increased expenditure on care and treatment raises Return to labor (wages) declines in all scenarios public and private consumption and decreases sav- except the optimistic one. Separating out the effects ings and investments. Private investment falls by of reduced labor supply and lower productivity, we 9 percent in the pessimistic scenario, following the find that as skilled and highly skilled labor becomes 7 percent reduction in real rate of return to capital. scarcer, their factor payment (wages) goes up. The latter is due to the reduced marginal product of Nevertheless, the labor productivity factors act in capital following the loss of labor. the opposite direction, pushing wages down, and the combined effect is a fall in wages. Note from The reduction in labor productivity (modeled as a Table 5-4 that both exports and imports fall, driven shock to TFP) is the strongest driver of the negative by the changes in domestic supply and demand. impact, followed by the reduction in labor supply. Table 5-5. CGE Model: Sectoral Implications of HIV/AIDS Epidemic, Scenario Analysis Scenarios Medium subscenarios Higher OUTPUT Reduced Lower public INDEX Benchmark Pess-c Med-m Opt-c labor TFP spending Agriculture, hunting 1.00 1.02 1.01 1.00 1.00 1.01 1.00 Fishery 1.00 0.99 1.00 1.00 0.99 1.00 1.00 Mining of coal and peat 1.00 0.91 0.95 0.98 0.97 0.97 1.00 Production of non-energy materials 1.00 0.67 0.81 0.91 0.91 0.84 1.00 Food-processing industries 1.00 0.98 0.99 1.00 0.99 1.00 1.00 Textile and leather industry 1.00 0.99 1.02 1.01 1.01 1.05 0.98 Woodworking, pulp and paper industry, publishing 1.00 1.03 1.03 1.01 1.00 1.04 1.00 Petroleum refinement 1.00 0.95 0.97 0.99 0.99 0.98 1.00 Manufacture of chemicals, rubber and plastic products 1.00 0.89 0.95 0.98 0.98 0.96 1.00 Manufacture of other non-metallic products 1.00 0.90 0.95 0.98 0.97 0.97 1.00 Metallurgy and metal processing 1.00 0.63 0.78 0.91 0.90 0.81 1.00 Manufacture of machinery and equipment 1.00 0.91 0.96 0.97 0.97 0.99 1.01 Other 1.00 0.74 0.84 0.93 0.93 0.86 1.00 Electric energy 1.00 0.90 0.95 0.98 0.97 0.96 1.00 Public utilities 1.00 0.94 0.97 0.98 0.98 0.98 1.00 Construction 1.00 0.93 0.96 0.98 0.98 0.98 1.00 Trade 1.00 0.94 0.97 0.98 0.98 0.98 1.00 Hotels and restaurants 1.00 1.21 1.12 1.04 1.04 1.12 1.00 Transport 1.00 1.20 1.11 1.04 1.04 1.10 1.00 Post and telecommunications 1.00 1.01 1.01 1.00 1.00 1.02 1.00 Other services 1.00 1.00 1.00 1.00 0.99 1.01 1.00 Note: All benchmark indexes equal unity (or 100 percent), reflecting the starting (benchmark) position for the change. Source: Authors' calculations. 30 Sectoral Implications peat and production of non-energy materials sectors The HIV/AIDS epidemic is estimated to have impor- (the share of skilled and highly skilled is about 44 per- tant sectoral impacts, depending on the factor intensi- cent in each). Separate simulations under the medium ty of the sector and the distribution of skill classes scenario reveal that decline in labor supply and labor within the sector labor force (see Table 5-5). Sectors productivity has a relatively equal impact on sector with labor-intensive production and a high share of output, given a slightly higher weight on labor produc- skilled and highly skilled labor appear to be the most tivity in some sectors. Similar results are observed for affected. Examples include the mining of coal and sectoral exports (see Annex 8 Table A8-5). 31 CHAPTER 6 Policy Implications and Conclusions O ur findings after assessing the short- to medi- additional 1-2 percent reduction in labor force um-term (2004-14) socioeconomic impact of nationwide. The epidemic could contribute to labor the HIV/AIDS epidemic in Ukraine demon- force shrinkage in the worst-affected oblasts at rates strate that if not curtailed, the spread of this disease of 2.7-3.6 percent for Donetsk and 2.2-4.2 percent for is likely to have grave effects on the population and Odesa. The worst-affected oblasts in terms of HIV economy. Allowed to grow at its current rate, the prevalence have the highest output per worker and epidemic will have a long-lasting and destructive the highest agricultural wages that drive inward effect not only at the individual level but to the soci- migration. Younger people are most affected, with a ety at large. AIDS has become a reality of life in pronounced gender differential (the sharpest decline many countries, impeding human development, limit- is for females in the 15-19 age group). Longer-term ing the rights of children and adults to healthy and negative demographic consequences follow from the productive lives, and affecting living standards. As reduced fertility among young, HIV-infected women. this study shows, Ukraine's potential epidemic would likely undermine the economy, reducing the As labor takes its hit, so do families and children. labor force and revenues and increasing government The medium scenario posits that 42,000 orphans will costs. have lost both parents to AIDS by 2014, with another 105,000-169,000 having lost one parent, depending on The cumulative number of people infected with HIV the scenario. Those children are at risk of impeded is estimated to reach 479,000-820,000 by 2014, with access to quality education, health care, and even another 29,000-94,000 contracting it each year. The basic needs, unless they receive adequate assistance adult prevalence rate may reach 1.9-3.5 percent that from the government. Implications include not only same year, and those needing ART may increase to increased HIV/AIDS but even greater threats to 130,000 (77,000 in the optimistic scenario). Annually, society. AIDS would cause an estimated 35,000-65,000 deaths, with similar numbers developing the disease each While its revenue shrinks with the workforce, the year. AIDS would account for almost a third of all government would also experience increased med- male deaths and a staggering 60 percent of female ical expenditure and social security outlays. The deaths in the 15-49 age group by 2014, reducing life study finds that depending on the cost scenario for expectancy by 2-4 (3-5) years for males (females). ART and hospitalizations, annual AIDS care expendi- The spread of HIV/AIDS would exacerbate Ukraine's ture may be reach 630 million UAH by 2014 (esti- already-adverse demographic situation: without mate range: 41-629 million UAH, quite wide due to AIDS, low fertility rates would drive the Ukrainian the high degree of uncertainty about exogenous fac- population down to 44.2 million by 2014; with it, an tors such as future treatment costs). additional 300,000-500,000 would be lost, leaving a total population of 43.9-43.7 million. Revenue losses through the fall in employment due to HIV/AIDS, forgone income taxes, and unpaid pen- Echoing the underlying demographic decline of 10.4 sion and social security (temporary disability and percent from 2004 to 2014, HIV/AIDS will cause an unemployment) levies are estimated to reach 263-418 32 million UAH (in optimistic-pessimistic scenarios). sectors such as production of non-energy materials At the same time, projected additional budget expen- as well as metallurgy and metal processing would be diture in 2014 will require an extra 109-200 million most affected, with output falling by up to a third in UAH for permanent disability pensions due to the worst-case scenario. Given the relative share of HIV/AIDS, 20-35 million UAH in additional pensions these sectors in the country's trade structure, the from the social protection fund, 7-12 million UAH in worst-case scenario anticipates a fall of 40 percent in temporary HIV disability payments, and 3-8 million exports of these sectors, which translates into 5.5 UAH in AIDS orphan pensions. Thus, the total esti- percent fall in GDP, an 8 percent fall in total welfare, mated HIV/AIDS-related additional benefits are and a 9 percent fall in investment. The macroecono- 139-255 million UAH per year by 2014. metric model produces stronger effects for the agri- culture sector, due to the higher estimated labor The study applied several macroeconomic models share in the production function. to estimate the magnitude of the macroeconomic effects likely to be caused by the HIV/AIDS epidemic In line with other international studies, the modeling in Ukraine based on the range of plausible scenarios. results for Ukraine demonstrate that the HIV/AIDS Implications from the models include the following epidemic could lead to potentially catastrophic con- mid-term effects (by 2014 with AIDS, compared to sequences without an effective and timely national the "no-AIDS" baseline scenario): response. Even within the short to medium term, the study shows that the cost of inaction would be high 1-6 percent reduction in the level of output and the long-term implication could be even higher. (GDP in constant prices), 2-8 percent reduction in total welfare, and The epidemic's distribution as reported here calls for 1-9 percent reduction in investment. attention to and effective targeting of the young, females, and those in the worst-affected oblasts. The CGE analysis also demonstrates a decline in Prevention and treatment programs need to reach wages for unskilled, skilled, and highly skilled work- these groups and areas, and the messages and servic- er groups, driven by the HIV/AIDS-associated decline es must fit their needs. In addition, the pattern of in labor productivity. On a sectoral level, the labor- transmission requires a prevention strategy focused intensive industries with the greater share of skilled on harm-reduction programs as well as sex educa- and highly skilled workers proved to be especially tion for young populations. Even though the mode of vulnerable in terms of the production and export transmission is evolving toward heterosexuals, IDUs indicators.14 still constitute the majority of new infections. Special effort will have to be made to reach this mar- Sectoral analysis suggests that labor-intensive sec- ginalized group. tors whose labor inputs suffer from the epidemic will be among the worst affected. Based on the CGE, Given the important role they play in the Ukrainian economy, the worst-affected oblasts of Donetsk, Dnipropetrovsk, Odesa, and Mykolaiv should be treated with priority in implementing the HIV pre- 14 The most affected sectors in the CGE model are production of vention, education, and treatment measures. non-energy materials; mining of coal and peat; manufacture of chemicals and rubber; metallurgy and metal processing; and elec- tric energy. The result of the macroeconometric model is the oppo- Due to the data limitation, this study could model site, with agriculture posting the largest losses (1.2-2.3 percent of only the impact of ART, one of many possible inter- the baseline), followed by transport and communications (1.1-2.2 percent). Sectoral output in mining is estimated to decline by 0.1- ventions. Nevertheless, the study demonstrates that 0.2 percent. The model applied the largest losses to the labor force prevention and treatment could be cost-effective, in the agricultural sector, which coincidentally is concentrated in the eastern oblasts with the highest HIV prevalence rates. See Annex 7 for details. 33 and scaling up the treatment could avoid the other- Last, timing is crucial: as study results demonstrate, wise expected overburdening of the health system the epidemic is still spreading, so timely, effective and escalating costs. Although not modeled here, interventions, including making ART available, could preventive education measures must complement reverse the epidemic and reduce its negative impact ART to enhance its impact. on socioeconomic development in Ukraine. 34 ANNEX 1 HIV/AIDS in Ukraine: Official Data Table A1-1. Reported New Cases of HIV Infection, AIDS Cases, and AIDS Deaths, Ukraine, 1987-2004 Year 1987-95 1996 1997 1998 1999 2000 2001 2002 2003 2004 Total New HIV cases Total, including 1,897 5,422 8,934 8,590 5,830 6,216 7,009 8,761 10,013 12,494 75,166 Ukrainian nationals 1,673 5,400 8,913 8,575 5,827 6,212 7,000 8,756 10,009 12,491 74,856 Foreign nationals 224 22 21 15 3 4 9 5 4 3 310 Children 28 99 210 402 549 737 937 1,379 1,844 2,293 8,478 New AIDS cases Total, including 82 146 193 399 586 648 868 1,356 1,916 2,745 8,939 Ukrainian nationals 77 143 189 398 586 647 867 1,353 1,915 2,743 8,918 Foreign nationals 5 3 4 1 0 1 1 3 1 2 21 Children 7 10 4 14 15 13 30 47 68 96 304 AIDS deaths Total, including 38 70 85 150 253 415 474 837 1,285 1,775 5,382 Ukrainian nationals 34 69 82 148 253 414 473 834 1,285 1,775 5,367 Foreign nationals 4 1 3 2 0 1 1 3 0 0 15 Children 5 6 4 9 12 9 11 23 38 33 150 Source: Ukrainian AIDS Center. Table A1-2. National HIV Serosurveillance Data by Category/Code, Ukraine, 2002-04 2002 2003 2004 CODE CATEGORY Tested HIV+ % Tested HIV+ % Tested HIV+ % 100 Total Ukrainian 2,299,981 15,572 0.68 2,459,784 18,522 0.75 2,501,132 23,087 0.92 nationals, including 101 Sexual partner of 4,143 515 12.43 4,496 627 13.95 5,573 862 15.47 an HIV+ 101`1 Children born to 12,940 1,163 8.99 14,727 1,216 8.26 16,182 1,729 10.68 HIV+ mothers 102 IDUs 36,286 4,765 13.13 33,004 4,855 14.71 32,184 4,754 14.77 (continued on page 36) 35 Table A1-2. National HIV Serosurveillance Data by Category/Code, Ukraine, 2002-04 (continued from page 35) 2002 2003 2004 CODE CATEGORY Tested HIV+ % Tested HIV+ % Tested HIV+ % 104 Those with STIs 67,921 584 0.86 61,674 685 1.11 59,960 720 1.20 105 Multiple sexual 16,050 160 1.00 17,196 187 1.09 16,715 243 1.45 partners 108 Donors of blood & products, organs, 939,108 927 0.10 958,205 1,182 0.12 941,524 1,209 0.13 tissues 109 Pregnant women 808,632 1,874 0.23 924,099 2,555 0.28 965,405 3,252 0.34 112 Prisoners 12,770 1,192 9.33 17,782 1,603 9.01 25,638 3,273 12.77 113 Examined by 116,878 2,256 1.93 121,350 3,030 2.50 134,528 3,895 2.90 clinical indications 114 Examined on voluntary basis (anonymously or 36,679 1,145 3.12 34,326 1,303 3.80 38,326 1,588 4.14 confidentially) 115 Occupational exposure 4,677 26 0.56 4,938 7 0.14 5,648 2 0.04 (medical contacts) 120 Other 243,897 965 0.40 267,987 1,272 0.47 259,449 1,560 0.60 200 Foreign nationals 3,542 30 0.85 5,289 17 0.32 5,899 25 0.42 300 Total tested 2,303,523 15,602 0.68 2,465,073 18,539 0.75 2,507,031 23,112 0.92 Source: Ukrainian AIDS Center. Table A1-3. Reported HIV Cases by Mode of Transmission, 1987-2004 Year Mode of transmission 1987-95 1996 1997 1998 1999 2000 2001 2002 2003 2004 Total IDU 1,024 4,360 7,448 6,516 3,771 3,881 3,964 4,587 4,815 5,778 46,144 Heterosexual 433 709 1,007 1,385 1,323 1,427 1,885 2,499 3,043 4,041 17,752 MTCT 19 92 196 378 527 727 914 1,371 1,830 2,273 8,327 Unknown 161 236 260 294 202 173 231 295 314 389 2,555 Blood transfusion 4 0 0 1 0 0 3 2 3 1 14 Medical contact 5 0 0 0 3 0 0 0 1 0 9 MSM 27 3 2 1 1 4 3 2 3 9 55 Total 1,673 5,400 8,913 8,575 5,827 6,212 7,000 8,756 10,009 12,491 74,856 Source: Ukrainian AIDS Center. 36 Table A1-4. Reported Cases of HIV Infection in Ukraine, by Gender, 1995-2004 Total infected, Year including: Males Females Persons % Persons % 1995 1,490 936 62.8 554 37.2 1996 5,400 4,130 76.5 1,270 23.5 1997 8,913 6,569 73.7 2,344 26.3 1998 8,575 5,763 67.2 2,812 32.8 1999 5,827 3,757 64.5 2,070 35.5 2000 6,212 3,947 63.5 2,265 36.5 2001 7,000 4,326 61.8 2,674 38.2 2002 8,756 5,278 60.2 3,478 39.8 2003 10,009 5,695 56.9 4,314 43.1 2004 12,491 7,245 58 5,247 42 Total 74,673 47,646 63.8 27,028 36.2 Source: Ukrainian AIDS Center. Table A1-5. Reported HIV and AIDS Incidence in Ukraine, by Region, 2001-04 Officially registered HIV incidence Officially registered AIDS incidence Per 100,000 Per 100,000 Regions 2001 2002 2003 2004 2001 2002 2003 2004 Crimea 24.1 31.7 33.4 37.2 3.8 9.2 13.4 16.8 Vinnitsa 6.2 4.7 9.8 11.1 0.6 1 2.3 2.8 Volyn 5.9 12.1 9.2 12.7 0.2 1.4 1.8 1.6 Dnipropetrovsk 26.2 41.8 49.7 59.8 0.7 4 6.4 8.8 Donetsk 31.2 32.1 38.7 52.9 4.3 6.3 8.9 15 Zhytomyr 8.7 8 9.4 11.8 1.3 1.1 0.8 1.9 Zakarpatska 1.7 1 1 1.3 0.2 0.1 0 1.9 Zaporizhya 9.8 8.1 13 16.9 1.8 2 2.5 4.4 Ivano-Frankivsk 1.4 1.6 3.3 4.8 0.3 0.1 0.6 0.6 Kyiv 12.6 17.4 18.8 21.6 1 1.8 2.1 4 Kirovograd 8 11 11.7 17.5 0.4 0.8 0.6 0.4 Lugansk 8 15.1 13.9 23 0.5 0.6 1.3 2.2 Lviv 4.4 5.4 6.6 8.5 0.4 1 1 1.5 Mykolayiv 32 42.7 48.1 57.5 2.6 4.4 3.1 11.3 Odesa 36.3 46.5 48.3 58.9 11.6 14.5 16.7 15.8 Poltava 8.3 6.8 9.1 8.8 0 0 0.9 2.4 (continued on page 38) 37 Table A1-5. Reported HIV and AIDS Incidence in Ukraine, by Region, 2001-2004 (continued from page 37) Officially registered HIV incidence Officially registered AIDS incidence Per 100,000 Per 100,000 Regions 2001 2002 2003 2004 2001 2002 2003 2004 Rivne 2.2 7.4 8.2 9.1 0 0.1 1.5 0.6 Sumy 5.8 5.6 6.8 5.7 1.9 1.2 1.2 1.2 Ternopil 3.5 2.6 4.4 4.7 0.2 0.4 0.4 0.7 Kharkiv 5.4 7.5 13.6 14.2 0.2 0.4 1 0.7 Kherson 10.8 18.6 18.9 20 0.4 2.5 5.9 8.4 Khmelnytsky 8.9 21.2 16 14.2 0 0 3.3 3.1 Cherkasy 12.8 16 17 18.3 0.5 0.9 2.3 3.4 Chernivtsi 2.4 4 4.2 5.4 0.1 1 0.3 0.8 Chernigiv 10 9.6 12.9 20.1 0.2 0.5 0.2 0.4 Kyiv city 13.1 15.9 15.9 24 1.6 1.7 2.3 5.2 Sevastopol city 32.7 26 34.2 50.4 4.9 1.1 3.5 8.2 Total 14.2 18.2 20.8 25.9 1.8 2.8 4 5.7 Source: Ukrainian AIDS Center. Table A1-6. Officially Reported HIV Incidence among Blood Donors and Pregnant Women, by Region, 2004 % among donors % among pregnant % among donors % among pregnant AR Crimea 0.13 0.27 Odesa 0.25 0.60 Vinnitsa 0.06 0.09 Poltava 0.09 0.24 Volyn 0.05 0.11 Rivne 0.05 0.06 Dnipropetrovsk 0.24 0.7 Sumy 0.01 0.07 Donetsk 0.22 0.61 Ternopil 0.02 0.04 Zhytomyr 0.10 0.29 Kharkiv 0.09 0.13 Zakarpatska 0.01 0.10 Kherson 0.15 0.25 Zaporizhya 0.06 0.14 Khmelnytsky 0.08 0.42 Ivano-Frankivsk 0.06 0.05 Cherkasy 0.11 0.23 Kyiv 0.20 0.53 Chernivtsi 0.05 0.04 Kirovograd 0.17 0.48 Chernigiv 0.22 0.55 Lugansk 0.08 0.14 Kyiv city 0.14 0.54 Lviv 0.08 0.08 Sevastopol city 0.04 0.30 Mykolayiv 0.33 0.79 Total 0.13 0.34 Source: Ukrainian AIDS Center. 38 Table A1-7. HIV Seroprevalence Surveys among Table A1-8. HIV Seroprevalence Surveys among Injecting Drug Users (IDUs) Commercial Sex Workers (CSWs) City Year Sample size % HIV+ City Year Sample size % HIV+ 1999 259 37.8 2000 53 13.2 2000 259 41.7 2002 102 31.4 2002 250 31.6 Donetsk 2004 103 30.1 Poltava 2004 250 28. 2002 51 3.9 2000 252 39.7 Lutsk 2004 51 9.8 2002 250 40.0 2002 103 22.3 Donetsk 2004 250 41.6 Odesa 2004 124 31.1 Kryvyi Rig 2000 249 28.1 2002 100 17.0 2000 293 64.0 Poltava 2004 100 18.0 2002 259 58.3 2002 100 6.0 Odesa 2004 250 58.1 Simferopol 2004 100 19.0 2000 261 27.2 Mykolaiv 2002 100 30.0 2002 250 28. Kharkiv 2002 90 12.2 Simferopol 2004 363 59.0 Sumy 2004 20 10.0 2000 250 17.8 Kherson 2004 100 11.0 2002 250 16. Source: Artyukh et al. (2005b). Kharkiv 2004 241 14. Mykolaiv 2002 250 53 Table A1-9. HIV Seroprevalence Surveys among Patients with Sexually Transmitted 2002 250 32. Infections (STIs) Lutsk 2004 241 32.8 Sumy 2004 164 11. City Year Sample size % HIV+ Kherson 2004 250 31.2 1999 252 2.4 Kyiv 2000 500 1.8 Source: Artyukh et al. (2005a). 2000 476 1.3 2002 482 1.2 Donetsk 2004 300 2.0 2002 310 1.3 Lutsk 2004 300 1.3 2002 333 9.6 Odesa 2004 327 9.0 2002 300 1.7 Poltava 2004 300 3.7 2002 300 12.3 Simferopol 2004 356 4.8 2002 300 0.3 Kharkiv 2004 300 1.0 Sumy 2004 200 4.5 Kherson 2004 300 2.7 Source: Artyukh et al. (2005b). 39 ANNEX 2 Demographic Forecast: Methodology and Assumptions W e used the Futures Group15 modeling tool rate, mortality (based on Life Tables for Ukraine), Spectrum (DemProj and AIM modules) to and net migration. Demographic forecast is based on estimate demographic, economic, and social three major hypotheses: costs of the HIV/AIDS epidemic, projecting it from its onset in 1994 to 2014. The medium-term forecast Total fertility rate (TFR): Input TFR assumes that horizon allows us to remain confident about the key the decline observed in the 1990s has stabilized, with parameters and variables in the future: uncertainty a gradual increase to follow (to 1.33 children per about future values increases as we project out from female in 2010 and 1.41 in 2014). The share of the the baseline year. youngest females in total births decreases, while the proportion of births attributed to the 30+ age groups To forecast population and its age-gender composi- rises. tion, the medium demographic projection was used based on expert estimates by the Institute of Improved mortality indicators: Based on the Demography, National Academy of Sciences (NAS), analysis of the current socioeconomic situation and of the future level of demographic indicators. The the prospects of its improvement, we assumed that demographic scenario is based on assumptions about in the near future further deterioration of population future fertility, mortality, and migration. It also health subsides, mortality rates fall, and life assumes gradual improvement in Ukraine's socioeco- expectancy at birth improves. This hypothesis nomic situation, including sustained economic assumes that by 2014, life expectancy in Ukraine growth, increases in living standards, poverty reduc- reverts to its pre-crisis level of the early 1990s. tion, and improvement in the quality and accessibility Reduction in mortality rates and increased life of medical services. expectancy will be more pronounced for males. By 2014, forecast life expectancy at birth is 65.6 years The forecast is based on the assumption of the grad- for males and 75.8 for females. ual decrease in negative balance of migration, to a net outflow of 5,000 persons in 2009-10 and zero net Migration: There is a significant degree of uncer- outflow over 2011-14. Migration mobility of males tainty in making assumptions about future migration and their contribution to the balance of external patterns. Justification of any hypothesis with respect migration is projected to grow, causing the share of to migration flows requires consideration of many females among migrants to drop to 54 percent of the contradictory and sometimes ambiguous factors. In total in 2010. The age composition of migrants is cal- this analysis we accept the hypothesis of increased culated by extrapolating the smoothed annual data immigration and receding emigration. Among the on age distribution of migrants for 1994-2003. factors contributing to such dynamics are: Improved economic conditions that weaken the "push" factor for out-migration, Demographic Assumptions Baseline demographic inputs included Ukraine's 15 http://www.futuresgroup.com/Resources.cfm?area=2a&get= 1994 population by age and gender, total fertility Spectrum. 40 Table A2-1. Spectrum AIM Inputs: Adult HIV Prevalence, 1994-2014 In percentages 1994-2003 Year 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Adult HIV prevalence 0.11 0.16 0.24 0.34 0.48 0.65 0.87 1.13 1.40 1.67 2004-14 Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Adult HIV prevalence Medium 1.84 2.07 2.24 2.37 2.44 2.48 2.48 2.46 2.43 2.38 2.33 Optimistic 1.72 1.76 1.81 1.86 1.91 1.95 2.00 1.98 1.95 1.92 1.90 Pessimistic 1.90 2.10 2.24 2.33 2.45 2.56 2.75 2.94 3.12 3.31 3.5 Source: Authors' calculations using the UNAIDS Workbook Method spreadsheet of concentrated epidemic. Repatriation policies with respect to ethnic in size at the same rate as total adult population (a Ukrainians in other Commonwealth of negative growth rate of -1 percent per annum 2000- Independent States (CIS) countries and other eth- 2010 and -0.9 percent per annum for 2010-14). nic groups (such as Crimean Tartars) previously deported from Ukraine, and To estimate year/level of saturation in HIV preva- lence in high-risk groups under low to high scenar- A gradual weakening of "pull" factors for migra- ios, we used the latest serosurveillance data and the tion through ethnic links abroad. results of the surveys among the risk groups (to esti- mate the year of peak prevalence). In the medium Epidemiological Assumptions epidemic scenario we assume that the year of satura- tion/peak prevalence is 2007 for IDUs and 2010 for Adult HIV Prevalence Rate other risk groups. Low-high estimates of the risk To calculate adult HIV prevalence, an essential input groups sizes and group HIV prevalence allowed us to into the Spectrum AIM module, we used the UNAIDS calculate three scenarios of adult HIV prevalence: Workbook Method spreadsheet. This tool (an Excel® medium, optimistic, and pessimistic. Three scenarios spreadsheet) is used internationally to estimate and take identical values for 1994-2003 and diverge from project adult HIV prevalence from surveillance data 2004 (see Table A2-1): in countries with a low-level or concentrated epi- demic.16 Modeling options include (1) using sentinel In the model the epidemic starts in 1994 with an surveillance data on HIV prevalence in risk groups or adult HIV prevalence rate of 0.11 percent. In the (2) using antenatal clinic (ANC) data on HIV preva- medium epidemic scenario this rate peaks at 2.48 lence among low-risk females (urban and rural). Our percent in 2009-10 and gradually decreases there- modelling used the former and estimated the size of after to 2.33 percent in 2014. In the optimistic sce- model risk groups such as IDUs, their partners, nario, a peak prevalence rate of 2 percent is reached MSM, their female partners, CSWs, their clients, and in 2010, with the gradual reduction to 1.9 percent in STI patients. We assumed the groups were growing 2014. In the pessimistic scenario this rate increases over 2004-14 reaching 3.5 percent in 2014. These data were used as inputs to Spectrum AIM module. 16 http://www.unaids.org/en/resources/epidemiology/epi_software- tools.asp. 41 of earY 2007 2010 2010 2010 2007 2010 2010 2010 2010 2010 saturation prev 50.0% 10.0% 20.0% 7.0% 20.0% 5.0% 5.0% 5.0% 2.20% 1.50% Saturation data High 45.0% 6.0% 18.0% 5.00% 5.0% 2.0% 1.00% 3.00% 1.30% 0.90% year ANC Prevalence base Low 20.0% 2.0% 8.0% 3.00% 3.0% 1.5% 0.70% 1.00% 0.90% 0.80% 2020-30 -0.9% -0.9% -0.9% -0.9% -0.9% Partners -0.9% -0.9% -0.9% -0.9% -0.90% -0.90% rate x 2010-20 -0.9% -0.9% -0.9% -0.9% -0.9% -0.9% -0.9% -0.9% -0.9% -0.90% -0.90% growth one Prevalence) Annual 2005-10 -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.00% -1.00% Select Peak and Size 2000-05 -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.0% -1.00% -1.00% chosen MARPs in (MARPs in High MARPs size 700,000 250,000 70,000 300,000 of 500,000 125,000 1,000,000 1,100,000 8,876,000 3,804,000 Inputs year included base Partners Population already orkbook Low 70% 500,000 150,000 50,000 200,000 not 300,000 75,000 W 25,460,000 700,000 600,000 8,862,000 3,798,000 are women that Epidemic (MARPs) (PLR) populations workers low-risk risk sex to population population centage MSM of 15-49 per lower of populations workers at high-risk STIs low-risk Concentrated IDUs clients applied low-risk calculations. Group sex of of of of partners with data female A2-2. Population population female Authors' workers Most-at-risk IDU MSM Clients Populations Partners Partners Female Partners Patients ANC ce: ableT Sex Urban Rural Population Adult Urban 1. 2. a. b. Sour 42 Vertical transmission (MTCT). In 1994-98 the period 1994-2004. The ratio of female to male preva- mother-to-child transmission (MTCT) rate was esti- lence was 0.35 in 1994, 0.57 in 2000, and 0.76 in 2003. mated at 32 percent, falling to 28 percent by 1999. The trend was extrapolated to accommodate for the Ukrainian AIDS Center data indicate that the MTCT growing share of heterosexual transmission, reach- rate dropped to 15.9 percent in 2003 due to the avail- ing a ratio of 1 in 2014. ability of pre- and perinatal ARV prophylaxis. We start with this MTCT rate and assume its further ARV availability. The medium scenario assumes reduction: gradually to 5 percent in 2014 (medium); that in 2004, 1 percent of those needing ARV therapy to 10 percent in 2004 and gradually to 5 percent in have access to it, rising to 5 percent in 2005, to 10 2014 (optimistic); and gradual reduction to 10 per- percent in 2008 and remaining at that level. The opti- cent in 2014 (pessimistic) scenarios. mistic scenario assumes that in 2004, 1 percent of those in need have access, rising to 30 percent in Percentage of infants with AIDS dying in 2010, and then to 50 percent in 2014. The pessimistic first year. We attempted to calculate this share scenario assumes that in 2004, 1 percent of those in based on the Ukrainian AIDS Center data but were need have access, rising to 5 percent in 2005, and concerned with the bias of our estimate, due to limit- remaining at that level thereafter. ed availability of data. As a result, a Spectrum default value of 67 percent was used. AIDS impacts. In addition to three epidemic sce- narios, three plausible cost of treatment scenarios Life expectancy after AIDS onset. International were considered, reflecting the degree of uncertainty clinical experience suggests a 6-18-month life with respect to cost of treatment and access to treat- expectancy after AIDS onset. We used the default ment by those who require it: parameter (1 year) due to the lack of data for Ukraine. Scenario A. ARV-low, hospitalization-low assumes availability of low-price ART (UAH 1,500 Fertility of HIV-infected women (TFR reduc- per annum in constant prices17) throughout the tion). Based on the Ukrainian AIDS Center esti- modeling horizon. It is also assumed that starting mate, the ratio of fertility of HIV-infected women to from 0 percent in 2004, 30 percent of AIDS cases fertility of uninfected women is 0.7 for all age groups are hospitalized by 2014, at an annual cost of except teenagers; the TFR ratio for women aged UAH 1,500 per case; 15-19 is 1.2. Scenario B. ARV-high, hospitalization-low HIV progression. The incubation period (the assumes that the cost of an ARV treatment per interval between infection and the start of AIDS year is UAH 7,50018 in constant prices throughout symptoms) has been assigned the following the modeling horizon; 50 percent of AIDS cases Spectrum scenarios: are hospitalized by 2014, at an annual cost of UAH 7,500 per case; Scenarios Adults Children Scenario C. ARV-high, hospitalization-high Medium Quick Slow assumes that the cost of an ARV treatment per Pessimistic Quick Slow year is UAH 7,500 in constant prices throughout Optimistic Slow Slow the modelling horizon; 100 percent of AIDS cases are hospitalized by 2014, at an annual cost of HIV age distribution. HIV age distribution was UAH 7,500 per case. re-calculated separately for male and females based on the Ukrainian AIDS Center data. 17 Data supplied by the World Health Organization (WHO) Kiev Office, based on the Clinton Foundation best negotiated price for Sex ratio (those infected with HIV). This ratio generics. Exchange rate 1US$ = 5.3 UAH. was calculated based on the actual data from the 18Ministry of Health estimate, 2004. 43 Table A2-3. Estimated Cumulative Number of Those Infected (in Thousands) and Corresponding Adult HIV Prevalence Rate (Percentage), Ukraine, 2004-14 Year AIDS Medium AIDS Optimistic AIDS Pessimistic Adult Adult Adult prevalence prevalence prevalence Infected with HIV rate, % Infected with HIV rate, % Infected with HIV rate, % PREGNANT PREGNANT PREGNANT TOTAL WOMEN TOTAL WOMEN TOTAL WOMEN 2004 476.69 8.95 1.83 447.78 8.26 1.72 491.16 9.28 1.89 2005 536.75 10.34 2.06 459.96 8.51 1.76 544.12 10.49 2.09 2006 582.53 11.47 2.23 474.42 8.88 1.81 582.51 11.45 2.23 2007 616.42 12.39 2.37 487.97 9.27 1.86 606.18 12.09 2.33 2008 634.17 12.98 2.44 500.85 9.71 1.92 634.97 12.99 2.44 2009 640.67 13.25 2.49 508.61 9.98 1.96 657.94 13.73 2.56 2010 634.36 13.27 2.49 517.31 10.41 2.02 695.83 14.99 2.74 2011 620.93 13.26 2.47 508.9 10.49 2.02 731.07 16.31 2.93 2012 603.26 13.03 2.44 497.03 10.42 2.00 760.86 17.36 3.11 2013 580.97 12.77 2.39 486.39 10.47 1.99 791.3 18.52 3.3 2014 558.3 12.45 2.34 478.5 10.54 1.98 820.42 19.51 3.49 Source: Authors' calculations. Table A2-4. Forecasted New Cases of HIV Infection and AIDS, and Annual AIDS Deaths In thousands Year AIDS Medium AIDS Optimistic AIDS Pessimistic NEW CASES NEW CASES NEW CASES HIV AIDS AIDS deaths HIV AIDS AIDS deaths HIV AIDS AIDS deaths 2004 63.32 23.84 18.38 27.24 13.7 9.78 77.69 23.76 18.3 2005 82.35 29.49 23.1 26.13 18.11 13.18 75.21 29.69 22.99 2006 72.92 35.71 28.8 31.6 22.59 17.19 65.46 35.91 29.46 2007 66.5 41.51 34.76 34.52 26.42 21.23 56.95 41.69 35.68 2008 56.16 46.89 40.68 37.67 30.4 24.67 68.11 46.97 41.49 2009 50.76 51.38 46.39 36.04 33.99 28.13 68.07 51.66 46.79 2010 43.37 55.08 50.96 40.2 36.63 31.26 88.08 55.71 51.52 2011 40.62 57.59 54.72 25.61 38.19 33.49 90.07 59.31 55.6 2012 39.88 58.9 57.32 23.9 38.74 34.82 88.58 62.4 59.21 2013 37.58 59.06 58.73 26.12 38.2 35.27 92.7 64.95 62.33 2014 38.32 58.16 58.99 29.33 36.77 34.81 94.35 67.26 64.91 Source: Authors' calculations. 44 Table A2-5. Forecasted Number of Those Infected with HIV, Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa Oblasts, 2004-14 Number infected with HIV, thousand Percentage of total 2004 2010 2014 2014 Oblast Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Dnipropetrovsk ­ 32.6 ­ 77.6 ­ 85.3 ­ 10.4 Donetsk 51.3 52.5 92.7 103.5 92.2 105.6 19.3 12.9 Mykolayiv 15.6 15.0 30.7 36.7 32.5 44.2 6.8 5.4 Odesa 29.9 25.1 49.5 72.6 48.9 116.1 10.2 14.2 4 oblasts in total 96.8 125.2 172.9 290.4 173.6 351.2 36.3 42.9 Ukraine 447.8 491.2 517.3 695.8 478.5 820.4 100.0 100.0 Source: Authors' calculations. Table A2-6. Forecasted Annual AIDS Deaths, Dnipropetrovsk, Donetsk, Mykolayiv and Odesa Oblasts, 2004-14 In thousands 2004 2010 2014 Oblasts Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Dnipropetrovsk ­ 1.4 ­ 4.1 ­ 6.7 Donetsk 1.0 1.8 3.6 6.2 5.5 9.2 Mykolayiv 0.3 0.5 1.1 1.6 1.9 2.9 Odesa 0.5 0.9 2.0 2.7 3.3 5.6 4 oblasts in total 1.8 4.6 6.7 14.6 10.7 24.4 Ukraine 9.8 18.3 31.3 51.5 34.8 64.9 Source: Authors' calculations. Table A2-7. Forecasted Accumulated AIDS Deaths, Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa Oblasts, 2004-14 Accumulated AIDS deaths, thousand Percentage of total 2004 2010 2014 2014 Oblasts Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Optimistic Pessimistic Dnipropetrovsk ­ 3.9 ­ 19.9 ­ 43.1 ­ 8.2 Donetsk 2.9 6.0 17.2 30.6 36.8 63.3 12.2 12. Mykolayiv 0.8 1.7 5.0 8.0 11.2 17.5 3.7 3.3 Odesa 1.6 3.3 9.6 14.3 20.8 31.7 6.9 6. 4 oblasts in total 5.3 14.9 31.8 72.8 68.8 155.6 22.8 29.5 Ukraine 27.2 56.4 162.9 284.3 301.3 526.4 100.0 100.0 Source: Authors' calculations. 45 Table A2-8. Estimated Indicators of AIDS-Related Mortality in the Working-age (15-59) Population, Dnipropetrovsk, Donetsk, Mykolayiv, and Odesa Oblasts, 2014 Annual AIDS deaths, Percentage of AIDS deaths in total 15-59 age group, `000 number of deaths,15-59 age group Oblasts Optimistic Pessimistic Optimistic Pessimistic Dnipropetrovsk ­ 6.5 ­ 38.8 Donetsk 5.3 8.9 22.8 33.2 Mykolayiv 1.8 2.8 26.9 36.1 Odesa 3.1 5.3 25.1 36.6 Ukraine 33.2 61.7 17.0 27.8 Source: Authors' calculations. Table A2-9. Estimated Reduction in Life Expectancy due to AIDS, Dnipropetrovsk, Donetsk, Mykolayiv and Odesa Oblasts, 2014 In years Life expectancy Life expectancy with AIDS AIDS-induced reduction in life without AIDS expectancy Oblasts Optimistic Pessimistic Optimistic Pessimistic Males Dnipropetrovsk 64.4 ­ 59.0 ­ 5.4 Donetsk 63.8 60.1 58.4 3.7 5.4 Mykolayiv 63.4 59.3 56.9 4.1 6.5 Odesa 64.1 60.7 57.0 3.4 7.1 Ukraine 65.6 63.4 61.6 2.2 4.0 Females Dnipropetrovsk 74.9 ­ 68.7 ­ 6.2 Donetsk 75.0 70.1 68.3 4.9 6.7 Mykolayiv 74.3 69.1 66.6 5.2 7.7 Odesa 73.9 69.6 66.1 4.3 7.8 Ukraine 75.8 72.9 71.0 2.9 4.8 Source: Authors' calculations. 46 Table A2-10. Forecasted Total Population in the Range of AIDS Scenarios, Dnipropetrovsk, Donetsk, Mykolayiv and Odesa Oblasts In millions 2004 2014 Oblasts No-AIDS No-AIDS AIDS optimistic AIDS pessimistic Dnipropetrovsk 3.49 3.27 ­ 3.23 Donetsk 4.69 4.25 4.21 4.19 Mykolayiv 1.23 1.17 1.16 1.15 Odesa 2.41 2.29 2.27 2.26 Source: Authors' calculations. 47 ANNEX 3 Methodology for Estimating Labor Force and Employment Analysis at National Level period. This category includes those employed or actively looking for a job (unemployed by the ILO T his section presents projectons of the follow- definition). Other categories, such as students, pen- ing indicators: working-age population, total sioners (disability and other categories), unpaid labor force, and employed and unemployed home-makers and domestic caregivers, discouraged population, by five-year age group and gender, over job seekers, and others not actively seeking a job are the 2004-14 period. classified as "not in labor force." Working-age Population Labor force forecast is implemented using the demo- The study assumes any person aged 15-70 is a work- graphic forecast of the working-age population and ing-age person ("adult") who may be either in or out the estimated population trend for groups outside of the labor force. The five-year age group and gen- the labor force. The projected labor force declines der projections of by 2.3 million persons (11.5 percent) over 2004-14. these adults are based Table A3-1. Estimated Increase in Based on the State Statistics Committee data for on actual labor force Population by Age 1998-2003, it has been declining, with the overall participation data for Group, 2004-14 reduction of 3.3 million (1.2 million males and 2.1 1998-2003 and the million females). The gender structure of labor force Percentage demographic forecast has also changed, with the share of males in total Age group increase from Chapter 3. It is labor force increasing from 49.2 percent in 1998 to 25-29 6.9 estimated that the 51.1 percent in 2003 (and a corresponding share of working-age popula- 30-34 4.8 females declining from 50.8 percent to 48.9 percent). tion declines by 2.4 This tendency is preserved over the ten-year forecast 35-39 5.7 million over the fore- horizon, with the share of males in total labor force 50-54 2.7 cast period. The num- reaching 52.3 percent by 2014. Figure A3-1 presents 55-59 39.6 ber of adult females the labor force projection by gender. falls faster than that of Source: Demographic forecast (Chapter 3). males. The share of The economically active population is estimated to males (females) in the decrease in virtually all age groups, due to the under- total adult population is 47 percent (53 percent) and lying demographic forecast and decreased labor is constant within the forecast period. Note that the force participation rates in some age groups. population in several five-year age groups increases However, we predict an increase in the economically (see Table A3-1). active population aged 35-39. Labor Force (Economically Employed Population Active Population) By definition, people aged 15-70 are considered Labor force, or the economically active population, employed if during the reference period (the week of is comprised of adults (15-70) of both sexes who the survey) they supplied at least one hour to a labor supplied labor in a labor market during a reference 48 market in paid employment or Figure A3-1. Labor Force Reductions by Gender, 2004-14 for an in-kind reward; or they were self-employed or worked 100.00 for their family business; or 99.50 they worked at least 30 hours 99.00 per week without pay for a fam- ily business or enterprise, 98.50 including own farm, in order to Index 98.00 sell the enterprise's product; or they were temporarily absent 97.50 Males Females from their official 97.00 workplace/own business due to 96.50 circumstances beyond their 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 control. Source: Authors' calculations. The analysis was conducted by five-year age group and gender, using actual 2014: 10.4 percent. In particular, the number of employment data by age-gender group, projected employed males will decrease over the forecast hori- GDP growth rates, and projected labor productivity. zon by 8.3 percent (from 10.4 million to 9.5 million) Employment in period t+1 is given by equation (1) with a corresponding 12.6 percent drop for females GDPt+1 (from 10.03 million to 8.8 million). Despite the gener- Eit+1 = Eit * , (1) PLt+1 al decline, absolute employment grows in some age groups. For instance, for males we observe an GDPt+1 increase in the number of employed by 6 percent in Eit * PLt+1 Eit the 20-24 age group during 2004-08; by 9.7 percent in if , where (2) EAit+1 EAit the 55-59 age group over 2004-09; by 8.3 percent in the 25-29 age group; and by 2.4 percent in the 35-39 Eit+1 , Eit ­ employment in the age group i in year age group over 2004-14. For females, we estimate an t+1 and t; increase in the number employed by 7.6 percent in GDPt+1 ­ change in GDP over [t; t+1); the 20-24 age group over 2004-09; by 12.4 percent in the 35-39 age group over 2004-14; and by 40.5 per- PLt+1 ­ change in labor productivity (value added cent in the 55-59 age group over 2004-11 (Table A3-2 per employee); on page 50). EAit , EAit+1 ­ number of economically active people from age group i during Figure A3-2 on page 50 shows that there are also period t and t+1; changes in the relative index of employment by gen- i ­ index of a five-year age group (from 15-70); der; the share of employed in the total of economi- t - base year; t+1 - forecast year. cally active population marginally increases (from 0.9055 to 0.9057) for males and decreases (from If inequality (2) does not hold, next period of 0.9124 to 0.9119) for females over 2004-14. employment is calculated using last period share of employment (equation (3)): Unemployed Population Eit Based on the ILO definition, the unemployed are per- Eit+1 = EAit+1 * (3) EAit sons 15-70 years of age (either registered with the State Employment Service or not) who meet all of According to our forecast, total employment will the following requirements: decline from 20.4 million in 2004 to 18.3 million in a) Not in paid employment or self-employed; 49 Table A3-2. Estimated Increase in Absolute Employment, by Age between the economically active and Group and Gender employed populations. We estimate that the unemployed population decreases over Males Females the forecast horizon from 2.1 million in Percentage Percentage 2004 to 1.8 in 2014, or by 10.2 percent. In Age group Period increase Age group Period increase particular, the number of unemployed 20-24 2004-08 6.01 20-24 2004-09 7.59 decreases over this period by 8.52 percent 25-29 2004-14 8.29 for males (from 1.1 million in 2004 to 994,000 in 2014) and by 12.1 percent for 35-39 2004-14 2.37 35-39 2004-14 12.36 females (from 963,000 in 2004 to 847,000 in 55-59 2004-09 9.74 55-59 2004-11 40.47 2014). This corresponds to the reduction in unemployment rate from 9.5 percent to 9.4 Source: Authors' calculations. percent for males over 2004-14, with only a marginal change for females whose unem- b) Actively looking for a job or trying to set up their ployment rate is about 8.8 percent (Figure A3-3) own business within the past 4 weeks prior to survey; Labor Force in Three HIV/AIDS c) Available to start at a new job within the following Epidemic Scenarios 2 weeks. Based on the demographic and epidemiological fore- casts from Chapter 3 and using the methodology of The unemployed category also includes those who the previous section, we constructed labor force pro- have received an offer of paid employment and will jections in three epidemic scenarios: medium, opti- start some time in the future, as well as those in mistic, and pessimistic. The forecast demonstrates training at the request of the State Employment that the decline in working-age, economically active, Service. In line with the ILO methodology, the unem- employed, and unemployed populations is occurring ployed population is calculated as the difference faster in all three "with-AIDS" scenarios than those in the "no-AIDS" scenario of the previous section. Figure A3-2. Share of Employed in Total Economically Active Population by Gender, 2004-14 Working-age population: 0.9140 Impact of HIV/AIDS. We esti- mate that during 2004-14 the 0.9120 total working-age population will decrease by 2.7-2.9 million 0.9100 persons, depending on scenario 0.9080 (see Figure A3-4). Share 0.9060 We calculate additional losses in the working-age population 0.9040 due to the impact of HIV/AIDS 0.9020 epidemic. By 2014, the reduc- tion in this population attrib- 0.9000 uted to HIV/AIDS constitutes 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 441,000 in the medium epidem- Males Females ic scenario (an additional 1.3 percent compared to the "no- Source: Authors' calculations. AIDS" scenario), 268,000 (an 50 additional 0.8 percent) in opti- Figure A3-3. Unemployment Rates by Gender, 2004-14 mistic and 472,000 (an additional 1.4 percent) in pessimistic scenar- 9.7 ios, respectively (see Table A3-3 9.5 on page 52). 9.3 In some five-year age groups where an absolute increase was 9.1 observed in the "no-AIDS" sce- Percent nario, the magnitude of the 8.9 increase diminished in all three epidemic scenarios with HIV/AIDS 8.7 (Table A3-4 on page 52). 8.5 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Labor force: Impact of HIV/AIDS. Total labor force Males Females decreases in all epidemic scenarios Source: Authors' calculations. over the forecast horizon by 2.5-2.7 million (Figure A3-5 on page 52). Figure A3-4. Projected Working-Age Population in the Range of AIDS Scenarios, 2004-14 36,500 36,000 35,500 35,000 34,500 34,000 Thousand 33,500 33,000 32,500 32,000 31,500 2004 2006 2008 2010 2012 2014 No-AIDS Medium Optimistic Pessimistic Source: Authors' calculations. 51 Table A3-3. Estimated Losses in the Working-Age Population due to the HIV/AIDS Epidemic, 2004-14 Losses in working-age Additional losses due to Working age Working age population HIV/AIDS population, population, 2004 2014 Total % of "no-AIDS" Total % of "no-AIDS" Scenario (million) (million) (million) scenario (thousand) scenario No-AIDS 36.17 33.75 2.42 -- -- -- AIDS Medium 36.15 33.29 2.86 18.21 441.1 1.31 AIDS Optimistic 36.13 33.44 2.69 11.07 268.1 0.79 AIDS Pessimistic 36.15 33.26 2.89 19.52 472 .8 1.40 Source: Authors' calculations. Table A3-4. Estimated Impact of HIV/AIDS on Selected Working-Age Groups, 2004-14 % Increase AIDS Optimistic AIDS Pessimistic in "no-AIDS" Age Group scenario % Increase Difference % Increase Difference 25-29 6.9 6.6 0.3 5.8 1.1 30-34 4.8 3.3 1.5 1.5 3.3 35-39 5.7 3.5 2.2 2.2 3.5 50-54 2.7 2.1 0.6 1.8 0.9 55-59 39.6 39.4 0.3 38.9 0.7 Source: Authors' calculations. Figure A3-5. Estimated Impact of HIV/AIDS on Labor Force, 2004-14 23.0 22.5 22.0 21.5 Thousand 21.0 Millions 20.5 No-AIDS Hundred AIDS Pessimistic 20.0 AIDS Optimistic 19.5 19.0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 52 By 2014, additional losses in the labor force due to Table A3-5. Estimated Losses in Labor Force due the HIV/AIDS epidemic are estimated to be 323,000 to the HIV/AIDS Epidemic, 2004-14 (an additional 1.6 percent compared to the "no- In thousands AIDS" demographic scenario) in the medium epi- demic scenario, 193,000 (an additional 1 percent) in Losses due to HIV/AIDS the optimistic one, and 351,000 (an additional 1.7 Total Percentage additional to percent) in the pessimistic scenario (Table A3-5). Scenario "no-AIDS" scenario AIDS Medium 323.4 1.6 Note that the HIV/AIDS epidemic causes reductions AIDS Optimistic 193.2 0.96 in the number of economically active in all age AIDS Pessimistic 351.0 1.74 groups, weakening the observed increase in the absolute number of economically active among cer- Source: Authors' calculations. tain age groups reported in the earlier section (see Table A3-6). Employed Population: Table A3-6. Estimated Impact of HIV/AIDS on the Economically Active Population Impact of HIV/AIDS n Certain Age Groups, 2004-14 The total employed popula- Age groups with Percentage Increase, Percentage Increase, Percentage Increase, tion is projected to decline with an increased "no-AIDS" AIDS Optimistic AIDS Pessimistic over the forecast horizon in economic activity Males Females Males Females Males Females all epidemic scenarios by 2.3- 25-29 8.1 -- 8.0 -- 7.5 -- 2.4 million (Figure A3-6). 35-39 2.4 12.4 -- 10.4 -- 9.7 Results by gender are pre- sented in Table A3-7 on page Source: Authors' calculations. 54. By 2014, additional losses in employed population due to the HIV/AIDS epidemic are Figure A3-6. Estimated Impact of HIV/AIDS on the Employed Population, 2004-14 estimated to be 171,000 (an additional 1 percent com- 21.0 pared to the "no-AIDS" demo- graphic scenario) in the opti- No-AIDS mistic epidemic scenario, and 20.5 AIDS Optimistic 302,000 (an additional 1.7 per- AIDS Pessimistic cent) in the pessimistic one 20.0 (Tables A3-7 and A3-8 on page 54). 19.5 Millions Analysis at 19.0 Regional Level To assess the impact of the 18.5 HIV/AIDS epidemic on the regional level, we selected 18.0 four oblasts: Donetsk, 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Dnipropterovsk, Odesa and Mykolayiv. The calculations Source: Authors' calculations. 53 Table A3-7. Percentage of Estimated Losses in the Employed Population due were based on four demographic to the HIV/AIDS Epidemic by Gender, 2004-2014 forecasts: the "no-AIDS" scenario and three epidemic scenarios with Percentage decrease in employment HIV (medium, optimistic, and pes- No-AIDS AIDS medium AIDS optimistic AIDS pessimistic simistic). We estimate the working- age, economically active, and Total employed 10.40 11.81 11.25 11.88 employed and unemployed popula- Males 8.28 10.05 9.23 10.65 tions by gender using the same Females 12.61 13.64 13.35 13.69 methodological approaches as in the national level analysis. Results Source: Authors' calculations. confirm the expected reduction in the size of the working-age population and the labor Table A3-8. Total Estimated Losses in the force, including a reduction in the employed popula- Employed Population due to the HIV/AIDS tions in all four oblasts. Epidemic, 2004-14 In thousands Donetsk Oblast The labor force forecast was conducted using three Losses in employed population scenarios: "no-AIDS" demographic scenario and two Scenario due to HIV/AIDS "with-AIDS" scenarios, optimistic and pessimistic. Total Percentage additional to The total working-age population of Donetsk Oblast "no-AIDS" scenario is projected to decrease within the forecast period AIDS medium 286.4 1.6 (2004-2014) by 4,005,000 in the "no-AIDS" scenario AIDS optimistic 170.4 1.0 and by 4,419,000 (4,725,000) in optimistic (pes- AIDS pessimistic 301.7 1.7 simistic) epidemic HIV/AIDS scenario, representing an additional loss of 414,000-720,000 persons of Source: Authors' calculations. working age by 2014 due to the epidemic. Table A3-9 compares projected working-age population under optimistic and pessimistic scenarios to the Table A3-9. Additional Losses in Working-age Population due to "no-AIDS" demographic scenario. the HIV/AIDS Epidemic, Donetsk Oblast, 2004-14 The projected labor force will decline in all In thousands scenarios: by 232,000 (10.1 percent in the "no-AIDS") and by 2,925,000 (3,131,000) or Losses in working-age population 12.8 percent (13.7 percent) in optimistic Working-age Working-age population, population, % of "no-AIDS" (pessimistic) epidemic scenarios, respec- 2004 2014 Total scenario tively. Thus, by 2014 the HIV/AIDS epidem- No-AIDS 3,686.7 3,286.1 400.5 -- ic causes and additional 2.7 percent (3.6 percent) decline in Donetsk's labor force in AIDS optimistic 3,686.7 3,244.7 441.9 10.3 the optimistic (pessimistic) epidemiologi- AIDS pessimistic 3,686.7 3,214.1 472.5 18.0 cal scenario. Estimated labor force by gen- der in the pessimistic scenario is plotted in Source: Authors' calculations. Figure A3-7. Figure A3-8 presents changes in the total labor force over 2004-14. Direct losses in labor force due to the impact of the HIV/AIDS epidemic constitute 605,000 in the opti- 54 mistic and 811,000 persons in Figure A3-7. Estimated Labor Force by Gender in Pessimistic Scenario, the pessimistic scenario. Donetsk Oblast, 2004-14 Projected employed popula- 1,200,000 tion in Donetsk Oblast also declines (Figure A3-9 on page 1,150,000 Males Females 56), with larger losses observed for males than females. 1,100,000 In the "no-AIDS" demographic 1,050,000 scenario, by 2014 the number of employed males (females) in Donetsk Oblast declines by 1,000,000 1,087,000 (1,049,000) persons, or 10.2 percent (10.0 percent) 950,000 compared to 2004. The corre- sponding decline in the AIDS 900,000 optimistic scenario is 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1,467,000 (1,227,000) for males (females), representing an Source: Authors' calculations. additional 3.5 percent (1.8 per- cent) decline compared to the demographic trend. In the pessimistic scenario, the projected decline in Figure A3-8. Estimated Total Labor Force in Three Scenarios ("No-AIDS," employed population is 1,478,000 AIDS Optimistic and AIDS Pessimistic), Donetsk Oblast, (1,404,000) for males (females), or an 2004-14 additional 3.7 percent (3.5 percent) decline compared to the demographic 2,350,000 trend. Thus, additional losses in 2,300,000 No-AIDS Donetsk's employed population due to AIDS Optimistic the epidemic are estimated at 558,000 in 2,250,000 AIDS Pessimistic the optimistic and 747,000 in the pes- 2,200,000 simistic epidemic scenarios (Table A3-10 on page 57). 2,150,000 2,000,000 Within the entire forecast period and 2,050,000 under all scenarios, there is the trend toward a decline in the total unemployed 2,000,000 population as a result of reduction in the 1,950,000 population as a whole. The forecast of unemployed persons is presented in 1,900,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Figure A3-10 on page 56. It demonstrates that in both epidemic scenarios, the num- Source: Authors' calculations. ber of unemployed decreases over the forecast horizon. 55 Figure A3-9. Projected Employed Population, Donetsk Oblast, 2004-14 2,150,000 2,100,000 No-AIDS AIDS Optimistic 2,050,000 AIDS Pessimistic 2,000,000 1,950,000 1,900,000 1,850,000 1,800,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. Figure A3-10. Projected Unemployed Population, Donetsk Oblast, 2004-14 185,000 No-AIDS 180,000 AIDS Optimistic AIDS Pessimistic 175,000 170,000 165,000 160,000 155,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 56 Table A3-10. Estimated Losses in Employed Population due to the HIV/AIDS Epidemic, Donetsk Oblast, 2004-14 In thousands Losses in employed population Losses in employed population due to HIV/AIDS Employed Employed Additional as % Additional as % Scenarios Population, 2004 Population, 2014 Total of "no-AIDS" increase Total of "no-AIDS" increase No-AIDS 2,112.6 1,899.0 213.6 -- -- -- AIDS optimistic 2,109.1 1,839.7 269.4 12.67 55.8 26.1 AIDS pessimistic 2,108.8 1,820.6 288.2 13.77 74.7 35.0 Source: Authors' calculations. Dnipropetrovsk Oblast additional 3.9 percent to the baseline demographic Using the same methodology as above, a hypotheti- losses. cal "no-AIDS" demographic forecast was constructed for Dnipropetrovsk Oblast, supplemented with an Projected employed population in Dnipropetrovsk epidemic forecast of HIV/AIDS in a pessimistic Oblast declines over 2004-14 (see Figure A4-14 on scenario only, due to data limitations. The total page 59), by 134,000 (8.2 percent) in the "no-AIDS," working-age population of the oblast is projected and by 198,000 (12.1 percent) in pessimistic epidem- to decrease within the forecast period (2004-2014) by 2,274,000 (8.2 percent) in the "no-AIDS" scenario and by Figure A3-11. Estimated Working-Age Population, Dnipropetrovsk Oblast, 2,894,000 (10.5 percent) in the pes- 2004-14 simistic epidemic scenario. This consti- tutes an additional loss of 62,000 per- 2,800,000 sons (2.3 percent) of working age due to HIV/AIDS over 2004-14 (Figure No-AIDS 2,750,000 A3-11). AIDS Pessimistic 2,700,000 Projected labor force declines in both scenarios, by 146,000 (8.2 percent) in 2,650,000 "no-AIDS" and by 214,000 (12.1 per- cent) in pessimistic epidemic scenar- ios. Estimated labor force by gender in 2,600,000 the pessimistic scenario is plotted in Figure A3-12 on page 58. Note that the 2,550,000 number of economically active males declines faster than that of females (a 2,500,000 reduction of 11.1 percent over 2004-14 compared to 9.9 percent). Figure A3-13 2,450,000 on page 58 depicts the total projected 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 labor force. Additional losses in labor force due to the HIV/AIDS epidemic are estimated at 68,000 persons, or an Source: Authors' calculations. 57 Figure A3-12. Estimated Labor Force by Gender in Pessimistic Scenario, Dnipropetrovsk Oblast, 2004-14 950,000 900,000 Males Females 850,000 800,000 750,000 700,000 650,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. Figure A3-13. Estimated Total Labor Force, Dnipropetrovsk Oblast, ic scenarios. The decline in employed 2004-14 population is more pronounced for males than females. Additional losses in 1,800,000 the employed population due to the HIV/AIDS epidemic amount to 66,000 by 1,750,000 2014, an additional 3.9 percent decline compared to the demographic trend. 1,700,000 Within the forecast period in both sce- narios, the unemployed population in 1,650,000 Dnipropetrovsk Oblast declines (see Figure A3-15). 1,600,000 Odesa Oblast No-AIDS 1,550,000 AIDS Pessimistic In the "no-AIDS" demographic scenario, the working-age population of Odesa 1,500,000 Oblast is projected to decrease by 2014 by 7.6 percent from its 2004 level of 1.9 million. This represents a loss of 1,450,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1,425,000, including 61,100 males (a 6.9 percent reduction) and 81,400 females Source: Authors' calculations. 58 (a reduction of 8.3 percent). When the Figure A3-14. Estimated Total Employment, Dnipropetrovsk Oblast, impact of HIV/AIDS is added, the 2004-14 working-age population is projected to decline over the period by 162,900 1,700,000 (8.7 percent) in the optimistic and 173,100 (10.2 percent) in the pes- No-AIDS 1,650,000 simistic epidemic scenarios. The AIDS Pessimistic labor force in Odesa Oblast is also projected to decrease over the fore- 1,600,000 cast horizon: by 85,100 (7.5 percent) in the "no-AIDS" and by 110,200 1,550,000 (132,900) or 9.7 percent (11.7 per- cent) in the optimistic (pessimistic) scenarios. Table A3-11 on page 60 1,500,000 compares projected labor force under the HIV/AIDS scenarios to the "no- AIDS" scenario, allowing us to esti- 1,450,000 mate additional losses in labor force due to the epidemic. 1,400,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 The employed population in Odesa Oblast under the "no-AIDS" scenario Source: Authors' calculations. will drop from 1,077,000 in 2004 to 995,000 in 2014, that is, by 81,500 (7.6 percent). The number of males will decline over this period by 34,900 (7.0 Figure A3-15. Projected Unemployed Population, Dnipropetrovsk Oblast, percent), the number of females by 2004-14 42,100 (8.2 percent). Figure A3-16 on page 60 presents the projected 145,000 employed population under the "no- No-AIDS AIDS" and two epidemic scenarios, 140,000 AIDS Pessimistic by gender. 135,000 In the "AIDS optimistic" scenario, the decline in the employed population 130,000 over the forecast horizon constitutes 93,000 persons (8.7 percent), or an 125,000 additional loss of 11,500 persons (1.2 percent) due to the epidemic. Similarly, employment losses under 120,000 the pessimistic epidemic scenario are 116,100 (10.9 percent), an additional 115,000 reduction of 23,100 (3.3 percent) com- pared to the "no-AIDS" trend over 110,000 2004-14. The unemployed population 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 will decrease by 2,700 (4.7 percent) in the "no-AIDS" scenario. The epidemic Source: Authors' calculations. 59 Table A3-11. Losses in Labor Force due to the HIV/AIDS Epidemic, Odesa Oblast, 2004-14 In thousands Losses in labor force Additional losses due to HIV/AIDS Additional % of % of 2004 Scenario Total % Total "no-AIDS" decrease no-AIDS No-AIDS Total 85.0 7.5 -- -- -- Males 40.9 6.9 -- -- -- Females 44.1 8.2 -- -- -- AIDS optimistic Total 110.2 9.7 25.2 2.2 29.6 Males 59.1 9.9 18.2 3.0 44.5 Females 51.1 9.5 7.0 1.3 15.9 AIDS pessimistic Total 133.0 11.7 48.0 4.2 56.5 Males 72.1 12.1 31.2 5.2 76.3 Females 60.9 11.3 16.8 3.1 38.1 Source: Authors' calculations. Figure A3-16. Projected Employed Population by Gender, Odesa Oblast, 2004-14 580,000 560,000 540,000 520,000 500,000 480,000 460,000 440,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Males, No-AIDS Females, AIDS Optimisitc Females, No-AIDS Males, AIDS Pessimistic Males, AIDS Optimisitc Females, AIDS Pessimistic Source: Authors' calculations. 60 results in an additional reduction Figure A3-17. Projected Unemployed Population, Odesa in unemployed population of Oblast, 2004-14 2,500-3,800 persons, or an addi- tional 4.3-6.6 percent, depending 75,000 on the epidemic scenario (Figure No-AIDS A3-17). 70,000 AIDS Optimistic AIDS Pessimistic Mykolaiv Oblast 65,000 Within the forecast period the working-age population in Mykolayiv Oblast decreases by 60,000 7.5 percent in the "no-AIDS" sce- nario, and by 8.6 percent in the 55,000 optimistic scenario with HIV/AIDS. By 2014, the labor 50,000 force declines by 46,100 (7.8 per- 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 cent) from its 2004 level of 591,400 in the "no-AIDS" opti- Source: Authors' calculations. mistic scenario. Figure A3-18 illustrates the reduction in labor force under the "no-AIDS" and two epidemic scenarios. Figure A3-18. Projected Labor Force, Mykolaiv Oblast, 2004-14 The HIV/AIDS epidemic induces a reduction in labor force partici- pation rate for Mykolayiv Oblast 600,000 No-AIDS AIDS Optimistic by 2 percentage points over 590,000 AIDS Pessimistic 2004-14, from 61.9 percent to 580,000 59.7 percent. Additional losses in the labor force due to the epi- 570,000 demic by 2014 are 11,500-24,400 560,000 persons (see Table A3-12 on 550,000 page 62). 540,000 The projected decrease in the 530,000 employed population in Mykolayiv Oblast within the ten- 520,000 year horizon is 39,200 or 7.7 per- 510,000 cent in the "no-AIDS" scenario, 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 8.7 percent in the optimistic, and 10.8 percent in the pessimistic Source: Authors' calculations. epidemic scenarios (Figure A3-19 on page 62). Additional losses due to the HIV/AIDS epidemic are 5,000-15,000 persons. 61 Table A3-12. Projected Labor Force in Three Scenarios, Mykolayiv Oblast, 2004-14 In thousands Additional losses in labor Losses in labor force force due to HIV/AIDS Labor force Labor force % "no-AIDS" % of losses under 2004 2014 Total 2004 Total "no-AIDS" scenario No-AIDS 591.4 545.3 46.1 -- -- -- AIDS optimistic 590.7 533.1 57.6 9.7 11.5 24.9 AIDS pessimistic 590.0 519.5 70.5 11.9 24.4 52.9 Source: Authors' calculations. Figure A3-19. Projected Employed Population, Mykolayiv Oblast, 2004-14 520,000 510,000 No-AIDS AIDS Optimistic 500,000 AIDS Pessimistic 490,000 480,000 470,000 460,000 450,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 62 ANNEX 4 Methodology and Assumptions for Estimating the Impact of HIV/AIDS on Government Budget and Social Insurance Funds MAIN OBJECTIVES: Estimation of the State Budget/Special Funds Revenue 1) Estimation of the state budget/special funds Forgone Due to Reduced Employment revenue forgone due to reduced employment; The following payroll deductions are made from the 2) Estimation of the additional expenditure by salaries of employed persons (see Table A4-1): the special fund (temporary disability) on temporary disability payments to those Personal income tax transferred to the state infected with HIV; and budget (13 percent of wages and salaries); 3) Estimation of the additional expenditure by Mandatory state pension insurance levy trans- the state budget and the special fund (perma- ferred to the pension fund (1-5 percent of salaries); nent disability) due to an increase in AIDS- Temporary disability insurance levy transferred related permanent disability cases. to the social fund (temporary disability) (0.5-1 percent of salaries); Table A4-1. Taxes and Levies on Wages and Income, Ukraine, 2004 Unemployment insurance levy transferred to the Deductions from wages and salaries social fund (unemploy- Income amount Deduction rate ment) (0.5 percent of salaries). Personal income tax (state budget) Any 13% Pension fund levy As Table A4-1 shows, the Public servants UAH 150-250 1% withholding rate for social UAH 250-350 2% funds depends on income UAH 350-450 3% level. The expert estimate was used to arrive at the UAH 450-500 4% average withholding rates of UAH 500+ 5% 2 percent for the pension Other employees Up to UAH 150 1% fund and 1 percent for the Above UAH 150 2% social insurance fund (tem- Social insurance fund Up to UAH 453 0.5% porary disability). (temporary disability) levy (official subsistence level for a working-age person) First, the number of Above UAH 453 1% employed was estimated Social insurance fund Any 0.5% under two epidemic scenar- (unemployment) levy ios (total and by gender). Source: Ministry of Labor and Social Policies. 63 Table A4-2. Reduction in Employment from "No-AIDS" Baseline, Two AIDS Scenarios, 2004-14 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1. Optimistic scenario (with AIDS) Total 24,680 37,010 44,853 52,184 70,354 89,997 110,242 130,125 141,100 181,053 195,135 (persons) Males 13,192 15,819 17,985 23,694 35,067 52,153 55,263 70,720 83,992 103,588 111,074 Females 11,487 21,191 26,868 28,490 35,286 37,844 54,979 59,404 57,108 77,466 84,061 2. Pessimistic scenario (with AIDS) Total 28,046 42,933 55,621 72,247 100,118 130,389 158,057 199,065 225,470 260,659 309,715 (persons) Males 12,271 27,004 35,756 44,999 61,238 81,491 95,725 120,909 140,943 168,784 196,309 Females 15,775 15,929 19,865 27,248 38,880 48,898 62,333 78,155 84,526 91,875 113,406 Difference between scenarios (persons) Total 3,366 5,923 10,767 20,063 29,764 40,392 47,816 68,940 84,370 79,606 114,581 Source: Authors' calculations. The analysis was conducted by age-gender groups. 195,100 in 2014 in the optimistic, and from 28,100 to After taking differences of these projections with the 309,700 in the pessimistic scenarios. Under both "no-AIDS" projection of employed population, an scenarios the losses among males exceed those estimated loss in employment due to the HIV/AIDS among females. Losses under pessimistic scenarios epidemic was calculated (for males, females, and exceed those under optimistic by 114,600 in 2014 total). Table A4-2 reports the final results under (Figure A4-1). two scenarios. After the reduction in employment is estimated, for- Over the forcast period losses in employment due to gone state budget revenue is calculated as the HIV/AIDS will increase from 24,000 in 2004 to amount of unpaid personal income tax; forgone pen- sion fund revenue (as unpaid Figure A4-1. Employment Losses from "No-AIDS" Baseline, Two AIDS fees for mandatory state pen- Scenarios, 2004-14 sion insurance); the forgone revenue to disability social 350,000 insurance fund (as unpaid premiums to the fund); and 300,000 AIDS Optimistic the forgone revenue to 250,000 AIDS Pessimistic unemployment social insur- ance fund (as unpaid levies 200,000 to this fund). 150,000 Using our estimates of reduc- 100,000 tion in employment, average 50,000 withholding rate and average monthly wages, we calculate 0 the resulting forgone revenue 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 64 to the state budget and special Figure A4-2. Share of Forgone Revenue due to AIDS by Category, funds due to HIV/AIDS (in opti- Pessimistic Epidemic Scenario, 2014 mistic and pessimistic epidemic scenarios). We also calculate the 100 difference between these scenarios to determine additional forgone rev- 90 enues to the budget and special 82.3% funds if the pessimistic epidemic 80 scenario is realized instead of the optimistic one (Table A4-3 on 70 page 66). Unemployment levy 60 Temporary disability levy Table A4-3 demonstrates that for- Pension levy 50 gone revenue will grow over time Percent Personal income tax as the decline in employment con- 40 tinues, with the amount in forgone taxes increasing 14-fold under the 30 optimistic scenario and more than 19-fold under pessimistic one with 20 respect to the HIV/AIDS epidemic. The structure of total forgone rev- 9.5% 10 enue is presented in Figure A4-2. 5.1% 3.2% 0 Estimated Additional (HIV/AIDS-Related) Source: Authors' calculations. Temporary Disability Payments from the Special Fund outlays from the special fund on temporary disability (Temporary Disability) benefits will grow. As part of the Ukrainian legislative package on The number of HIV-infected persons on temporary mandatory state social insurance, the Parliament disability benefit was estimated separately by age- passed the Law of Ukraine "On mandatory state gender group under pessimistic and optimistic epi- social insurance in the event of temporary disability demic scenarios. This allowed us to compare addi- and on the expenses associated with births and buri- tional expenditures associated with the realization of als" (18 January 2001, # 2240-III). The law defines the pessimistic scenario as opposed to the optimistic legal, organizational, and financial principles of one. We assumed that a person infected with HIV mandatory state social insurance of citizens in the receives his or her temporary disability payment event of temporary disability. In accordance with starting Year 5 from becoming infected. For instance, this law, financial support to those temporarily dis- the 2004 estimate of those who apply for a tempo- abled (including access to social services) is funded rary disability benefit for the first time is based on through the special fund (temporary disability). This the number of new infections that occurred in 1999. fund is financed through compulsory levies and in The results are presented in Figures A4-3 and A4-4 2003 disbursed UAH 1.1 billion in temporary disabili- on page 67. ty benefits. As a consequence of the HIV/AIDS epi- demic, an increasing number of employees infected According to these figures, the estimated number of with HIV will take some time off for medical exami- persons on disability benefit diverges in 2008, assum- nation and treatment, including hospital inpatient ing that 2004-08 payments are made to those who treatment as their disease progresses. As a result, 65 928 2014 728 216,935,429 25,031,017 13,440,453 8,343,672 344,317,372 39 21,332,524 13,242,976 154,871,179 127,381,893 14,697,911 7,892,071 4,899,304 2013 2004-14 192,613,598 22,224,646 11,933,566 7,408,215 277,302,374 31,996,428 17,180,543 10,665,476 102,964,796 84,688,776 9,771,782 5,246,977 3,257,261 2012 HIV/AIDS, 143,645,330 16,574,461 8,899,689 5,524,820 229,536,709 26,485,005 14,221,174 8,828,335 104,426,923 85,891,379 9,910,544 5,321,486 3,303,315 by Caused 2011 126,767,179 14,629,982 7,853,986 4,875,661 193,928,468 22,376,362 12,015,031 7,458,787 81,654,840 67,161,289 7,749,380 4,161,044 2,583,127, 2010 Employment 102,772,447 11,858,359 6,367,369 3,952,786 147,348,389 17,001,737 9,129,116 5,667,246 54,195,526 44,575,942 5,143,378 2,761,747 1,714,459 in 2009 80,133,264 9,246,146 4,964,736 3,082,049 116,097,889 13,395,910 7,192,960 4,465,303, 43,725,868 35,964,625 4,149,764 2,228,224 1,383,255 Reduction a to 2008 due 59,659,891 6,883,834 3,696,288 2,294,611 84,900,074 9,796,162 5,260,068 3,265,387 30,687,068 25,240,183 2,912,329 1,563,781 970,776 Budgets 2007 42,104,805 4,858,247 2,608,645 1,619,416, 58,292,315 6,726,036 3,611,558 2,240,012 19,680,810 16,187,510 1,867,790 1,002,913 622,597 Funds 2006 Special 34,400,899 3,969,334 2,131,342 1,323,111 42,659,161 4,922,211 2,642,991 1,640,737 10,040,413 8,258,262 952,876 511,649 317,625 and 2005 State 26,957,019 3,110,425 1,670,149 1,036,808 31,271,121 3,608,206 1,937,433 1,202,735 5,245,094 4,314,102 497,781 267,284 165,927 the to 2004 81,151 15,469,349 1,784,925 1,024,566 594,975 17,579,272 2,028,377 1,164,310, 676,126 2,574,271 2,109,923 243,453 139,744 Revenue tax tax tax levy levy levy Forgone income levy calculations. scenario scenario income levy income levy pension insurance insurance pension insurance insurance between pension insurance insurance levy AIDS) disability)y levy disability)y levy disability)y A4-3. personal state social social AIDS) personal state social social personal state social social Authors' yvnia ce: hr Optimistic (with Pessimistic (with Difference scenarios ableT In 1. Unpaid Unpaid insurance Unpaid (temporar Unpaid (unemployment) 2. Unpaid Unpaid insurance Unpaid (temporar Unpaid (unemployment) 3. Unpaid Unpaid insurance Unpaid (temporar Unpaid (unemployment) Sour 66 contracted the infection in Figure A4-3. Estimated New Male Recipients of Temporary Disability Payments 1999-2003 in the medium epi- due to HIV, 2004-14 demic scenario. Overall, 721,100 HIV-infected persons 50,000 will receive temporary disabil- AIDS Optimistic ity benefit over 2004-14 in the 45,000 AIDS Pessimistic pessimistic scenario and 524,000 in the optimistic one. 40,000 To evaluate the monetary 35,000 value of the additional disabil- ity payments, we made several assumptions: 30,000 1) We assume that all those 25,000 infected with HIV receive money from the social 20,000 insurance special funds. This assumption provides an upper estimate of the 15,000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 fund's potential additional expenditure associated Source: Authors' calculations. with HIV/AIDS; 2) In line with historic data, the average number of Figure A4-4. Estimated New Female Recipients of Temporary Disability Payments sick days per person per due to HIV, 2004-14 year is 3.9; 3) Based on the same data 50,000 for 2003, an average tem- AIDS Optimistic porary disability benefit is 45,000 AIDS Pessimistic UAH 22.41, which is assumed to be applicable 40,000 to those requiring this benefit because of their 35,000 HIV status; 30,000 4) Average daily wages and salaries are assumed to be 25,000 UAH 23.75, based on the fund's 2003 annual report. 20,000 This is the basis for calcu- lating the temporary dis- 15,000 ability benefit amount in lieu of actual data on 10,000 average monthly earnings 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 of those infected with HIV. Source: Authors' calculations. 67 Figure A4-5. Outlays in Temporary Disability Associated with HIV/AIDS, 2004-14 14,000,000 AIDS Optimistic 12,000,000 AIDS Pessimistic 10,000,000 8,000,000 6,000,000 4,000,000 2,000,000 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. Future disability benefits up to 2014 are indexed in Estimated Additional Payments from accordance with the projected average monthly the Special Fund (Permanent wages and salaries for the same period. Total Disability) to Those Disabled due amount of temporary disability benefit is calculated to AIDS as a product of the total number of HIV-infected per- According to Ukrainian law, disabled persons receive sons requiring a benefit in a relevant year, adjusted the following payments and allowances: for an unemployment, times average daily wages and Disability pensions from the Pension Fund of salaries adjusted for inflation, times the number of Ukraine; sick days per patient and the index of substitution Financial support from the special fund (perma- for temporary disability. The results are presented in nent disability) and allowances for additional Table A4-4. They include an estimated additional social services; expenditure on temporary disability payments through the special fund due to HIV/AIDS, total and State social assistance to disabled children and by gender, under two epidemic scenarios. The differ- persons disabled from childhood provided from ence between the pessimistic and optimistic scenar- the State budget. ios demonstrates savings to the special fund (tempo- rary disability) if the pessimistic scenario is avoided. Permanent Disability Pensions To calculate additional permanent disability pen- Based on the methodology discussed above, tempo- sions, we used the projected number of new AIDS rary disability outlays under pessimistic and opti- cases as the number of new recipients of the perma- mistic scenarios diverge considerably from 2009 nent disability payments and allowances. Total esti- (Figure A4-5). mated additional expenditure on permanent 68 2014 64,210 35,610 28,610 179.57 37,700 20,120 17,510 179.57 11,532,042 6,769,822 4,762,220 11,532,042 6,394,519 5,137,523 6,769,822 3,625,536 3,144,286 2004-14 2013 64,280 34,570 29,710 171.84 36,950 18,650 18,300 171.84 11,045,756 6,349,419 4,696,336 11,045,756 5,940,445 5,105,311 6,349,419 3,204,781 3,144,638 Disability), y 2012 53,730 30,580 23,160 164.44 33,800 18275 15,525 164.44 8,836,940 5,558,031 3,278,909 8,836,940 5,028,538 3,808,402 5,558,031 3,005,119 2,552,912 emporar (T 2011 62,110 33,900 28,220 157.36 30,980 15,570 15,410 157.36 9,775,011 4,874,917 4,900,094 9,775,011 5,334,399 4,440,612 4,874,917 2,450,047 2,424,870 Fund Special 2010 71,600 41,490 30,110 150.58 27,840 15,315 12,525 150.58 10,781,667 4,192,201 6,589,466 10,781,667 6,247,645 4,534,022 4,192,201 2,306,162 1,886,039 the from 2009 74,520 44,810 29,710 143.82 26,530 16,480 10,050 143.82 10,717,548 3,815,574 6,901,974 10,717,548 6,444,623 4,272,925 3,815,574 2,370,172 1,445,402 Payments 2008 126,016 77,810 35,500 42,310 136.97 76,890 33,460 43,430 136.97 10,657,932 10,531,916 10,657,932 4,862,570 5,765,362 10,531,916 4,583,144 5,948,773 Disability y 2007 9,829,175 9,642,809 186,366 75,420 41,670 33,750 130.33 73,990 40,160 33,830 130.33 9,829,175 5,430,678 4,398,497 9,642,809 5,233,886 4,408,923 emporarT 2006 9,145,136 9,011,341 133,795 73,820 43,220 30,600 123.88 72,740 41,820 30,920 123.88 9,145,136 5,354,277 3,790,858 9,011,341 5,180,840 3,830,501 2005 59,220 37,560 21,650 117.65 60,840 37,330 23,510 117.65 6,965,999 7,157,767 6,965,999 4,418,808 2,547,102 7,157,767 4,391,838 2,765,929 (HIV/AIDS-Related) 2004 44,380 26,510 17,870 101.25 45,740 26,490 19,250 101.25 4,493,262 4,630,955 4,493,262 2,684,010 1,809,252 4,630,955 2,681,985 1,948,970 Additional y to y to (pI) (oI) of of due optimistic (persons) benefit due (persons) benefit calculations. Estimated SCENARIO (pI) temporar SCENARIO (oI) temporar scenario scenario between and number of benefit total annual number of benefit total annual A4-4. Authors' additional yvnia person person ce: ableT hr In Pessimistic Optimistic Difference pessimistic scenarios PESSIMISTIC otal,additional,T expenditures, Males Females Projected recipients disability HIV/AIDS, Males Females verageA per OPTIMISTIC otalT expenditures Males Females Projected recipients disability HIV/AIDS, Males Females verageA per Sour 69 Figure A4-6. Estimated Additional Expenditure on Permanent Disability Pensions due to AIDS, 2004-14 In hryvnia 250,000,000 225,000,000 AIDS Optimistic 200,000,000 AIDS Pessimistic 175,000,000 150,000,000 125,000,000 100,000,000 75,000,000 50,000,000 25,000,000 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. Figure A4-7. Estimated Number Disabled due to AIDS, by Gender, 2004-14 40,000 Males 35,000 Females 30,000 25,000 20,000 15,000 10,000 5,000 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 70 disability pensions due to Figure A4-8. Estimated Additional Financial Assistance and Allowances Provided AIDS is presented in Figure by the Special Fund (Permanent Disability) to Those Disabled due A4-6. to AIDS, 2004-14 In hryvnia To construct this estimate, we used the projected new AIDS 40,000,000 cases under two epidemic sce- narios and the actual average 35,000,000 disability pension adjusted for projected inflation. Figure AIDS Optimistic 30,000,000 A4-7 presents the estimated AIDS Pessimistic number disabled as a result of 25,000,000 AIDS in pessimistic scenario, by gender. 20,000,000 We estimate that in 2005 addi- 15,000,000 tional disability pension pay- ments due to AIDS will total UAH 57.8 million under the 10,000,000 pessimistic scenario and 35.3 million under the optimistic 5,000,000 one. In the former, the disabili- ty pension outlays will 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 increase to UAH 199.9 million in 2014, corresponding to an 15 percent per annum growth Source: Authors' calculations. rate. In the latter, these pay- ments will reach UAH 109.2 million, a 13 percent per annum increase over 2005- provision of medical and social services to a person 14 (Figure A4-5). Results of calculation are presented disabled is calculated and adjusted for future in Table A4-5 on page 73. inflation. Financial Support from the Special Fund We estimate that in the pessimistic scenario the (Permanent Disability) and Allowances for total additional expenditure from the special fund Additional Social Services (permanent disability) due to AIDS amounts to UAH Pursuant to Ukraine's law "On basic principles of 5.8 million in 2005, UAH 17.3 million in 2010, and social protection of disabled persons in Ukraine," UAH 35.5 million in 2014, representing a 3-7 percent disabled persons have the right to receive financial increase in total fund expenditure. Figure A4-8 and support and allowances for health and social servic- Table A4-5 present the results under both scenarios. es from the special fund (permanent disability). To estimate additional outlays by the fund due to the Additional Assistance to HIV/AIDS epidemic, we used the projected new AIDS Disabled Children and Persons cases occurring among those aged 15 and above sep- Disabled Since Childhood Provided arately for males and females. This was taken as an from the State Budget estimate of the number of recipients of financial sup- To calculate the amount of additional assistance port and allowances from the special fund (perma- from the state budget to children disabled due to nent disability). An annual allowance covering AIDS, we assumed that all children who developed 71 Figure A4-9. Projected Number of Disabled Children (0-16) Receiving Assistance AIDS when they are 16 or due to Their AIDS Condition, 2004-14 younger, as well as adults who were infected with HIV in their 3500 childhood become disabled. In accordance with the Ukraine AIDS Optimistic law "On state social aid to per- 3000 AIDS Pessimistic sons disabled since childhood and disabled children," these 2500 categories of population have the right to receive state assis- 2000 tance provided from the state budget. 1500 To estimate the number of recipients in this category, we 1000 used the projected number of AIDS cases in the 0-16 age 500 group and added cases in the 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 17-21 age group. This selection is based on the assumption that Source: Authors' calculations. progression from infection with HIV at age 16 and younger to AIDS takes five years on aver- age. Figures A4-9 and A4-10 Figure A4-10. Projected Number of Persons Disabled Since Childhood (AIDS Patients Aged 17-21) Receiving Assistance under Optimistic and illustrate our estimate of the Pessimistic Scenarios, 2004-14 number of disabled children and those disabled since child- hood under both epidemic sce- 2000 narios. Total additional assis- tance to disabled children and 1750 those disabled since childhood as a result of AIDS was calcu- 1500 lated and inflated using CPI. Results are presented in AIDS Optimistic Table A4-5. 1250 AIDS Pessimistic 1000 750 500 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Authors' calculations. 72 74) page 2014 67,260 36,530 30,730 2,972 36,740 19,240 17,500 2,972 on 243,718,083 132,199,118 199,909,417 108,574,056 91,335,361 109,198,216 57,184,912 52,013,304 (continued 2013 64,950 35,500 29,450 2,844 38,180 20,180 18,000 2,844 222,333,806 129,754,108 184,730,776 100,969,093 83,761,684 108,591,548 57,395,952 51,195,596 2004-2014 2012 62,400 34,400 28,000 2,722 38,760 20,800 17,950 2,722 201,945,767 124,491,751 169,835,471 93,627,247 76,208,224 105,466,739 56,611,824 48,854,915 HIV/AIDS, to 2011 due 59,310 33,000 26,310 2,605 38,190 20,760 17,430 2,605 181,677,295 116,201,415 154,474,018 85,949,125 68,524,893 99,466,578 54,069,813 45,396,765 Funds 2010 55,710 31,350 24,360 2,492 36,600 20,130 16,480 2,492 161,639,414 105,609,184 138,849,521 78,135,568 60,713,953 91,245,395 50,171,259 41,074,136 Special and 2009 51,660 29,610 22,050 2,380 33,990 18,890 15,110 2,380 143,539,156 93,979,824 122,975,600 70,486,014 52,489,585 80,936,322 44,967,268 35,969,054 Budget State 2008 46,970 27,490 19,480 2,267 30,420 17,190 13,230 2,267 124,642,174 80,323,018 106,486,808 62,323,235 44,163,573 68,965,908 38,971,859 29,994,049 from Fund 2007 41,690 25,000 16,700 2,157 26,410 15,170 11,240 2,157 Payments 105,658,679 66,616,407 89,951,537 53,927,780 36,023,757 56,969,307 32,723,377 24,245,930 Pension the 2006 35,910 22,110 13,800 2,050 22,580 13,040 9,530 2,050 Disability total 86,844,409 54,486,002 from 73,632,949 45,336,244 28,296,706 46,279,467 26,738,336 19,541,131 AIDS, AIDS, to 2005 to 29,690 18,820 10,870 1,947 18,110 10,650 7,460 1,947 Permanent 68,509,962 41,782,943 due 57,814,746 36,647,811 21,166,934 35,265,242 20,738,533 14,526,709 due 2004 pensions 23,760 15,460 8,300 1,791 13,700 8,230 5,470 1,791 Additional 50,917,910 29,379,285 42,564,282 27,695,446 14,868,836 24,542,536 14,743,436 9,799,100 Expenditure disability pa pa Estimated scenario scenario Additional additional recipients, recipients, of payment of payment A4-5. (pI) (persons) (oI) (persons) ableT hyvnia In Estimated PESSIMISTIC (pI+pII+pIII) OPTIMISTIC (oI+oII+oIII) INCLUDING: Estimated PESSIMISTIC otalT Males Females Number total Males Females verageA OPTIMISTIC otalT Males Females Number total Males Females verageA 73 2014 551 551 64,420 34,910 29,510 35,480 18,490 16,970 35,480,773 19,227,473 16,253,301 19,530,398 10,183,786 9,346,612 8,327,893 5,241,768 2013 482 482 62,310 34,000 28,330 36,840 19,440 17,430 2014 30,047,236 16,390,278 13,656,958 17,773,810 9,371,383 8,402,428 7,555,793 4,673,155 2004-2014 2012 422 422 59,940 32,970 26,950 37,380 20,030 17,350 2013 25,282,165 13,911,098 11,371,067 15,771,818 8,451,298 7,320,520 6,828,132 4,165,930 (UAH) HIV/AIDS, DS AI to 2011 369 369 57,000 31,700 25,330 36,800 19,990 16,800 to 2012 21,061,091 11,706,761 9,354,330 13,586,490 7,382,276 6,204,214 6,142,186 3,707,424 due due Funds 2010 323 323 53,610 30,160 23,450 35,210 19,370 15,840 2011 17,328,395 9,748,637 7,579,759 11,380,951 6,260,978 5,119,974 childhood 5,461,497 3,224,794 Special since and 2009 312 312 49,750 28,550 21,220 32,640 18,200 14,470 2010 15,543,178 8,916,169 6,627,009 10,202,846 5,683,863 4,518,983 5,020,378 2,849,616 disabled Budget 2008 301 301 persons 45,260 26,540 18,720 29,160 16,530 12,650 2009 State 13,630,381 7,992,716 5,637,665 8,787,771 4,978,131 3,809,640 and 4,524,985 2,442,418 from children AIDS 2007 290 290 to 40,190 24,170 16,030 25,300 14,580 10,710 2008 11,663,319 7,012,498 4,650,821 7,337,446 4,230,129 3,107,317 4,043,823 2,052,624 Payments due disabled to Funds 2006 279 34,600 21,370 13,210 21,550 12,520 9,040 279 2007 9,656,184 5,967,399 3,688,785 6,020,455 3,496,108 2,524,346 3,555,276 1,727,174 budget Disability Special the State 2005 269 28,590 18,210 10,400 17,240 10,200 7,040 269 2006 7,681,836 4,889,418 2,792,419 4,628,971 2,738,718 1,890,253 from from 3,013,380 1,442,448 Permanent assistance 2004 22,800 14,930 7,890 256 12,940 7,830 5,110 256 assistance 2005 5,833,690 3,816,696 2,016,995 3,307,973 2,001,656 1,306,317 2,519,937 1,179,091 Additional financial financial 73) pa pa to page SCENARIO SCENARIO SCENARIO recipients, recipients, (pIII)- additional additional from of payment of payment children A4-5.Estimated (pII) (persons) (oII) (persons) (UAH) assistance ableT PESSIMISTIC otalT Males Females Number total Males Females verageA otalT Males Females Number total Males Females verageA Estimated 2004 PESSIMISTIC otalT otalT disabled (continued Estimated OPTIMISTIC 74 3,044 1,998 1,722 1,545 1,418 666 1,722 1,545 3,086,125 3,470,504 2,441,796 1,028,708 2,824 1,944 1,655 1,483 656 2,882,638 3,388,750 2,416,008 972,742 1,460 1,655 1,483 2004-2014 2,638 1,880 1,579 1,416 658 2,662,202 3,253,195 2,321,424 931,771 1,470 1,579 1,416 HIV/AIDS, to 2,452 1,796 1,512 1,356 694 2,434,762 3,148,347 2,207,520 940,827 1,460 1,512 1,356 due Funds 2,232 1,726 1,445 1,296 674 2,236,703 2,982,837 2,109,408 873,429 1,460 1,445 1,296 Special and 2,056 1,748 1,386 1,242 716 1,242 2,170,762 2,840,656 1,951,488 889,168 1,408 1,386 Budget 1,852 1,762 1,319 1,182 1,294 730 1,319 1,182 State 2,082,567 2,569,339 1,706,527 862,812 from 1,640 1,774 1,252 1,122 1,162 762 1,252 1,122 1,991,199 2,309,655 1,454,359 855,295 Payments 1,448 1,710 1,193 1,069 1,096 822 1,193 1,069 1,828,101 2,186,080 1,307,309 878,771 Disability 1,272 1,546 1,134 1,016 972 774 1,134 1,016 1,570,932 1,888,730 1,102,248 786,482 Permanent 1,132 1,436 1,042 934 842 698 1,042 934 1,340,846 1,528,775 877,027 651,748 year Additional per or to pa: to pa: 73) to since since since due childhood AIDS, to to due childhood AIDS, annum, calculations. page disabled (persons) to SCENARIO persons (oIII)- disabled (persons) to persons per is from disabled of (0-16) of since due assistance children disabled children disabled of (0-16) of since due assistance children disabled A4-5.Estimated total total Authors' assistance childhood (persons) (UAH) assistance assistance (persons) "Pa" ce: ableT otalT persons since Number children AIDS, Number disabled (17-21) total verageA Disabled Persons childhood OPTIMISTIC otalT otalT disabled otalT persons childhood Number children AIDS, Number disabled (17-21) total verageA Disabled Persons childhood (continued Note: Sour 75 ANNEX 5 Growth Model: Methodology and Assumptions T his simple aggregate, two-factor, closed-econo- ment as well as household consumption becomes my macroeconomic model is analogous in its the residual of aggregate output: structure to the World Bank model designed private public It = It + It for the Russian Federation by Ruehl, Pokrovsky and Vinogradov (2002). The major distinction is that It = St unlike the Russian model, the Ukrainian model does Yt = Ct + St not generate the population and epidemic dynamics within the model. Demographic and epidemiological The tax rate can be defined by the user. As a per- forecasts reported in Chapter 3 and the labor force centage of output, private investments are defined as forecasts of Chapter 4 are used as inputs into the fol- follows: lowing macroeconomic model. Number of ARV ther- . (1­ ) Yt. private private It = st apy recipients and total expenditure on AIDS hospi- talization and treatment are also determined outside As a percentage of output, public investments are the model: these series are outputs from the AIM defined as model and are exogenous to the macroeconomic public public It = s . max(0,CBSt) model. Here s public denotes the share of public investment, Model Description and CBSt is current budget surplus. Current budget surplus is defined as the tax revenues, .Yt , net of The output Yt is a function of labor Lt and capital Kt debt payment and minimum required public expendi- with the Cobb-Douglas production function: tures (MPEt) (all user-defined parameters), as well as the cost of antiretroviral therapy (defined as unit Yt = At . Kt . Lt , cost per annum, ARV_Cost, times the number At is the total factor productivity (TFP). The value of receiving ARV therapy, ARV_Recipients) and AIDS capital share is assumed at = 0.3, and the labor treatment (AIDS_Care, projected from AIM accord- share = 0.7. The TFP is assumed to be growing ing to two epidemic and three cost scenarios). The from 1 percent to 5 percent over 2004-14. All these budget level, B, is parameters are open to manipulation by users as the Bt = .Yt . public ­ It ­ (1 + i) Dt­1 ­ MPEt ­ ARV_ model is coded in an Excel spreadsheet. Cost*ARV _Recipientst ­ AIDS_Caret The growth rate of capital is by gross investment net Current budget deficit results in a debt D, the debt of depreciation: service is included in the next year budget. Kt = Kt­1 . (1 ­ ) + lt­1 where the depreciation rate is parameterized as Model Application = 0.05 but can be changed by the user. We applied the model under three demographic Investment consists of public (government) and pri- scenarios (benchmark "no-AIDS" and two epidemic vate investment. Investment equals savings; govern- scenarios: "AIDS Optimistic" and "AIDS Pessimistic") 76 and three cost scenarios (ARV Low- Table A5-1. Simple Growth Model Parameters for Various Scenarios HOSP(italization) Low, ARV High- In hryvnia HOSP Low, and ARV High-HOSP High). Model parameters and impli- MODEL PARAMETERS No-AIDS Low High cations are presented in Tables A5- HIV treatment costs scenarios 1 and A5-2 on page 78. Annual cost of ARV medication per HIV+ receipient 0 1500 7500 Economy Recall that the epidemic scenarios used in the modelling of Chapter 3 Public investment as % of government revenue 8.1 8.1 8.1 are based on the following assump- Private saving (investment) as % of after-tax income 30 30 30 tions about availability of ARV ther- Domestic interest rate, %, i 5 5 5 apy: Household tax rate, %, tau 26 26 26 The "AIDS medium" scenario Miscellaneous assumes that in 2004, 1 percent Budgetary costs of HIV prevention programs, billion UAH 0.01 0.01 0.01 of those in need of ARV therapy Minimum public expenditure as % of GDP, MPE 10 10 10 have access to it, with the pro- portion receiving it rising to 10 Depreciation rate, delta, % 5 5 5 percent in 2008 and remaining at Capital share, alpha (CRS function) 0.3 0.3 0.3 that level thereafter; Public debt, % DP 29 29 29 The "AIDS optimistic" scenario Source: State Statistics Committee of Ukraine, Ministry of Health, WHO (Kiev office), estimates of assumes that in 2004, 1 percent the Institute for Economic Forecasting of the Academy of Sciences of Ukraine, and authors' of those in need of ARV therapy calculations. have access to it, with the pro- portion receiving it rising to 30 percent in 2010 and then to 50 percent in 2014; The "AIDS pessimistic" scenario assumes that in also assumed that starting from 0 percent in 2004, 2004, 1 percent of those in need of ARV therapy 30 percent of AIDS cases are hospitalized by have access to it, with the proportion receiving it 2014, at an annual cost of UAH 1,500 per case; rising to 5 percent in 2005 and remaining at that Scenario B. ARV-High, Hospitalization- level thereafter. Low. Assumes that the cost of an ARV treatment The optimistic scenario also assumes lower MTCT per year is UAH 7,500 in constant prices through- rates starting in 2004 than the pessimistic scenario. out the modelling horizon; 50 percent of AIDS We considered three plausible cost scenarios, reflect- cases are hospitalised by 2014, at an annual cost ing the degree of uncertainty with respect to the cost of UAH 7,500 per case; of treatment and access to treatment by those who Scenario C. ARV-High, Hospitalization- require it: High. Assumes that the cost of an ARV treat- Scenario A. ARV-low, Hospitalization- ment per year is UAH 7,500 in constant prices Low: Assumes availability of low-price ARV throughout the modelling horizon; 100 percent of treatment (UAH 1,500 per year in constant AIDS cases are hospitalised by 2014, at an annual prices19) throughout the modelling horizon. It is cost of UAH 7,500 per case. 19Data supplied by WHO Kiev Office, based on the Clinton Foundation best negotiated price for generics. Exchange rate 1US$ = 5.3 UAH. 77 Table A5-2. Simple Growth Model Output, AIDS/Cost Scenario Analysis A. ARV low -H low B. ARV high -H low C. ARV high -H high 2004 2014 Diff(Base) 2004 2014 Diff(Base) 2004 2014 Diff(Base) Output (billion UAH) No AIDS 345.94 359.64 0.00 345.94 359.64 0.00 345.94 359.64 0.00 AIDS optimistic 345.94 357.09 -2.54 345.94 357.08 -2.55 345.94 357.08 -2.56 AIDS pessimistic 345.94 355.20 -4.44 345.94 355.19 -4.45 345.94 355.19 -4.45 Output (Index, 2004=100) No AIDS 100.00 103.96 0.00 100.00 103.96 0.00 100.00 103.96 0.00 AIDS optimistic 100.00 103.22 -0.74 100.00 103.22 -0.74 100.00 103.22 -0.74 AIDS pessimistic 100.00 102.68 -1.28 100.00 102.67 -1.29 100.00 102.67 -1.29 Output per capita (Index, 2004=100) No AIDS 100.00 110.80 0.01 100.00 110.80 0.00 100.00 110.80 0.00 AIDS optimistic 100.06 110.79 -0.01 100.06 110.79 -0.01 100.06 110.79 -0.01 AIDS pessimistic 100.13 110.00 -0.80 100.13 110.78 -0.02 100.13 110.78 -0.02 GDP growth rate, % Average No AIDS 12.10% 0.10% 1.45% 12.10% 0.10% 1.45% 12.10% 0.10% 1.45% AIDS optimistic 12.10% 0.02% 1.39% 12.10% 0.02% 1.39% 12.10% 0.02% 1.39% AIDS pessimistic 12.10% -0.13% 1.34% 12.10% -0.13% 1.34% 12.10% -0.13% 1.34% GDP per capita growth rate, % Average No AIDS 10.00% 0.64% 1.85% 10.00% 0.64% 1.85% 10.00% 0.64% 1.85% AIDS optimistic 10.00% 0.66% 1.84% 10.00% 0.66% 1.84% 10.00% 0.66% 1.84% AIDS pessimistic 10.00% -0.13% 1.77% 10.00% 0.58% 1.83% 10.00% 0.58% 1.83% Capital stock (billion UAH) No AIDS 1,032.04 1,290.01 0.00 1,032.04 1,290.01 0.00 1,032.04 1,290.01 0.00 AIDS optimistic 1,032.04 1,288.09 -1.92 1,032.04 1,287.97 -2.04 1,032.04 1,287.93 -2.08 AIDS pessimistic 1,032.04 1,286.77 -3.24 1,032.04 1,286.69 -3.32 1,032.04 1,286.62 -3.39 Capital stock (Index, 2004=100) No AIDS 100.00 125.00 0.00 100.00 125.00 0.00 100.00 125.00 0.00 AIDS optimistic 100.00 124.81 -0.19 100.00 124.80 -0.20 100.00 124.80 -0.20 AIDS pessimistic 100.00 124.68 -0.31 100.00 124.67 -0.32 100.00 124.67 -0.33 Effective labor supply (million persons) No AIDS 20.44 18.31 0.00 20.44 18.31 0.00 20.44 18.31 0.00 AIDS optimistic 20.42 18.12 -0.20 20.42 18.12 -0.20 20.42 18.12 -0.20 AIDS pessimistic 20.43 18.00 -0.31 20.43 18.00 -0.31 20.43 18.00 -0.31 (continued on page 79) 78 Table A5-2. Simple Growth Model Output, AIDS/Cost Scenario Analysis (continued from page 78) A. ARV low -H low B. ARV high -H low C. ARV high -H high 2004 2014 Diff(Base) 2004 2014 Diff(Base) 2004 2014 Diff(Base) Effective labor supply (Index, 2004=100) No AIDS 100.00 89.60 0.00 100.00 89.60 0.00 100.00 89.60 0.00 AIDS optimistic 99.88 88.64 -0.95 99.88 88.64 -0.95 99.88 88.64 -0.95 AIDS pessimistic 99.96 88.08 -1.52 99.96 88.08 -1.52 99.96 88.08 -1.52 Investment (billion UAH) No AIDS 82.29 84.50 0.00 82.29 84.50 0.00 82.29 84.50 0.00 AIDS optimistic 82.29 83.90 -0.60 82.29 83.86 -0.64 82.29 83.85 -0.65 AIDS pessimistic 82.29 83.45 -1.05 82.29 83.43 -1.07 82.29 83.41 -1.09 Investment (Index, 2004=100) No AIDS 100.00 102.69 0.00 100.00 102.69 0.00 100.00 102.69 0.00 AIDS optimistic 100.00 101.96 -0.74 100.00 101.91 -0.78 100.00 101.90 -0.79 AIDS pessimistic 100.00 101.42 -1.27 100.00 101.39 -1.30 100.00 101.37 -1.33 Cumulative HIV+ (thousands) No AIDS 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AIDS optimistic 447.78 478.50 478.50 447.78 478.50 478.50 447.78 478.50 478.50 AIDS pessimistic 491.16 820.42 820.42 491.16 820.42 820.42 491.16 820.42 820.42 Public expenditure on ARV (million UAH) No AIDS 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AIDS optimistic 0.41 70.61 70.61 2.03 353.03 353.03 2.03 353.03 353.03 AIDS pessimistic 0.72 10.38 10.38 3.60 51.90 51.90 3.60 51.90 51.90 Public expenditure on AIDS care (million UAH) No AIDS 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AIDS optimistic 0.00 16.55 16.55 0.00 137.88 137.88 0.00 275.76 275.76 AIDS pessimistic 0.00 30.27 30.27 0.00 252.23 252.23 0.00 504.47 504.47 Total direct expenditure on HIV/AIDS (million UAH) No AIDS 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AIDS Optimistic 0.41 87.16 87.16 2.03 490.91 490.91 2.03 628.79 628.79 AIDS Pessimistic 0.72 40.65 40.65 3.60 304.13 304.13 3.60 556.37 556.37 1 Data supplied by WHO Kiev Office, based on the Clinton Foundation best negotiated price for generics. Exchange rate 1US$ = 5.3 UAH. Source: Authors' calculations. 79 ANNEX 6 Measuring the Burden of HIV/AIDS in Ukraine: Methodology H artunian, Smart, and Thompson (1981) distin- and current cost estimates, so it is not suitable for guish between two major methods of measur- evaluation of prevention programs or programs ing cost-of-illness: prevalence based and inci- aimed at slowing down illness progression. It is even dence based. less useful for conditions whose prevalence changes significantly with time (e.g., cardiovascular disease). The prevalence approach to measuring economic The incidence approach allows estimation of avoid- costs of a disease in a period of time (usually a year) able costs and hence economic comparison of inter- is based on 1) identifying all health care costs and ventions that alter future disease incidence. productivity losses accruing to the sick during the year (the stock of people with the condition, or The incidence-based methodology for the economic prevalence), and 2) calculating expected future evaluation of the HIV intervention programs used in income lost by those who die from the disease in this report follows that methodology of the Global this year. Burden of Disease (GBD) study (see Murray, Lopez, and WHO [1994]; Murray and Acharya [1997]; Murray The incidence approach focuses only on the new et al. [2000]; and Murray and Lopez [2000] for cases of the disorder diagnosed over a certain period details). (a year), i.e., on the incidence of the disease. The present value of total lifetime health care expendi- The burden of a disease, including HIV/AIDS, is ture, morbidity, and mortality costs are then calculat- measured in Disability Adjusted Life Years (DALYs). ed, based on assumptions about the survival proba- The DALY measure combines potential years lost to bilities and an appropriate discount factor. premature mortality (Years of Life Lost, or YLL), and the loss of healthy life due to a nonfatal condi- The drawback of the latter approach to estimating tion/disability (Years Lost to Disability, or YLD): economic burden of disease is the uncertainty asso- DALY = YLL + YLD. ciated with the future treatment paths: a significant medical technology breakthrough would require ex- To calculate the loss of healthy life years, YLD, the post re-estimation of the future health care costs and incidence of the disease is multiplied by the average survival probabilities. The prevalence approach duration of each stage of the disease and by a dis- involves projected future earnings that are also ability (severity) weight for this stage. The year of uncertain. In both approaches sensitivity analysis is perfect health has a disability weight of zero; disabil- needed to assess various scenarios within the ity weight of one represents an equivalent of death. bounds of realization of stochastic variables. Neither In calculating DALYs for the GBD, a standard life the prevalence or incidence approach is a dominate table is used for all countries. In studies such as methodology. Mathers, Vos, and Stephenson (1999) and DHS (1999), national or state life tables (cohort life The prevalence approach is usually used to control expectancies) are used to calculate YLL. The GBD health care costs and loss of productivity within a uses a 3 percent time discount rate to calculate dis- year. It uses retrospective dynamics of the disease counted DALYs. It also applies non-uniform age 80 Table A6-1. Duration of HIV/AIDS Stages and Disability Weights for YLD Calculation Duration Disability weight Asymptomatic AIDS Asymptomatic AIDS HIV HIV-disease AIDS terminal HIV HIV-disease AIDS terminal Age 0-4 5-9 10-14 15-15.5 group years years years years 0-4 3 3 1 0.3 0.20 0.31 0.56 0.93 5-14 4 4 4 0.4 0.20 0.31 0.56 0.93 15-24 4 4 4 0.4 0.20 0.31 0.56 0.93 25-34 4 4 4 0.4 0.20 0.31 0.56 0.93 35-44 4 4 4 0.4 0.20 0.31 0.56 0.93 45-54 4 4 4 0.4 0.20 0.31 0.56 0.93 55-64 4 4 4 0.4 0.20 0.31 0.56 0.93 65-74 4 4 4 0.4 0.20 0.31 0.56 0.93 75+ 4 4 4 0.4 0.20 0.31 0.56 0.93 Source: Adapted from DHS (1999) and Stouthard et al. (1997). Table A6-2. Calculation of the Years of Life Lost to weights that give more value to young Premature Mortality, YLLs Undiscounted and and mid-adult years while discounting Discounted at 3 Percent very young and older ages. The studies mentioned chose to avoid non-uniform YLL_9 (0,0) YLL_9 (3,0) age weights for discounting purposes, the approach undertaken in this study, Male Female Male Female too. 0-4 60.56 70.70 0-4 27.92 29.34 5-14 54.34 64.43 5-14 26.79 28.50 The severity weights for HIV/AIDS are 15-24 44.81 54.69 15-24 24.62 26.85 from the Dutch study (Stouthard et al. 25-34 35.95 45.10 25-34 21.97 24.70 1997) under the assumption that these are the best available weights in the 35-44 27.70 35.70 35-44 18.79 21.89 Ukrainian context. Duration of the stages 45-54 20.40 26.74 45-54 15.23 18.36 has been adjusted to match Ukrainian 55-64 14.24 18.50 55-64 11.57 14.16 epidemiological evidence. The burden of 65-74 9.18 11.26 65-74 8.01 9.53 HIV/AIDS in Ukraine is estimated using 75+ 4.65 5.38 75+ 4.32 4.95 ten-year age groups. Source: Authors' calculations. 81 83) 2014 1.89 0.55 0.79 0.00 0.01 1.32 0.13 0.11 0.56 0.44 0.54 page 341.92 60.23 13.72 46.51 774.02 22.93 35.18 77.81 17.71 27.16 60.08 on 2,626.42 1,712.05 (continued 2013 1.37 0.40 -0.08 0.00 0.00 1.06 0.10 0.08 0.39 0.32 0.39 304.91 52.94 10.79 42.15 799.95 19.48 30.02 66.17 15.51 23.90 52.68 2,717.33 1,763.12 2012 1.39 0.40 0.01 0.00 0.00 0.77 0.07 0.08 0.41 0.32 0.39 8.44 263.83 44.76 36.32 784.56 16.81 26.03 57.05 13.64 21.12 46.29 2,663.05 1,719.49 2011 0.96 0.28 -0.07 0.00 0.00 0.58 0.06 0.05 0.26 0.22 0.27 6.54 36.06 29.52 13.61 21.14 46.23 11.15 17.31 37.84 222.17 780.07 2,648.71 1,705.52 scenarios 2010 0.86 0.25 -0.05 0.00 0.00 0.40 0.04 0.05 0.23 0.20 0.24 5.03 27.38 22.35 14.08 21.71 47.72 11.49 17.72 38.95 178.52 573.66 pessimistic 1,943.91 1,260.80 minus 2009 0.93 0.27 0.04 0.00 0.00 0.19 0.02 0.06 0.27 0.22 0.26 3.82 19.28 15.46 15.14 23.21 51.17 12.14 18.61 41.03 149.33 830.38 376.67 1,273.39 2004­14 Optimistic 2008 0.40 0.12 -0.07 0.00 0.00 0.10 0.01 0.01 0.07 0.09 0.11 2.82 9.18 9.99 7.64 12.00 15.27 33.67 11.68 25.76 134.12 785.83 356.37 1,201.13 Scenario, 2007 0.58 0.17 0.01 0.00 0.00 -0.03 0.00 0.03 0.15 0.14 0.16 6.08 1.88 4.20 7.02 4.85 7.38 10.68 23.56 16.27 118.21 865.52 568.78 257.81 Pessimistic 2006 -0.26 -0.07 -0.19 0.00 0.00 0.03 0.00 -0.04 -0.18 -0.06 -0.07 1.83 1.02 0.81 1.38 2.12 4.70 0.61 0.94 2.08 108.09 863.32 389.74 1,321.86 versus 2005 0.13 0.04 -0.05 0.00 0.00 0.00 0.00 -0.01 -0.03 0.03 0.04 84.16 -1.28 0.35 -1.63 -0.66 -1.00 -2.24 -0.84 -1.28 -2.85 569.85 1,937.21 1,269.62 Optimistic 2004 0.00 0.00 -0.06 0.00 0.00 0.00 0.00 -0.02 -0.08 0.00 0.00 43.38 -0.32 0.00 -0.32 -0.16 -0.24 -0.51 -0.16 -0.24 -0.51 623.01 Analysis, 1,957.12 1,328.82 Y DAL low ,V UAH UAH on UAH UAH of 2004) (`000) AR UAH y capita index- on averted mil (0,0), (3,0), (3,0) Y(0,0), Y(3,0), Y(3,0) (`000) (`000) Y Y Y (Index, per averted DAL DAL DAL care, expenditure Summar per per per hospitalization, GDP GDP index- DAL DAL DAL supply supply capita index- UAH HIV+ averted averted discounted per per per expenditure AIDS UAH mil low y index- V V V per rate, rate, stock stock labor labor in A6-3. additional (0,0) (3,0) (3,0) AR AR AR add_exp add_exp add_exp ARV UAH Ys Ys Ys of of of UAH ableT A. Summar Output Output Output Growth Growth Capital Capital Effective Effective Investment Investment Cumulative Additional mil Savings otalT HIV/AIDS, DAL DAL DAL (`000) Cost ostC Cost discounted, otalT otalT otalT disc, 82 84) 1.89 0.55 0.01 0.00 0.00 1.28 0.12 0.11 0.56 0.43 0.52 page 68.77 875.75 676.24 492.52 446.02 2014 341.92 301.13 114.35 186.78 774.02 114.65 175.89 on 2,626.42 1,712.05 (continued 1.37 0.39 0.00 0.00 1.03 0.10 0.08 0.39 0.31 0.37 70.20 754.05 600.35 355.82 313.68 2013 -0.08 89.63 97.40 304.91 264.68 175.05 799.95 150.12 2,717.33 1,763.12 1.38 0.40 0.01 0.00 0.00 0.75 0.07 0.08 0.41 0.31 0.38 67.86 659.61 535.24 360.54 324.22 2012 69.80 84.04 263.83 223.80 154.00 784.56 130.16 2,663.05 1,719.49 0.96 0.28 0.00 0.00 0.57 0.05 0.05 0.26 0.21 0.26 67.09 537.49 440.01 248.91 219.39 2011 -0.07 53.85 68.07 222.17 180.30 126.45 780.07 105.72 scenarios 2,648.71 1,705.52 82) page 0.86 0.25 50.38 2010 -0.05 0.00 0.00 0.39 0.04 0.05 0.23 0.19 0.24 41.02 95.86 70.41 from 543.39 443.55 222.83 200.49 178.52 136.88 573.66 108.56 pessimistic 1,943.91 1,260.80 (continued minus 0.93 0.27 0.04 0.00 0.00 0.19 0.02 0.06 0.27 0.21 0.26 33.56 574.40 460.56 240.67 225.22 2009 96.38 30.92 65.46 75.68 149.33 830.38 376.67 116.06 1,273.39 2004­14 Optimistic 0.40 0.12 0.00 0.00 0.10 0.01 0.01 0.07 0.09 0.11 31.96 94.58 375.52 287.27 103.76 2008 -0.07 60.00 22.37 37.63 49.95 76.35 134.12 785.83 356.37 1,201.13 Scenario, 0.58 0.17 0.01 0.00 0.00 0.00 0.03 0.15 0.13 0.16 23.27 261.10 180.30 150.10 145.90 2007 -0.03 30.38 14.51 15.87 35.09 53.40 118.21 865.52 568.78 257.81 Pessimistic 0.00 0.00 0.03 0.00 9.15 7.33 1.82 6.92 34.79 52.60 23.28 -67.45 -68.26 2006 -0.26 -0.07 -0.19 -0.04 -0.18 -0.06 -0.07 10.60 108.09 863.32 389.74 1,321.86 versus 0.13 0.04 0.00 0.00 0.00 0.00 0.03 0.04 1.73 51.27 -24.87 -31.69 32.88 34.51 2005 -0.05 -0.01 -0.03 84.16 -6.38 -8.11 -3.29 -5.02 569.85 1,937.21 1,269.62 Optimistic 52.03 -6.05 -6.05 0.00 0.32 0.00 0.00 2004 -0.06 0.00 0.00 0.00 0.00 -0.02 -0.08 0.00 0.00 43.38 -1.58 0.00 -1.58 -0.80 -1.19 623.01 Analysis, 1,957.12 1,328.82 Y DAL UAH UAH low UAH ,V on UAH UAH of mil 2004) (`000) AR y `000 infec, infec, capita index- on averted mil (0,0), (3,0), (`000) (`000) Y Y avert (Index, per averted Summar averted, avert collected, per position care, expenditure DAL DAL per y hospitalization GDP GDP index- supply supply taxes capita index- HIV+ averted averted discounted per per cases V expenditure AIDS AR high, y index- V V per rate, rate, stock stock labor labor in AR AR A6-3. HIV of add_exp (0,0) (3,0) (3,0) budgetar UAH additional ARV Ys Ys Ys of of ableT New Cost otalT Additional Net B. Summar Output Output Output Growth Growth Capital Capital Effective Effective Investment Investment Cumulative Additional mil Savings otalT HIV/AIDS DAL DAL DAL (`000) Cost Cost 83 85) page 71.11 68.77 2014 1.89 0.55 0.01 0.00 0.00 1.31 0.13 0.11 0.56 0.44 0.53 72.42 389.04 109.09 241.31 491.67 304.90 341.92 301.13 228.71 on 4,289.47 2,660.57 2,626.42 (continued 64.42 99.28 70.20 2013 1.37 0.40 -0.08 0.00 0.00 1.05 0.10 0.08 0.39 0.31 0.38 85.87 330.87 218.82 355.18 180.14 304.91 264.68 178.81 3,900.42 2,579.58 2,717.33 57.83 89.56 67.86 2012 1.39 0.40 0.01 0.00 0.00 0.76 0.07 0.08 0.41 0.32 0.39 84.99 285.25 196.29 360.09 206.09 263.83 223.80 138.81 3,335.82 2,295.42 2,663.05 47.74 74.14 67.09 2011 0.96 0.28 -0.07 0.00 0.00 0.57 0.06 0.05 0.26 0.22 0.27 73.66 231.13 162.10 248.63 122.18 222.17 180.30 106.64 3,578.94 2,510.02 2,648.71 83) page 49.31 76.03 50.38 2010 0.86 0.25 -0.05 0.00 0.00 0.40 0.04 0.05 0.23 0.20 0.24 80.78 56.10 from 238.60 167.09 222.68 126.82 178.52 136.88 4,078.88 2,856.48 scenarios 1,943.91 (continued 51.40 78.83 33.56 2009 0.93 0.27 0.04 0.00 0.00 0.19 0.02 0.06 0.27 0.21 0.26 96.38 60.37 36.01 255.86 173.77 240.61 175.15 149.33 3,015.87 2,048.29 pessimistic 1,273.39 2004­14 minus 31.33 47.89 31.96 66.12 2008 0.40 0.12 -0.07 0.00 0.00 0.11 0.01 0.01 0.07 0.09 0.11 60.00 43.09 16.91 168.36 105.59 103.75 134.12 2,578.76 1,617.31 1,201.13 Scenario, Optimistic 18.33 27.89 61.54 23.27 2007 0.58 0.17 0.01 0.00 0.00 -0.03 0.00 0.03 0.15 0.14 0.16 3.14 30.38 27.24 117.82 873.00 455.97 150.11 134.25 118.21 865.52 Pessimistic 1.38 2.11 4.67 23.48 34.79 35.50 2006 -0.26 -0.07 -0.19 0.00 0.00 0.03 0.00 -0.04 -0.18 -0.06 -0.07 9.15 -3.73 178.45 -67.43 -69.25 12.88 108.09 1,321.86 versus -4.18 -6.38 0.13 0.04 0.00 0.00 0.00 0.00 0.03 0.04 1.73 -11.19 -14.22 51.27 32.88 40.99 2005 -0.05 -0.01 -0.03 84.16 -6.38 -8.11 -122.53 -155.78 1,937.21 Optimistic -2.53 -0.80 -1.19 -2.53 0.00 1.58 52.03 2004 0.00 0.00 -0.06 0.00 0.00 0.00 0.00 -0.02 -0.08 0.00 0.00 0.00 -30.27 -30.27 43.38 -1.58 -1.58 Analysis, 1,957.12 Y DAL UAH UAH UAH UAH high UAH mil,V of mil 2004) (`000) AR y `000 infec, (3,0) Y(0,0), Y(3,0), Y(3,0) infec, capita index- on mil Y (`000) DAL DAL DAL avert (Index, per averted DAL Summar per per per averted, avert collected, per position care, expenditure hospitalization index- supply supply per UAH per y GDP GDP capita index- HIV+ averted V cases V taxes expenditure AIDS AR AR high, y index- per rate, rate, stock stock labor labor in A6-3. of add_exp add_exp add_exp UAH HIV of add_exp additional (0,0) budgetar ARV HIV/AIDS Ys ableT Cost discounted, otalT otalT otalT disc, New Cost otalT Additional Net C. Summar Output Output Output Growth Growth Capital Capital Effective Effective Investment Investment Cumulative Additional Savings otalT on DAL 84 2014 774.02 114.65 175.89 389.04 27.57 42.30 93.56 68.77 492.20 419.78 1,712.05 4,289.47 1,031.54 2013 799.95 97.40 150.12 330.87 31.60 48.70 70.20 107.34 355.57 269.70 1,763.12 3,900.42 1,265.36 2012 784.56 84.04 130.16 285.25 31.91 49.43 67.86 108.33 360.37 275.38 1,719.49 3,335.82 1,266.81 2011 68.07 780.07 105.72 231.13 27.81 43.19 94.43 67.09 248.82 175.16 1,705.52 3,578.94 1,462.15 84) page 2010 70.41 28.86 44.49 97.78 50.38 from 573.66 108.56 238.60 222.80 166.71 1,260.80 4,078.88 1,671.63 (continued 2009 75.68 28.27 43.36 95.59 830.38 376.67 116.06 255.86 33.56 240.68 204.67 3,015.87 1,126.71 2004­14 2008 49.95 76.35 14.08 21.52 47.45 31.96 86.87 785.83 356.37 168.36 726.78 103.78 2,578.76 Scenario, 2007 3.62 5.51 35.09 53.40 12.16 23.27 90.10 568.78 257.81 117.82 873.00 150.12 146.99 Pessimistic 2006 6.92 10.60 23.48 -2.82 -4.32 -9.57 34.79 863.32 389.74 178.45 -72.75 -67.43 -63.70 versus 2005 -3.29 -5.02 -4.18 -6.38 569.85 -11.19 -14.22 51.27 32.88 40.99 1,269.62 -122.53 -155.78 Optimistic 2004 -0.80 -1.19 -2.53 -0.80 -1.19 -2.53 0.00 1.58 52.03 623.01 -30.27 -30.27 Analysis, 1,328.82 Y DAL UAH UAH UAH UAH UAH UAH UAH of mil y averted `000 infec, (0,0), (3,0), (3,0) Y(0,0), Y(3,0), Y(3,0) infec, (`000) Y Y Y DAL DAL DAL avert DAL DAL DAL Summar per per per averted, avert collected, per position calculations. averted discounted per per per UAH per y V V V cases V taxes y A6-3. (3,0) (3,0) AR AR AR AR Authors' Ys Ys of of of add_exp add_exp add_exp UAH HIV of add_exp budgetar ce: ableT Summar DAL DAL (`000) Cost Cost Cost discounted, otalT otalT otalT disc, New Cost otalT Additional Net Sour 85 ANNEX 7 Macroeconometric Model: Methodology, Assumptions, and Estimation T he model is based on a system of equations nous variables such as interest rate, exchange rate, linking sectors (types) of economic activity in inflation, and sectoral deflators. Assumptions of bal- Ukraine. It includes aggregate national output anced budget and accounting identities close the (real sector), consumption, investments, employ- model. The block diagram representing theoretical ment, foreign trade, government budget, and mone- structure of the model and linkages between its com- tary sector. The reduced form model is econometri- ponents is in Figure A7-1. cally estimated. Using economic policy variables as controls, we solve the model for endogenous vari- Real sector model links though the accounting ables such as private and public consumption, gross identities estimated GDP measures. The model has investment, tax rates, export and import of goods four blocks. The aggregate supply block forms the and services, using the projected values of exoge- production function (the sum of domestic production Figure A7-1. Macroeconometric Model for Ukraine: Sectoral Structure Real Sector Sectoral and Blocks: World GDP, import/export deflator of world GDP deflators Aggregate supply; CPI, PPI. Aggregate demand; other deflators Foreign trade; Types of economic activity (sectors) Inflation rate, GDP, CPI, unemployment, Deposits, interest exchange rate, interest pensioners rate, export/import rate (National Bank), Basic money emission, Subsidies macroeconomic required reserves Household income indicators and consumption Interest rate Monetary sector GDP, tax arrears, GDP, Wages and salaries, budget deficit, Government budget employment private consumption tax rates by type of activity Source: Institute for Economic Forecasting. 86 and imports) as a function of gross capital forma- of state property, etc. These components represent tion, employment, and import of goods and services more than 80 percent of the budget revenue and are for final consumption. Estimated equations include included in the model as endogenous variables. time-lagged variables. Employment and investment equations are estimated within this block. The model of monetary sector is based on the equilibrium assumption (money supply equals money The block of aggregate demand represents the demand). Output variables include the projected expenditure measure of GDP and includes final con- money aggregate M2 and base money, given GDP sumption, government expenditure, gross fixed capi- and inflation rate. The model allows incorporation of tal formation, changes in inventories, and export of monetary policy instruments such as money emis- goods and services. sion, interest rate, and velocity of money. The foreign trade block defines export, import, and Sectoral macroeconometric models are implemented trade balance for goods and services in constant and using the E-Views-3 econometric modelling package. current prices. The export supply function depends Simulations were run based on the estimated model on domestic and world GDP and relative prices. coefficients, projected values of exogenous vari- Import demand function depends on real domestic ables, and a range of scenarios with respect to con- GDP, exchange rate, and terms of trade. trols. Direct relations and feedbacks between the sectoral models allow us to study direct and indirect The disaggregation block breaks down aggregate impact of change in exogenous variables/controls on GDP into sectoral components using the production endogenous variables in the model. Linkages approach. Sectors explicitly modelled are agricul- between the sectors are presented in Figure A7-1. ture, hunting, forestry, and fishing; mining; manufac- For example, changes in GDP in constant prices and turing; energy, gas and water industry; construction; interest rates generated in Real and Monetary blocks wholesale and retail trade and repair services; trans- enter the Household income and consumption. There port and telecommunications; education; health care is a feedback from the government budget sector and social services; etc. Sectoral models represent through total budget revenue and a resulting dispos- an econometric estimation of two-factor production able income that enters the household's saving and functions (sectoral capital and labor), with or with- consumption functions. Control variables in the out time trend, fitted to the 1994-2004 data. Monetary sector are tax rates and the taxation base. Changes in total budget revenues feed into the The household income and consumption Household income and consumption block, while model defines the disposable income, labor produc- changes in government deficit feed into the tivity, and unemployment levels. Average monthly Monetary block. As a feedback, real sector supplies real wages, pensions, and saving are evaluated in the GDP in current prices to the government budget this block. Saving is a function of disposable income, block. Other endogenous variables used in the interest rate, and time trend. Monetary block are unemployment rate, budget deficit, GDP in current prices, and exchange rate. The model of government budget describes budg- Estimated macroeconometric model based on et revenue, expenditure, and balance. Budget rev- 1986-2003 data is presented in Appendix A7 on enue is a sum of taxes applied to the corresponding page 92. tax base plus other administrative levies. Based on the current budget structure, the model includes the Model Assumptions following tax components: corporate profit tax; value added tax (VAT); excise duty; personal income Modelling was undertaken using what we estimate tax; land tax; deductions for geological survey and as the most plausible assumptions regarding exploration; stamp duty; proceeds from privatization future economic growth in Ukraine, future invest- 87 Table A7-1. Projected Macroeconomic Indicators until 2014 Years and periods ACTUAL FORECAST Indicators 1999 2000 2001 2002 2003 2004 2005-09 2010-14 Annual change in real indicators GDP, % -0.2 5.9 9.2 5.2 9.6 12.1 7.2 5.7 Population, `000,000 49.5 49.0 48.5 47.9 47.5 47.1 46.1 44.8 Aggregate supply Fixed assets, % 1.3 1.0 1.1 1.9 2.3 3.5 2.5 2.2 Labor force, `000,000 21.8 21.3 21.02 21.38 21.45 21.46 21.45 21.44 Final consumption, % -3.7 2.0 9.3 5.0 12.8 13.0 5.5 4.5 private -1.9 2.5 9.6 6.0 12.4 16.3 6.8 5.4 public -7.9 1.0 10.4 -0.1 14.8 4.7 1.6 1.7 Gross fixed capital formation, % 0.1 12.4 6.2 5.3 15.8 10.2 11.8 10.8 Investments in fixed capital 0.4 14.4 20.8 8.9 27.6 28.9 12.6 11.8 Foreign trade Export of goods and services (US$ `000,000), % -3.2 14.4 8.0 11.1 24.1 18.1 6.6 5.3 Import of goods and services (US$ `000,000), % -19.1 18.9 13.0 7.4 34.7 13.6 6.8 5.4 Prices and exchange rate CPI, % pa 22.7 28.2 12.0 0.8 5.2 9.1 6.2 5.0 PPI, % pa 31.1 20.9 8.7 3.0 7.7 20.4 7.0 5.1 Exchange rate, average, UAH/US$ 4.13 5.44 5.37 5.32 5.33 5.32 5.29 5.25 Money and credit extension NBU refinancing rate, % pa 44.0 29.6 20.2 9.2 7.0 9.0 7.0 5.3 Interest rate on loans, % pa 53.6 40.3 31.9 24.8 18.0 16.5 13.0 9.2 Interest rate on deposits, % pa 20.8 13.5 11.2 7.8 6.5 6.3 6.2 6.0 Money base, % pa 39.0 40.0 37.4 33.6 30.1 34.2 25.5 17.2 Money supply (M3), % pa 40.5 46.1 41.9 41.8 46.5 32.1 27.7 18.1 Government budget Total revenue, % GDP 25.2 28.9 26.9 27.4 28.5 24.5 24.7 24.9 Total expenditure, % GDP 26.7 28.3 27.2 26.7 28.7 25.4 24.9 25.0 Balance, % GDP -1.5 0.6 -0.3 0.7 -0.2 -3.2 -2.2 -0.1 Social indicators Real monthly wages, % -8.9 -0.9 19.3 18.2 15.2 8.8 7.7 6.3 Unemployment rate (ILO methodology), % 11.4 11.1 10.9 10.1 9.3 9.4 9.3 9.1 Notes: CPI = Consumer Price Index; PPI = Producer Price Index; NBU = National Bank of Ukraine; M3 = Broad Money Aggregate. Sources: Derzhkomstat (2004a and b), Ministry of Economy and European Integration (2004), NBU (2004), and Institute for Economic Forecasting of the National Academy of Sciences of Ukraine. 88 ment activity, and increased interna- Table A7-2. Macroeconometric Model: Estimated Percentage Difference in tional competitiveness of the Selected Macroeconomic Indicators in Three Epidemic Scenarios, Ukrainian economy. The following 2004-14 assumptions about the external and internal factors were made: Variable 2005 2014 2005 2014 2005 2014 External factors: Global future AIDS scenario Optimistic Medium Pessimistic economic development (average per GDP (production) GDP96 -0.44 -1.57 -0.39 -2.48 -0.40 -3.01 annum over 2005-14): growth rates GDP (expenditure) GDP96_SUM -0.41 -1.43 -0.37 -2.25 -0.37 -2.74 of the world GDP 2.5-3.1 percent; Investments I96 -0.43 -1.46 -0.39 -2.28 -0.39 -2.78 EU GDP 1.6-2.3 percent; Russian Federation GDP 4.5-5.6 percent; Imports M96 0.06 0.17 0.05 0.28 0.05 0.34 world GDP deflator 2.0-2.4 percent; GDP per employee P96 -0.12 -0.51 -0.11 -0.80 -0.11 -0.99 global crude oil prices US$ 38-42 per Savings S96f -0.02 -0.06 -0.02 -0.09 -0.02 -0.11 barrel; world trade growth 6.1-8.0 Budget revenue REV -0.39 -1.34 -0.36 -2.11 -0.36 -2.57 percent. Source: Authors' calculations. Internal factors: Annual average Note: All variables are in 1996 constant prices. Percentage difference relates to the difference with the inflation rates in 2005-06 increase to underlying "no-AIDS" demographic scenario. 14-10 percent, returning to 5.0-4.5 percent over 2007-14; factor increase gap between real wages and labor productivity in price of natural monopolies products/services growth rates; decline in the share of imports in total (2014 compared 2004): electricity 1.6, natural gas 1.8, consumption despite the real exchange rate appreci- railroad cargo transportation 1.4; slower real appreci- ation; anticipatory growth of investments versus ation of UAH (based on the expected USA inflation) GDP dynamics; and increased share of debt and by 1.0-1.1 percent per annum over 2004-14; reduction FDIs in total investment. The baseline scenario in the nominal interest rate on loans from 18.5 per- assumes successful implementation of the macro- cent in 2005 to 9.5 percent in 2014; a corresponding economic program, focussing on improved efficien- reduction in an average interest rate on deposits from cy of enterprises, their restructuring and moderniza- 6.5 percent to 6.0 percent; gradual increase in wages tion, banking system and financial stability, and a as share of GDP to 46.2-46.5 percent; reduction in favorable business climate attracting foreign and budget deficit from 3.2 percent in 2005 to the bal- expatriated domestic capital. anced budget in 2014; net foreign direct investments of at least US$ 1.5-2.0 billion per annum over 2004-14. The estimated macroeconometric model was applied, using the labor force projections discussed These are the underlying model assumptions of the in Chapter 4, under the three epidemic scenarios. baseline macroeconomic scenario of the Ukrainian Changes in employment impact aggregate supply, economy for 2005-14, without the HIV/AIDS effects GDP, sectoral value-added, labor productivity, and (Table A7-1). through these indicators, other variables, such as import of goods and services, household savings, The baseline scenario is based on the following budget balance, etc. Macroeconomic costs of the assumptions: high average annual rates of economic HIV/AIDS epidemic are quantified as the difference growth (through innovation/technical change); between the projected endogenous variables in increase in real household incomes; increase in three epidemic scenarios compared to the baseline investments in fixed capital at a rate faster than GDP (no-AIDS) scenario. growth; growth in net exports; moderate government spending; labor productivity growth; decrease in the 89 Table A7-3. Rating of Oblasts (Major Agricultural Producers) with Respect to HIV Prevalence, 2004 Per thousand HIV prevalence Labor Oblast group Employment* Unemployment* force AR of Crimea 5 1,108.0 79.3 1,187.3 Vinnitsa 2 826.7 49.2 875.9 Volyn 3 445.6 61.4 507.0 Dnipropetrovsk 5 1,645.4 130.9 1,776.3 Donetsk 5 2,129.0 183.9 2,312.9 Zhytomyr 2 509.2 74.9 584.1 Kyiv 4 739.6 81.5 821.1 Kirovograd 3 462.1 51.4 513.5 Lviv 2 1,110.4 130.0 1,240.4 Odesa 5 1,065.2 60.3 1,125.5 Poltava 2 677.4 61.6 739.0 Sumy 1 564.3 76.0 640.3 Ternopil 1 374.5 56.1 430.6 Kharkiv 3 1,308.8 139.2 1,448.0 Khmelnytsky 3 543.5 83.3 626.8 Cherkasy 4 549.9 69.2 619.1 Chernigiv 4 539.6 61.3 600.9 * Source: Derzhkomstat (2004, p. 850). Weighted average ranking 3.6, Gross agricultural product in selected regions UAH 43 billion, 78 percent of total agricultural production in 2003. Model Application into the macroeconometric model, we quantify the Under the employment scenarios (without AIDS and effects of HIV/AIDS on GDP (in production and three "with AIDS" scenarios based on medium, opti- expenditure measures), accumulated fixed capital, mistic and pessimistic epidemic forecasts) devel- import of goods and services, labor productivity, oped in Chapter 4, national labor force is predicted household savings, and budget balance. All of these to decline over 2004-14 by 2.3 million (11.6 percent) indicators (in constant 1996 prices) decline over under the "no-AIDS, underlying demographic sce- 2004-14 below what would have been their bench- nario, 2.5 million (12.7 percent) under the AIDS opti- mark value in the "no-AIDS" scenario (see Table A7-2 mistic scenario, taking into account epidemiologic on page 89). data, and 2.7 million (13.6 percent) under the pes- simistic scenario. Feeding the labor force projections 90 Table A7-4. Regional Trends in Average Monthly Nominal Wages, 1999-2003 1999 2000 2001 2002 2003 National average 100.0 100.0 100.0 100.0 100.0 Crimea 94.4 97.8 96.8 95.1 93.7 Vinnitsa 72.5 69.1 69.1 70.4 72.1 Volyn 66.3 65.1 64.6 67.2 69.0 Dnipropetrovsk 117.4 118.8 119.0 116.4 113.8 Donetsk 123.6 127.1 123.2 120.1 119.0 Zhytomyr 75.3 71.3 70.7 71.2 72.2 Zakarpatska 73.0 74.7 76.5 78.4 81.9 Zaporizhya 120.8 125.8 121.9 118.2 117.1 Ivano-Frankivsk 78.7 81.8 83.3 84.5 86.9 Kyiv 107.3 104.9 101.9 100.4 101.7 Kirovograd 77.0 73.8 74.3 74.9 76.3 Lugansk 103.4 101.0 102.9 104.4 102.4 Lviv 85.4 85.4 87.5 90.1 90.7 Mykolayiv 94.9 98.8 105.1 105.7 101.8 Odesa 102.8 102.5 98.4 100.7 98.3 Poltava 97.2 95.6 93.9 94.1 94.5 Rivne 75.8 75.1 78.8 82.9 84.5 Sumy 84.3 84.1 83.3 81.6 82.1 Ternopil 62.9 58.6 61.1 63.0 65.8 Kharkiv 103.4 100.0 99.7 98.3 98.3 Kherson 80.3 75.1 74.9 76.8 76.9 Khmelnytsky 71.3 67.7 67.8 68.5 69.8 Cherkasy 82.0 76.1 73.6 73.3 75.8 Chernivtsi 69.1 68.2 70.1 72.0 74.4 Chernigiv 79.2 76.7 75.6 73.6 74.0 The city of Kyiv 170.2 176.0 176.5 170.8 164.6 The city of 105.1 109.0 104.5 101.2 105.2 Sevastopol Source: State Statistics Committee of Ukraine. 91 APPENDIX A7. ESTIMATED MACROECONOMETRIC MODEL Coefficients are estimated using 1986-2003 data REAL SECTOR MODEL (14.47) (-3.16) (2.15) (-1.99) Aggregate supply block R2 = 0.981 DW = 1.46 S.E. = 1444.747 Total supply, at 1996 prices: Gross domestic product (demand), at 1996 prices: D(LOG(AS96)) = 0.4350·D(LOG(I96)) +0.4215·D(LOG(L)) 2GDP96 = C96 + I96 + INV96 + X96 - M96. + 0.3971·D(LOG(M96)) - Gross domestic product, at the prices of current period: (5.23) (1.28) (5.23) GDP = C + I + INV + X - M. 0.1626·F94 Total consumption, at the prices of current period: C = CG + CP. (-3.21) Government consumption, at the prices of current period: R2 = 0.855 DW = 2.52 S.E. = 0.034 CG = CG96·DEFCG. Total employment: Private consumption, at the prices of current period: L = 4.5516 + 9.25e-06·GDP96 + 0.7462·L(-1) CP = CP96·DEFCP. (1.51) (1.59) (4.86) Gross investments at the prices of current period: R2 = 0.905 DW = 2.14 S.E. = 0.587 I = I96·DEFI. Fixed assets, at 1996 prices: Change in current assets stock [inventory], at the prices of current period: K96=K96(1)+DK96 INV = INV96·DEFINV. Growth of fixed assets, at 1996 prices: DK96=I96 (1-0.6) GDP (supply), at the prices of current period: 1GDP = GDP96 · DEFGD. Gross domestic product (supply), at 1996 prices: 1GDP96 = AS96 - M96 GDP (demand), at the prices of current period: 2GDP = CP1 + CG + I + INV + X - M. Aggregate demand block Total consumption, at 1996 prices: Foreign trade block C96 = CG96 + CP96. Export of goods and services, at 1996 prices: Government consumption, at 1996 prices X96 = 3.0627·WGNP96 + 73.6083· CG96 = 0.1964·GDP96 +612.8428·F92 - 200.4952 (EO·(US_DEFGDP/DEFX)) - 44274.9266 (21.66) (1.33) (-0.16) (5.93) (1.15) (-3.03) R2 = 0.968 DW = 1.79 S.E. = 176 2.59 R2 = 0.851 DW = 2.15 S.E. = 2847.32 Private consumption, at 1996 prices: Import of goods and services, at 1996 prices CP96 = 10194.982 + 0.4795· GDP96 + 1.2969·(RDN-INFCPI) M96 = -0.0555·GDP96 - 355.6163·(EO·(DEFM/DEFGDP)) +1385.1064·TREND+25684.5864 (5.41) (34.21) (2.43) R2 = 0.987 (-1.11) (-1.24) (1.48) 2.18 DW = 1.77 S.E. = 2692.36 R2 = 0.769 DW = 2.68 S.E. = 3308.206 Gross investments, at 1996 prices: I96 = 0.228·GDP96 - 1.0957·(RKN-INFPPI) + Export of goods and services, at the prices of current period: 0.0312·K96(-1) -17713.3034 X = X96·DEFX. 92 Import of goods and services, at the prices of current period: DEFGDP - GDP deflator, 1996 = 1; M = M96·DEFM. US_DEFGDP - global GDP deflator, 1996 = 1; Trade balance, at 1996 prices: DEFCP - private consumption deflator, 1996 = 1; TB96 = X96 - M96. DEFCG - state consumption deflator, 1996 = 1; DEFI - gross investments deflator, 1996 = 1; Trade balance, at the prices of current period: DEFINV - deflator of change in current assets stock, TB = X - M. 1996 = 1; Export of goods and services at current prices, USD: DEFM - import deflator, 1996 = 1; X$ = X96·PX$. PM$ - import price index, USD; Import of goods and services at current prices, USD: DEFX - export deflator, 1996 = 1; M$ = M96·PM$. PX$ - export price index, USD; Trade balance at current prices, USD: DD - total amount of deposits; TB$ = X$ - M$. RDN - deposit interest rate; RKN - credit interest rate; EO - exchange rate, UAH for USD 1; LIST OF VARIABLES TREND - time factor; Endogenous variables F92, F93, F94, F95 - dummy variables (to allow for L - total employment, mil; the structural change in GDP - GDP at current prices; 1992-1995). GDP96 - GDP at 1996 prices; CO - total consumption, at current prices; DISAGGREGATE VARIABLES BLOCK CO96 - total consumption, at 1996 prices; (models of appraisal of gross value added (GVA) as CP - private consumption, at the prices of current period; per the kinds of economic activity). CP96 - private consumption, at 1996 prices; Economic and mathematical models in macrosectors have CG - state consumption, at the prices of current period; been constructed as the models of forecasting the GDP in CG96 - state consumption, at 1996 prices; terms of production functions applying relevant behavior DK96 - growth of fixed assets, at 1996 prices; regression equations relating to the assessment of employment I - gross investments, at the prices of current period; and fixed assets as per the basic kinds of economic activity I96 - gross investments, at 1996 prices; (time horizons: 1994-2004). REV - gross budget receipts; Model specification has been elaborated pursuant to the fol- INV - change in current assets stock, at the prices of lowing: the first two letters show the name of economic indi- current period; cator (VA - gross value added, EM - employment, K - fixed INV96 - change in current assets stock, at 1996 prices; assets, DK - net annual growth of basic assets); the next three- K96 - fixed assets, at 1996 prices; four letters indicate relevant kind of economic activity; the M - import, at the prices of current period; remaining symbols are the prices for economic indicators. All M96 - import, at 1996 prices; indicators in the models of relevant blocks are calculated at M$ - import, USD; basic prices of 1996, so the identifiers of relevant variables, AS96 - total supply, at 1996 prices; excluding the number of employed persons in macrosectors, X - export, at the prices of current period; end in figures 96. X96 - export, at 1996 prices; Specification of block of the models intended for X$ - export, USD; assessment of gross value added (GVA) as per the Exogenous variables kinds of economic activity WGNP96 - global GDP at 1996 prices, USD; 1. Economy as a whole: INFCPI - growth of consumer price index; VA_ALL96 = 0.3710*K1_ALL96 + 8201.0788*EM_ALL - INF_PI - growth of producer price index; 430735.2378 93 (10,18) (3,62) (-5,41) (-0,5735) (3,5278) (-1,8336) (-4,6155) R2 = 0,949 DW = 1,73 S.E. = 4271,36 R2 =0,862 DW = 2,40 S.E. = 0,1482 EM_ALL = -1.0306e-05*VA_ALL96 + 3.2402e-05*K1_ALL96 K1_INA96 = K_INA96(-1) + DK_INA96 .8345 - 0.7893*TREND 5. Production and distribution of electric power, gas (-0,69) (3,38) (-0,13) (-7,08) and water: R2 = 0,982 DW = 2,61 S.E. = 0,1485 VA_INE96 = 0.1763*K1_INE96 + 2268.3435*EM_INE - 7939.2236 K1_ALL96 = K_ALL96(-1) + DK_ALL96 (6,39) (2,65) (-3,06) 2. Agriculture, hunting, forestry and fishing R2 = 0,889 DW = 2,39 S.E. =194,5263 industry: VA_AGR96 = 0.0392*K1_AGR96 + 3341.6137*EM_AGR - LOG(EM_INE) = 0.6253*LOG(VA_INE96) + 12203.0545 + 200.9895*LOG(TREND) 59556.3731*(1/(K1_INE96)) - 6.0995 - 0.0288*TREND (0,57) (2,10) (-1,03) (0,38) R2 = 0,450 DW = 1,66 S.E. = 765,795 (1,26) (0,66) (-1,10) (-1,58) R2 =0,860 DW = 1,83 S.E. = 0,050 EM_AGR = 0.000134*VA_AGR96 - 4.3735e-06*K1_AGR96 + 4.3738 - 0.0133*TREND K1_INE96 = K_INE96(-1) + DK_INE96 (2,23) (-0,34) (2,83) (-0,76) 6. Construction: R2 = 0,470 DW = 2,05 S.E. = 0,1474 VA_CON96 = 0.2546*K1_CON96 + 4810.7954*EM_CON - 5026.1759 K1_AGR96 = K_AGR96(-1) + DK_AGR96 (1,35) (4,32) (-1,51) 3. Mining industry: R2 = 0,732 DW = 2,03 S.E. =590,3322 VA_INI96 = 0.1401*K1_INI96 + 1068.8495*EM_INI - 4183.2123 EM_CON = 7.6729e-05*VA_CON96 + 0.00013*K1_CON96 - 0.8327 - 0.0651*TREND (11,01) (1,87) (-4,08) R2 = 0,959 DW = 1,78 S.E. = 134,0027 (1,19) (0,97) (-0,52)(-1,35) R2 =0,7963 DW = 1,25 S.E. = 0,099 EM_INI = 2.3338e-05*VA_INI96 + 3.7847e-05*K1_INI96 - 0.7579 - 0.08499*TREND K1_CON96 = K_CON96(-1) + DK_CON96 (0,4864) (3,8371) (-2,4373) (-10,7983) 7. Wholesale and retail trade, trade in transport R2 =0,980 DW = 2,70 S.E. = 0,0173 facilities, repair services: VA_TR96 = 0.5133*K1_TR96 + 1575.18601*EM_TR - K1_INI96 = K_INI96(-1) + DK_INI96 5900.9751 + 141.2162*TREND 4. Manufacturing industry (2,11) (0,62) (-2,39) (0,52) VA_INA96 = 0.3858*K1_INA96 + 3542.3093*EM_INA - R2 =0,9303 DW = 1,70 S.E. =934,4289 53568.8255 EM_TR = 3.8287e-05*VA_TR96 + 2.8623e-05*K1_TR96 + (12,24) (2,97) (-6,97) 0.8123 - 0.0026*TREND R2 = 0,961 DW = 1,84 S.E. =952,2076 (0,62) (0,59) (1,10) (-0,59) EM_INA = -7.4318e-07*VA_INA96 + 5.383e-05*K1_INA96 - R2 =0,7660 DW = 2,51 S.E. =0,1456 3.5575 - 0.2949*TREND K1_TR96 = K_TR96(-1) + DK_TR96 94 8. Transport and communications: EM_GOV = 8.1139e-05*VA_GOV96 + 1.9106e-05*K1_GOV96 - 0.1142 LOG(VA_TCO96) = 0.6540*LOG(K1_TCO96) + 1.0976*LOG(EM_TCO) + 1.3929 (1,16) (0,85) (-0,21) (3,00) (3,02) (0,56) R2 =0,783 DW =1,93 S.E. =0,0521 R2 =0,844 DW = 1,22 S.E. =0,0540 K1_GOV96 = K_GOV96(-1) + DK_GOV96 EM_TCO = 1.1279e-05*VA_TCO96 + 1.4958e-05*K1_TCO96 - 12. Education: 0.0864 - 0.0503*TREND VA_EDU96 = 0.0538*K1_EDU96 + 432.0102*EM_EDU - (0,57) (1,84) (-0,16) (-1,97) 602.6311 R2 =0,7736 DW = 2,43 S.E. =0,0445 (0,89) (1,68) (-0,15) K1_TCO96 = K_TCO96(-1) + DK_TCO96 R2 =0,612 DW =0,71 S.E. =468,8820 9. Financial activity: LOG(EM_EDU) = -0.4201*LOG(VA_EDU96) + 0.1215*LOG(K1_EDU96) + 3.4753 - LOG(VA_FIN96) = 0.8387*LOG(K1_FIN96) + 1.1783*EM_FIN 0.34098*LOG(TREND) + 0,51 (-1,05) (0,71) (( (0,28) (-4,25) (9,95) (2,78) (0,32) R2 =0,553 R2 =0,842 DW = 1,56 S.E. =0,0939 DW = 0,94 S.E. =0,1626 K1_EDU96 = K_EDU96(-1) + DK_EDU96 LOG(EM_FIN) = -0.0866*LOG(VA_FIN96) + 0.03994*LOG(K1_FIN96) - 1.2761 - 13. Health care and social aid: 0.0053*TREND VA_HEC96 = 0.15004*K1_HEC96 + 2575.3893*EM_HEC - (-0,96) (0,45) (-0,76) (-0,23) 3024.6736 R2 =0,546 DW = 2,86 S.E. =0,0641 (2,16) (1,43) (-1,23) K1_FIN96 = K_FIN96(-1) + DK_FIN96 R2 =0,517 DW =0,81 S.E. =257,5275 10. Operations with real estate, leasing and services EM_HEC = 4.2034e-05*VA_HEC96 + 9.2602e-06*K1_HEC96 to legal entities: + 1.1866 - 0.01852*TREND VA_REL96 = 0.16797*K1_REL96 + 1423.2994*EM_REL - (1,26) (0,89) (6,18) (-4,49) 21028.7486 R2 =0,853 DW = 2,56 S.E. =0,0245 (4,62) (0,68) (-5,49) K1_HEC96 = K_HEC96(-1) + DK_HEC96 R2 = 0,975 DW = 2,13 S.E. =285,7419 14. Collective, public and personal services, house- EM_REL = 2.3452e-05*VA_REL96 + 1.4348e-06*K1_REL96 + maid services, exterritorial activity: 0.2387 + 0.0264*TREND LOG(VA_SER96) = 1.5499*LOG(K1_SER96) - (0,91) (0,75) (0,20) (2,34) 0.3157*EM_SER - 500.5321 R2 =0,9498 DW = 2,55 S.E. =0,0356 (1,75) (-0,77) (-0,88) K1_REL96 = K_REL96(-1) + DK_REL96 R2 =0, 0,507 DW =0,72 S.E. =0,1422 11. State management: EM_SER = 2.6679e-05*VA_SER96 + 1.4354e-05*K1_SER96 + 0.5839 - 0.0262*TREND LOG(VA_GOV96) = 2.1161*LOG(K1_GOV96) + 0.4685*LOG(EM_GOV) - 11.7199 (0,76) (0,60) (1,45) (-2,33) (2,28) (1,08) (-1,42) R2 =0,712 DW = 1,23 S.E. =0,0446 R2 =0,862 DW = 2,84 S.E. =0,0765 95 K1_SER96 = K_SER96(-1) + DK_SER96 (3.492) (10.883) (-3.329) THE MODEL OF CONSUMPTION SECTOR AND 62140.0199·RATE1. PERSONAL INCOME: (-3.1345) Disposable income: R2 = 0.976 DW =1.907 S.E. = 690.9 DI = GDP-REQ. Real average monthly wages and salaries, at 1996 VAT losses: prices: NTAX1=NTAX1_PROC_M·TAX1/100 . @PCH(WG96) = -0.1654·U_REAL - 0.0002·INFCPI + 0.1154·TREND Personal income tax: TAX2 = 0.1037387371·(ZARP/ZABZP) - (-2.58) (-3.48) (2.81) 231.5200362·(@PCH(GOV2)/@PCH(DEFGDP)). R2 = 0.731 DW =2.39 S.E. = 0.2796 (62.709) (-2.298) Real average monthly pension, at 1996 prices: R2 = 0.996 DW =2.575 S.E. = 151.374 PENS96 = 0.1804·WG96 + 3.2552 Excise duty: (11.26) (0.54) LOG(TAX3) = -2.95933241 + R2 = 0.894 DW = 0.83 S.E. = 19.1399 0.9397966208·LOG(ZARP/ZABZP) + 0.2665973369·GOV3. Savings, at 1996 prices: (-6.396) (15.278) (0.991) S96 = 0.01638955483·DI96 + 55.67839394·RDR + 1664.713306·TREND - 11008.73595 R2 = 0.988 DW =2.018 S.E. = 0.336 (0.45) (1.56) (4.41)(-1.50) Land rent: R2 = 0.855DW =1.09 S.E. = 3083.401 LOG(TAX5) = -1.547444297 + 1.001769604·LOG(RATE5). (-8.023) (37.35) Difference between investments and savings, at 1996 prices: S96_DIF = (I96 - S96)/S96 ·100 R2 = 0.995 DW =1.427 S.E. = 0.209 Labor productivity, at 1996 prices: State duty: P96 =GDP96/L/1000 1/LOG(TAX6) = 0.05979593283·LOG(GOV2) + Unemployment level: 0.006271083018·(1/LOG(GOV6)). LOG(U_REAL) = 0.2301574098·LOG(DEFGDP) + (2.775) (42.49) 2.14804849 R2 = 0.739 DW =1.899 S.E. = 0.236 THE MODEL OF PUBLIC FINANCES SECTOR: Income tax: Total budget receipts: LOG(TAX7) = -4.171087259 + 1.113438194·LOG(BASE7) + REV = TAX1+TAX2+TAX3+TAX4+TAX5+TAX6+TAX7+ 5.304431121·RATE7 + 0.0238788037·TREND. TAX8+TAX09. (-16.317) (20.35) (3.73) (1.44) Total budget expenditures: R2 = 0.998 DW =2.537 S.E. = 0.067 EXPD=REV+SLGDP·GDP. Budgetary balance: Tax on foreign economic activity: SALDO=REV-EXPD. 1/LOG(TAX8) = -0.06397661699 + 2.878426109( (1/LOG(M$·EO)) - 0.06248051978·GOV8 Value added tax: TAX1+(NTAX1-NTAX1(-1)) = 19538.81354 + (-3.397) (18.417) (-3.679) 0.06423515397·GDP - 13880.34951·GOV1 - 0.001634614739·TREND. 96 Money emission: (-1.956) EMMO =EMMO(-1) · (1+EMMO_D/100) R2 = 0.998 DW =3.079 S.E. = 0.003 Credit interest rate: Other proceeds: LOG(RKN) = 0.0001971569775·INFCPI + TAX09=TAX09_STR·REV/100, 0.9670027797·LOG(RKN(-1)) where: TAX1, TAX2, TAX3, TAX5, TAX6, TAX7, TAX - the pro- (-2.99) (46.32) ceeds according to the kinds of taxation; RATE5, RATE7 -tax R2 = 0.772 DW =1.71 S.E. = 0.3882 rates; GOV1, GOV3, GOV6, GOV8 -artificial instrumental vari- ables indicating the administration (fiscal) level pursuant to Money circulation rate (_2): the kinds of taxes; ZARP, ZABZP - wages and salaries fund and VM2f = 0.182280325·RKNf - 0.1732829754·RDN + wage arrears. 2.938641311 THE MODEL OF MONETARY AND CREDIT SECTOR: (4.41) (-4.06) (1.89) Money supply: R2 = 0.783 DW =0.85 S.E. = 1.5081 M2/DEFGDP = 1.121960075·GDP96 - 116.6708509·RKN - GDP_MV = M2·VM2 737.1711593·TREND2. GDP_MV_DIF = (GDP - GDP_MV)/GDP_MV·100 (8.2706) (-1.536) (-4.891) R2 = 0.905 DW = 2.06 S.E. = 14907.1 PCM2 =@PCH(M2) ·100 97 ANNEX 8 CGE Model: Methodology and Assumptions Model Description Households are endowed with labor and capital and To study the economy-wide impact of HIV/AIDS in receive transfers. They spend a constant share of Ukraine, we use a Computable General Equilibrium their income for investment goods. Final consump- (CGE) model. The literature contains a large range tion is modeled by a Cobb-Douglas function of a rep- of different models (See e.g., de Melo [1988], resentative household. Francois and Shiells [1994] or Devarajan and Robinson [2002]) for general surveys. The theoretical The government receives revenue from taxes and basis of the present modeling exercise is the applied tariffs as well as from factor endowments. They use general equilibrium framework discussed by Shoven income to provide public goods. In all scenarios, the and Whalley (1992). Based on this framework, we indirect tax rate adjusts endogenously so that the use a standard specification as, e.g., used in real value of public goods remains constant. Total Harrison, Rutherford, and Tarr (1997a and b), the investments equal the sum of depreciation, public static model of Pavel (2001) or in the basic static and private savings, and the current account specification of Jensen, Rutherford, and Tarr (2003). balance. The model is programmed in GAMS/MPSGE as described in Rutherford (1999), an algebraic form of Since all supply and demand functions in our model this standard specification can be found in Pavel are homogeneous of degree zero in prices, one price (2001) or Rutherford and Paltsev (1999). (the so-called numeraire) has to be fixed exogenous- ly while all other endogenous price variables define An overview of the model structure is in Figure A8-1. the change relative to this numeraire. The choice of Production takes place under Constant Returns to the numeraire as such has no impact on the results. Scale and all production factors are perfectly mobile. In our model, we chose the price index for invest- Consumers treat imported and domestically pro- ment goods. duced goods as imperfect substitutes while produc- ers regard sales on domestic markets or exports as imperfect alternatives. This standard assumption is Data Requirements and Sources based on Armington (1996). Exports and imports are The basis for our modeling exercise is a Social disaggregated into different trading partners, mod- Accounting Matrix (see e.g., Pyatt and Round [1985]) eled with constant elasticities of transformation and compiled on the basis on Ukraine's National substitution. Direct taxes are modeled as an activity- Accounts and Input-Output tables for 2001 (in basic specific tax on the use of labor and capital (tax rates and consumer prices). are negative if input/output tables report net subsi- dization). Indirect taxes are modeled as a commodi- Remuneration for labor has been disaggregated into ty-specific tax on private (household) and invest- three different categories using the 2001 edition of ment demand (tax rates are negative if input/output Labor of Ukraine (Pratsya Ukrainy) and the tables report net subsidization). Import tariffs are Statistical Yearbook 2002 of Ukraine. commodity- and region-specific and apply for all imports. 98 Figure A8-1. CGE Model: Structure Domestic Demand ­ Indirect taxes on private and investment demand ­ No indirect taxes on public and intermediate demand s = 5 Region 1 Region 7 (see Table 2) ··· (see Table 2) t = 3 Imports Exports Domestic sales ­ Tariffs by goods and regions s = 3 t = 5 Region 1 ··· Region 7 (see Table 2) (see Table 2) Domestic Output s = 0 Value Added Intermediate ­ Direct tax on labor demand and capital s = 1 Unskilled Skilled High skilled Capital Labor Labor Labor Note: s denotes elasticity of substitution; t denotes elasticity of transformation. Source: Institute for Economic Research and Policy Consultations model. Parameter Parameters (taken from Jensen, Rutherford, and Tarr [2003]) Elasticity of substitution between labor and capital .................................................... 1 Elasticity of substitution between Value Added and Intermediates............................... 0 Elasticity of substitution between imports and domestic goods ................................... 5 Elasticity of transformation between domestic output and exports ............................. 5 Elasticity of substitution between imports of different origin........................................ 3 Elasticity of transformation between exports to different destinations ......................... 3 99 Table A8-1. CGE Model: Composition of Labor Endowment Data Decline National Accounts/Input-Output table for 2001 In percentage Labor of Ukraine 2001 (Pratsya Ukrainy 2001) Scenario Labor endowment decline Statistical Digest, Kiev 2002 Unskilled Skilled High-skilled Foreign trade of Ukraine in 2001, vol.2, Statistical Pessimistic 0.016 0.063 0.063 publication, State Statistics Committee of Ukraine, Medium 0.008 0.031 0.031 Kiev, 2002. Optimistic 0.006 0.023 0.023 Cooperation between Ukraine and EU countries in 2002, Statistical publication, State Statistics Source: Authors' calculations. Committee of Ukraine, Kiev, 2003. Foreign trade of goods and services of Ukraine in 2002, Vol. 1, Statistical publication, State Statistics Committee of Ukraine, Kiev, 2003. Imports and exports in the Input-Output tables have been disaggregated into different origins and destina- tions using Ukraine's foreign trade statistics of the Model Application and Results State Statistics Committee. Reduction in Labor Supply. Input-Output tables include 38 activities/commodi- The main channel through which the HIV/AIDS epi- ties. They have been aggregated to 20 sectors as demic affects economic development is increased explained in Table A8-6. mortality and morbidity, which directly afflicts the labor force supply and productivity. A smaller overall Exports and imports by activity and commodity are population and labor force have their repercussions aggregated into the seven main trading regions on the production side (the economy's production (European Union, EU Accession Candidates, Russian potential) and the expenditure side of the economy Federation, CIS countries, Asia, North America, and (final household consumption, residential fixed rest of the World). investment, and demand for government services). The Input-Output table contains information about According to the forecast by the Center for Social two types of taxes, revenue from "taxes/subsidies on Studies of the Ministry of Labor and Social Policy of production" (direct taxes) and from "taxes/subsidies Ukraine (reported in Chapter 4), labor force endow- on commodities" (indirect taxes). ment will decrease by 2 percent by 2014 in the base- line scenario. We consider 1.5 percent, 2 percent, and The rates of taxes and subsidies on production are 4 percent decline in labor force endowment for opti- calculated in percentage of labor and capital costs mistic, medium, and pessimistic scenarios respective- per industry. ly, based on the labor force projections of Chapter 4. The most pessimistic scenario in terms of labor sup- Taxes and subsidies on commodities are split into ply shock is based on the projected magnitude of import tariffs (commodity- and trade region-specific labor force decline in the most affected regions. apply to all imports) and consumption taxes, includ- ing mainly VAT as well as excises and transit and The prevalence of HIV is uneven within the labor road fees on specific products, (commodity specific force across different skill classes. We have assumed on final and investment demand). in what follows that the decline for skilled and high- skilled labor is four times larger than that for 100 Table A8-2. Structure of Ukrainian Economy (CGE Model) Structure of value added (in %) TFP index Labor Unskilled Skilled High skilled SCENARIOS Sectors Capital (total) labor labor labor Pessimistic Medium Optimistic Agriculture, hunting 0.81 0.19 0.15 0.02 0.02 0.997 0.998 0.999 Fishery 0.29 0.71 0.47 0.14 0.10 0.983 0.990 0.996 Mining of coal and peat 0.24 0.76 0.43 0.18 0.15 0.977 0.987 0.995 Production of non-energy materials 0.12 0.88 0.50 0.21 0.18 0.973 0.985 0.994 Food-processing industries 0.49 0.51 0.29 0.12 0.10 0.984 0.991 0.997 Textile and leather industry 0.44 0.56 0.32 0.13 0.11 0.983 0.990 0.996 Woodworking, pulp and paper 0.47 0.53 0.30 0.13 0.11 0.984 0.991 0.997 industry, publishing Petroleum refinement 0.76 0.24 0.13 0.06 0.05 0.993 0.996 0.998 Manufacture of chemicals, rubber and 0.39 0.61 0.34 0.14 0.12 0.982 0.989 0.996 plastic products Manufacture of other non-metallic 0.30 0.70 0.40 0.17 0.14 0.979 0.988 0.995 products Metallurgy and metal processing 0.43 0.57 0.32 0.14 0.11 0.983 0.990 0.996 Manufacture of machinery and 0.33 0.67 0.38 0.16 0.13 0.980 0.988 0.996 equipment Other 0.23 0.77 0.43 0.18 0.15 0.977 0.987 0.995 Electric energy 0.74 0.26 0.15 0.06 0.05 0.992 0.995 0.998 Public utilities 0.15 0.85 0.48 0.21 0.16 0.974 0.985 0.994 Construction 0.40 0.60 0.37 0.12 0.10 0.984 0.991 0.997 Trade 0.57 0.43 0.17 0.15 0.10 0.982 0.990 0.996 Hotels and restaurants 0.50 0.50 0.25 0.16 0.09 0.982 0.990 0.996 Transport 0.56 0.44 0.28 0.09 0.07 0.988 0.993 0.997 Post and telecommunications 0.61 0.39 0.21 0.10 0.08 0.987 0.993 0.997 Other services 0.32 0.68 0.18 0.18 0.31 0.966 0.980 0.993 Source: CGE model for Ukraine SAM 2001. unskilled, replicating the actual epidemic patern in that people with university education are significant- Ukraine, with urban population (predominantly ly less afflicted from HIV/AIDS, we did not distin- skilled labor) being significantly more affected com- guish between skilled and high-skilled, but assumed pared to the rural population (treated here as they incur identical HIV/AIDS prevalence and mortal- unskilled). This could mean that the average produc- ity rates. This resulted in the distribution of labor tivity of skilled and high-skilled labor would endowment decline among different skill classes. decrease and that overall demand for skilled and high-skilled labor could increase, making those skill As mentioned above, HIV/AIDS affects the availabili- classes scarcer. Since we did not find any evidence ty of employees with specific qualifications. This 101 Table A8-3 CGE Model: Macroeconomic Implications of HIV/AIDS Epidemic, Scenario Analysis Scenarios Medium subscenarios Higher Reduced Lower public Macro Indicators Benchmark Pessimistic Medium Optimistic labor TFP spending Welfare (equivalent variation, change in %) - -8.3 -4.6 -2.2 -2.6 -3.3 0.0 GDP Index (change in %) - -5.5 -3.1 -1.6 -1.8 -2.3 0.2 Private Investment (change in %) -9.0 -5.0 -2.4 -2.8 -3.6 0.0 Real factor return (change in %): ­ Return to capital - -7.03 -3.87 -1.90 -2.22 -2.55 -0.11 ­ Wage rate for unskilled labor - -7.46 -4.17 -1.78 -1.93 -3.55 -0.03 ­ Wage rate for skilled labor - -2.58 -1.70 0.07 0.55 -3.56 0.05 ­ Wage rate for highskilled labor - -1.42 -1.05 0.37 0.89 -3.18 0.14 Aggregate exports (UAH billion) 113.24 102.54 107.12 110.40 110.05 108.52 113.18 Aggregate imports (UAH billion) 109.92 99.09 103.74 107.05 106.69 105.18 109.84 Total exports (change in %) - -9.46 -5.41 -2.51 -2.82 -4.17 -0.06 Total imports (change in %) - -9.86 -5.63 -2.61 -2.94 -4.31 -0.08 Tariff revenue (share of public budget) 10% 9% 9% 10% 10% 10% 10% Indirect tax revenue (share of public budget) 49% 55% 52% 51% 51% 51% 50% Indirect tax rate (weighted average) 12% 15% 13% 12% 12% 12% 12% Consumer Price Index (change in %) - -0.75 -0.42 -0.18 -0.20 -0.34 -0.01 Producer Price Index (change in %) - -3.01 -1.59 -0.66 -0.77 -0.90 -0.29 Real exchange rate (change in %) - -1.60 -0.77 -0.34 -0.43 -0.25 -0.19 Source: CGE model for Ukraine simulations. follows from the changes in the working-age popula- Reduction in Labor Productivity tion brought about by increased mortality, but HIV/AIDS, by increasing morbidity and mortality, HIV/AIDS also affects the return to investment in affects both the productivity of employees living skills, i.e., wages. In general, workers' deaths shrink with the disease and productivity in general. the aggregate supply of labor, meaning that wages Productivity gradually declines because of increasing for workers with specific skills are likely to rise, or absenteeism and mortality, inflicting repercussions that workers who die will be replaced with others to both public and private sectors. This decline in having less skill and less experience. On the other worker productivity grows clearer as symptomatic hand, the age structure of the working-age popula- AIDS progresses. tion changes: increased mortality means that employees are, on average, younger, and that individ- Since there is no evidence for Ukraine on the magni- uals with substantial experience in their profession, tude of decline of labor productivity due to normally a prerequisite for leading positions in a HIV/AIDS, we used figures from a World Bank study company, become scarcer. 102 Table A8-4. CGE Model: Sectoral Implications of HIV/AIDS Epidemic, Scenario Analysis Scenarios Medium subscenarios Higher Reduced Lower public Output index Benchmark Pessimistic Medium Optimistic labor TFP spending Agriculture, hunting 1.00 1.02 1.01 1.00 1.00 1.01 1.00 Fishery 1.00 0.99 1.00 1.00 0.99 1.00 1.00 Mining of coal and peat 1.00 0.91 0.95 0.98 0.97 0.97 1.00 Production of non-energy materials 1.00 0.67 0.81 0.91 0.91 0.84 1.00 Food-processing industries 1.00 0.98 0.99 1.00 0.99 1.00 1.00 Textile and leather industry 1.00 0.99 1.02 1.01 1.01 1.05 0.98 Woodworking, pulp and paper industry, 1.00 1.03 1.03 1.01 1.00 1.04 1.00 publishing Petroleum refinement 1.00 0.95 0.97 0.99 0.99 0.98 1.00 Manufacture of chemicals, rubber and plastic 1.00 0.89 0.95 0.98 0.98 0.96 1.00 products Manufacture of other non-metallic products 1.00 0.90 0.95 0.98 0.97 0.97 1.00 Metallurgy and metal processing 1.00 0.63 0.78 0.91 0.90 0.81 1.00 Manufacture of machinery and equipment 1.00 0.91 0.96 0.97 0.97 0.99 1.01 Other 1.00 0.74 0.84 0.93 0.93 0.86 1.00 Electric energy 1.00 0.90 0.95 0.98 0.97 0.96 1.00 Public utilities 1.00 0.94 0.97 0.98 0.98 0.98 1.00 Construction 1.00 0.93 0.96 0.98 0.98 0.98 1.00 Trade 1.00 0.94 0.97 0.98 0.98 0.98 1.00 Hotels and restaurants 1.00 1.21 1.12 1.04 1.04 1.12 1.00 Transport 1.00 1.20 1.11 1.04 1.04 1.10 1.00 Post and telecommunications 1.00 1.01 1.01 1.00 1.00 1.02 1.00 Other services 1.00 1.00 1.00 1.00 0.99 1.01 1.00 Source: CGE model for Ukraine simulations. Note: All benchmark indexes equal unity (or 100 percent) reflecting the start position for the change. on economic consequences of HIV in Russia (Rhuel, We should mention that the HIV/AIDS epidemic has Pokrovsky, and Vinogradov 2002). They used a 13 uneven sectoral impacts for the economy (e.g., Sharp percent reduction in productivity of the HIV-positive 2002). Since HIV/AIDS directly harms labor produc- population. Nevertheless, due to low prevalence rate tivity, we could expect that labor-intensive sectors in Ukraine (1.98 percent, 2.34 percent, and 3.49 per- (for example, metallurgy, mining, and non-energy cent, respectively, by 2014), we assumed 1.5 percent, sectors) will be afflicted most. On the other hand, 4 percent, and 7 percent labor productivity reduction specifics of each sector contribute to uneven for optimistic, medium, and pessimistic scenarios, HIV/AIDS effects. For example in the construction respectively. 103 Table A8-5. Implications for Sectoral Exports Scenarios Medium subscenarios Higher Reduced Lower public Exports Benchmark Pessimistic Medium Optimistic labor TFP spending Agriculture, hunting 1.00 1.27 1.15 1.07 1.07 1.12 1.00 Fishery 1.00 1.00 1.01 1.00 1.00 1.01 1.00 Mining of coal and peat 1.00 0.85 0.92 0.96 0.96 0.94 1.00 Production of non-energy materials 1.00 0.62 0.78 0.90 0.89 0.81 1.00 Food-processing industries 1.00 1.06 1.04 1.02 1.02 1.03 1.00 Textile and leather industry 1.00 0.98 1.02 1.02 1.01 1.05 0.98 Woodworking, pulp and paper industry, 1.00 1.06 1.05 1.01 1.01 1.06 1.00 publishing Petroleum refinement 1.00 0.97 0.99 1.00 1.00 0.99 1.00 Manufacture of chemicals, rubber and 1.00 0.86 0.93 0.97 0.97 0.94 1.00 plastic products Manufacture of other non-metallic products 1.00 0.87 0.93 0.97 0.96 0.95 1.00 Metallurgy and metal processing 1.00 0.59 0.76 0.90 0.89 0.79 1.00 Manufacture of machinery and equipment 1.00 0.89 0.95 0.97 0.96 0.98 1.01 Other 1.00 0.65 0.79 0.91 0.90 0.81 1.00 Electric energy 1.00 1.00 1.00 1.00 1.00 1.01 1.00 Construction 1.00 0.95 0.98 0.99 0.98 1.00 1.00 Trade 1.00 1.02 1.02 1.00 1.00 1.03 1.00\ Hotels and restaurants 1.00 1.28 1.16 1.06 1.06 1.15 1.00 Transport 1.00 1.35 1.18 1.07 1.07 1.17 0.99 Post and telecommunications 1.00 1.16 1.09 1.03 1.03 1.09 1.00 Other services 1.00 0.93 0.97 0.98 0.97 1.00 1.01 Source: CGE model for Ukraine simulations. Note: All benchmark indexes equal unity (or 100 percent) reflecting the start position for the change. industry, the following factors facilitate the spread commercial sex workers, many of them injecting of epidemics: 1) work sites are often far from the drug users. homes of the laborers, 2) temporary housing for workers, and 3) few opportunities for leisure and Increase in Public Spending entertainment. Employees in the transport sector, HIV/AIDS also erodes the government's financial long distance truck drivers in particular, are also resources, from both the revenue and the expendi- susceptible to HIV infection. They too spend a great ture sides. Countries afflicted by the epidemic see deal of time away from home and often for long peri- their tax base and thus their domestic revenue grow ods. Many of them have frequent contacts with more slowly or even shrink, even as demand for 104 Table A8-6. Classification of Activities and Goods in the Model and in Ukraine's Input-Output Tables Model Classification Input-Output Classification a01 Agriculture, hunting a01 Agriculture, hunting a03 Fishery a03 Fishery a04 Mining of coal and peat a04 Mining of coal and peat a06 Production of non-energy materials a06 Production of non-energy materials a07 Food-processing a07 Food-processing a08 Textile and leather a08 Textile and leather a09 Forestry, wood working, pulp and paper industry, publishing a09 Wood working, pulp and paper industry, publishing a02 Forestry a11 Petroleum refinement, manufacture of coke products, a11 Petroleum refinement production of hydrocarbons a10 Manufacture of coke products a05 Production of hydrocarbons a12 Manufacture of chemicals, rubber and plastic products a12 Manufacture of chemicals, rubber and plastic products a13 Manufacture of other non-metallic mineral products a13 Manufacture of other non-metallic mineral products a14 Metallurgy and metal processing a14 Metallurgy and metal processing a15 Manufacture of machinery and equipment a15 Manufacture of machinery and equipment a16 Other production a16 Other production a17 Electric energy a17 Electric energy a20 Utility supply (water, gas, heat) a20 Water supply a18 Gas supply a19 Heat supply a21 Construction a21 Construction a22 Trade a22 Trade a23 Hotels and restaurants a23 Hotels and restaurants a24 Transport a24 Transport a25 Post and telecommunications a25 Post and telecommunications a26 Other services a26 Financial intermediation a27 Real estate transaction a28 Renting a29 Information activities a30 Research and developmen a31 Services to legal entities a32 Public administration a33 Education a34 Health care and social assistance a35 Sewage, cleaning of streets and refuse disposal a36 Social activities a37 Recreational, entertainment, cultural and sporting activities a38 Other activities a39 Financial intermediation services indirectly measured EXPLANATIONS: Service sectors (Financial intermediation, Real estate transactions, Renting, Information activities, Research and development, Services to legal entities, Public administration, Education, Health care and social assistance, Sewage, cleaning of streets and refuse disposal, Social activities, Recreational, entertainment, cultural and sporting activities, Other service activities) are added to a single service activity / commodity. "Manufacture of coke products" (Activity/Commodity 10) has been added to "Production of hydrocarbons" and "Petroleum refinement" (Activity/Commodity 5 and 11, respectively) because of negative return to capital for "Manufacture of coke products." "Heat supply" (Activity/Commodity 19) has been added to "Water supply" <20> and "Gas Supply" (Activity/Commodity 20 and 18, respectively) because of negative return to capital for "Heat supply." "Forestry" (Activity/Commodity 2) has been added to "Wood working, pulp and paper industry, publishing" (Activity/Commodity 9) since a separation of exports and imports of "Forestry" by destination and origin is not possible. Source: Authors. 105 Table A8-7. Structure of Ukraine's Economy (2001) Share in... Total exports Total imports Activities / Commodities Value added (by activity) (by commodity) Agriculture, hunting 0.16 0.05 0.01 Fishery 0.00 0.00 0.00 Mining of coal and peat 0.02 0.01 0.02 Production of non-energy materials 0.01 0.03 0.02 Food-processing industries 0.05 0.07 0.05 Textile and leather industry 0.01 0.04 0.06 Woodworking, pulp and paper industry, publishing 0.02 0.02 0.04 Petroleum refinement 0.03 0.05 0.31 Manufacture of chemicals, rubber and plastic products 0.02 0.08 0.11 Manufacture of other non-metallic products 0.01 0.01 0.01 Metallurgy and metal processing 0.04 0.30 0.08 Manufacture of machinery and equipment 0.05 0.12 0.18 Other 0.00 0.03 0.00 Electric energy 0.05 0.00 0.00 Public utilities 0.01 0.00 0.00 Construction 0.04 0.00 0.02 Trade 0.12 0.00 0.00 Hotels and restaurants 0.01 0.01 0.01 Transport 0.10 0.16 0.02 Post and telecommunications 0.03 0.00 0.00 Other services 0.22 0.02 0.06 Total 1.00 1.00 1.00 Source: National Accounts / Input Output table for 2001; authors' calculations. health-related goods and services grows. The sector and cost scenario projections constructed in the ear- most directly affected by HIV/AIDS is the health sec- lier chapters, the public expenditure for HIV/AIDS tor. The demands on the public health service rise treatment will increase within the range 0.2 percent, sharply with the spread of the epidemic. Ukraine has 2 percent, and 3.5 percent of the Ministry of Health started to make theses treatments available through budget, for optimistic, medium, and pessimistic sce- the public health service. According to the epidemic narios, respectively. 106 Table A8-8. Structure of Output and Input (2001) Output (in %) Input (in %) Domestic Intermediate Value Code sales Exports Total demand added Depreciation Total a01 b01 Agriculture, hunting 0.91 0.09 1.00 0.56 0.44 0.00 1.00 a03 b03 Fishery 0.83 0.17 1.00 0.65 0.30 0.05 1.00 a04 b04 Mining of coal and peat 0.95 0.05 1.00 0.70 0.29 0.01 1.00 a06 b06 Production of non-energy 0.54 0.46 1.00 0.75 0.25 0.00 1.00 materials a07 b07 Food-processing industries 0.83 0.17 1.00 0.82 0.18 0.00 1.00 a08 b08 Textile and leather industry 0.29 0.71 1.00 0.60 0.27 0.13 1.00 a09 b09 Woodworking, pulp and 0.65 0.35 1.00 0.64 0.36 0.00 1.00 paper industry, publishing a11 b11 Petroleum refinement 0.80 0.20 1.00 0.81 0.17 0.02 1.00 a12 b12 Manufacture of chemicals, 0.35 0.65 1.00 0.78 0.22 0.00 1.00 rubber and plastic products a13 b13 Manufacture of other 0.82 0.18 1.00 0.68 0.32 0.00 1.00 non-metallic products a14 b14 Metallurgy and metal 0.21 0.79 1.00 0.78 0.19 0.03 1.00 processing a15 b15 Manufacture of machinery 0.41 0.59 1.00 0.60 0.36 0.04 1.00 and equipment a16 b16 Other 0.44 0.56 1.00 0.87 0.13 0.00 1.00 a17 b17 Electric energy 0.98 0.02 1.00 0.53 0.47 0.00 1.00 a20 b20 Public utilities 1.00 0.00 1.00 0.75 0.25 0.00 1.00 a21 b21 Construction 0.99 0.01 1.00 0.60 0.40 0.00 1.00 a22 b22 Trade 1.00 0.00 1.00 0.43 0.57 0.00 1.00 a23 b23 Hotels and restaurants 0.43 0.57 1.00 0.56 0.44 0.00 1.00 a24 b24 Transport 0.46 0.54 1.00 0.42 0.58 0.00 1.00 a25 b25 Post and telecommunications 0.94 0.06 1.00 0.33 0.67 0.00 1.00 a26 b26 Other services 0.97 0.03 1.00 0.41 0.59 0.00 1.00 Source: National Accounts/Input-Output table for 2001; own calculations. 107 Bibliography Ainsworth, M., L. Fransen, and M. Over, eds. 1998. ------. Forthcoming. "Behaviour Surveillance of Confronting AIDS: Evidence from the Commercial Sex Worker Populations as a Developing World. Selected background papers Component of the Second Generation HIV for the World Bank Research Report, "Confronting Surveillance." International HIV/AIDS Alliance in AIDS: Public Priorities in a Global Epidemic," Ukraine, Kiev, in Ukrainian. European Commission and The World Bank. Bailey, N. T. J. 1988. "Simplified Modelling of the Amdzhadin, L., K. Kaschenkova, T. Konoplytska, O. Population Dynamics of HIV/AIDS." Journal of Lysenko, A. Marusov, Y. Privalov, Y. Sayenko and the Royal Statistical Society, Series A (Statistics O. Trofimenko. Forthcoming. "Behaviour in Society) 151(1): 31-34. Surveillance of Men Having Sex with Men Barnett, T., A. Whiteside, O. Balakireva, A. Populations as a Component of the Second Scherbinskaya, et al. 2001. "The Social and Generation HIV Surveillance." International Economic Impact of HIV and AIDS in Ukraine: A HIV/AIDS Alliance in Ukraine, Kiev, in Ukrainian. Re-Study." The British Council, Kiev, Ukraine. Anderson, R. M. 1988. "The Epidemiology of HIV Becker, N. G., and L. R. Egerton. 1994. "A Infection: Variable Incubation Plus Infectious Transmission Model for HIV with Application to Periods and Heterogeneity in Sexual Activity." the Australian Epidemic." Mathematical Journal of the Royal Statistical Society, Series Biosciences 119(2): 205-24. A (Statistics in Society) 151(1): 66-98. Becker, N. G., and I. C. Marschner. 1993. "A Method Armington, P. 1969. "A Theory of Demand for for Estimating the Age-Specific Relative Risk of Products Distinguished by Place of Production." HIV Infection from AIDS Incidence Data." International Monetary Fund Staff Paper 16, Biometrika 80(1): 165-78. Washington, DC. Becker, N. G., and M. Motika. 1993. "Smoothed Non- Arndt, C., and J. D. Lewis. 2001. "The HIV/AIDS parametric Back-Projection of AIDS Incidence Pandemic in South Africa: Sectoral Impacts and Data with Adjustment for Therapy." Mathematical Unemployment." Journal of International Biosciences 118(1): 1-3. Development 13(4): 427-49. Becker, N. G., L. Watson, and J. Carlin. 1991. "A Artyukh, O. R., O. M. Balakireva, L. V. Bochkova, Y. P. Method of Non-parametric Back-Projection and Its Galilch, Y. M. Galustyan, D. M. Dikova-Favorska, Application to AIDS Data." Statistics in O. G. Zlobina, N. M. Levchuk, V. P. Lyutyi, V. A. Medicine 10: 1527-42. Martsynovska, T. P. Mykytyuk, V. F. Morozov, O. M. Petrovskyi, T. S. Shamota and O. O. Yaremenko. Bell, C., S. Devarajan, and H. Hersbach (2004). 2005a. "Behaviour Surveillance of Injecting Drug "Thinking about the long-run economic costs of User Populations as a Component of the Second AIDS," in The Macroeconomics of HIV/AIDS, M. Generation HIV Surveillance." International Haacker (ed). Washington, DC, International HIV/AIDS Alliance in Ukraine, Kiev, 68 pages, in Monetary Fund: 96-144. Ukrainian. Bhargava, A., et al. 2001. "Modeling the Effects of Health on Economic Growth." Journal of Health Economics 20(3): 423-40. 108 Bloom, D. E., and J. D. Sachs. 1998. "Geography, DeBell, D., and R. Carter. 2005. "Impact of Transition Demography, and Economic Growth in Africa." on Public Health in Ukraine: Case Study of the Brookings Papers on Economic Activity 0(2): HIV/AIDS Epidemic." British Medical Journal 207-73. 331: 216-19. Boldsen, J. L., J. L. Jensen, J. Sogaard, and M. Dee, P., K. Hanslow, and T. Phamduc. 2003. Sorensen. 1988. "On the Incubation Time "Measuring the Costs of Barriers to Trade in Distribution and the Danish AIDS Data." Journal Services." In Trade in Services in the Asia- of the Royal Statistical Society, Series A Pacific Region, eds. Takotoshi Ito and Anne (Statistics in Society) 151(1): 42-43. Krueger. Chicago: University of Chicago Press. Brainerd, E., and D. M. Cutler. 2005. "Autopsy of an Derzhkomstat (Ukrainian Statistics Committee). Empire: Understanding Mortality in Russia and 2004a. Statistical Yearbook of Ukraine, 2003. the Former Soviet Union." Journal of Economic Derzhkomstat Ukrainy, Kiev, in Ukrainian. Perspectives 19(1): 107-30. ------. 2004b. National Accounts of Ukraine, 2003. Brown, D., and R. Stern. 2001. "Measurement and Derzhkomstat Ukrainy, Kiev, in Ukrainian. Modeling of the Economic Effects of Trade and Derzhkomstat (Ukrainian Statistics Committee). Investment Barriers in Services." Review of 2005. Population of Ukraine, 2004: Demographic International Economics 9(2): 262-86. Yearbook. Derzhkomstat Ukrainy, Kiev, in Castillo-Chávez, C. 1989. Mathematical and Ukrainian. Statistical Approaches to AIDS Epidemiology. Devarajan, S., and S. Robinson. 2002. "The Influence New York: Springer Verlag. of Computable General Equilibrium Models on ------. 2002. Mathematical Approaches for Policy." TMD Discussion Paper No. 98. Emerging and Reemerging Infectious Diseases: International Food Policy Research Institute Models, Methods, and Theory. New York: (IFPRI), The World Bank, Washington, DC. Springer. DHS (Department of Human Services). 1999. Cuddington, J. T. 1993a. "Further Results on the Victorian Burden of Disease Study: Morbidity. Macroeconomic Effects of AIDS: The Dualistic, Public Health Division, Victorian Government Labour-surplus Economy." World Bank Economic Department of Human Services, Melbourne, Review 7(3): 403-17. Victoria. ------. 1993b. "Modeling the Macroeconomic Ethier, W. J. 1982. "National and International Effects of AIDS, with an Application to Tanzania." Returns to Scale in the Modern Theory of World Bank Economic Review 7(2): 173-89. International Trade." American Economic Review 72(2): 389-405. Cuddington, J. T., and J. D. Hancock. 1994. "Assessing the Impact of AIDS on the Growth Feshbach, M., and C. Galvin. 2005. "HIV/AIDS in Path of the Malawian Economy." Journal of Ukraine-An Analysis of Statistics." Washington, Development Economics 43(2): 363-68. DC, Woodrow Wilson International Center for Scholars: 34. Cuddington, J. T., J. D. Hancock, and C. A. Rogers. 1994. "A Dynamic Aggregate Model of the AIDS FitzSimons, D., V. Hardy, and K. Tolley. 1995. The Epidemic with Possible Policy Interventions." Economic and Social Impact of AIDS in Journal of Policy Modeling 16(5): 473-96. Europe. London: Cassell. de Melo, J. 1988. "Computable General Equilibrium Francois, J.F., and C.R. Shiells. 1994. Modeling Models for Trade Policy Analysis in Developing Trade Policy: Applied General Equilibrium Countries: A Survey." Journal of Policy Modeling Models of North American Free Trade, Vol. 10. Cambridge: Cambridge University Press. 109 Freire, S. (2004). "Impact of HIV/AIDS on saving Helpman, E., and P. Krugman. 1985. Market behaviour in South Africa". Forum paper for the Structure and Foreign Trade. Cambridge, MA: conference African development and poverty MIT Press. reduction: the macro-micro linkage, Lord Ianchovichina, E., and W. Martin. 2001. "Trade Charles Hotel, Somerset West, South Africa, 13-15 Liberalization in China's Accession to the WTO." October. Journal of Economic Integration 16: 421-45. German Advisory Group. 2004. "Ukrainian Steam IDSS (Institute for Demography and Social Studies). Coal: Not Competitive or Just Mismanaged?" 2003. Demographic Crisis in Ukraine: Causes Advisory Paper T19, Kiev. Available at and Consequences. Institute for Demography and http://www.ier.kiev.ua/English/advis_eng.cgi. Social Studies, National Academy of Sciences, Grossman, G. M., and E. Helpman. 1991. "Innovation Kiev, Ukraine. State Statistics Committee and Growth in the Global Economy." Cambridge (Derzhkomstat), in Ukrainian. MA: MIT Press. ------. 2005. Concept of Demographic Haacker, M. 2002a. "The Economic Consequences of Development, Ukraine, 2005-2015. HIV/AIDS in Southern Africa." IMF Working Demography and Social Economics. Vol. 1. Paper W/02/38: 41-95. Institute for Demography and Social Studies, National Academy of Sciences, Kiev, Ukraine, in ------. 2002b. "Modeling the Macroeconomic Impact Ukrainian. of HIV/AIDS." IMF Working Paper W/02/195: 41- 95. ILO (International Labor Organization). 2004. HIV/AIDS and Work: Global Estimates, Impact ------. 2004a. "HIV/AIDS: Impact on the Social and Response. Geneva: ILO: 101. Fabric and the Economy." In The Macroeconomics of HIV/AIDS, ed. M. Haacker. Isham, V. 1988. "Mathematical Modelling of the Washington, DC: International Monetary Fund. Transmission Dynamics of HIV Infection and ------. ed. 2004b. The Macroeconomics of AIDS: A Review." Journal of the Royal HIV/AIDS. Washington, DC: International Statistical Society. Series A (Statistics in Society) Monetary Fund. 151(1): 5-30 and 44-49. Harris, R. 1984. "Applied General Equilibrium IUSSP (International Union for the Scientific Study Analysis of Small Open Economies with Scale of Population). 1993. `AIDS Impact and Economies and Imperfect Competition." The Intervention in the Developing World: The American Economic Review 74(5): 1016-32. Contribution of Demography and Social Science." Papers communications. Leige, Belgium: IUSSP. Harrison, G. W., T. F. Rutherford, and D. Tarr. 1997a. "Economic Implications for Turkey of a Customs Jager, J. C., E. J. Ruitenberg, and Commission of the Union with the European Union." European European Communities, Working Party on AIDS. Economic Review 41(3-5): 861-70, April. 1988. Statistical Analysis and Mathematical Modelling of AIDS. Oxford University Press. ------. 1997b. "Quantifying the Uruguay Round." Economic Journal 107: 1405-30, September. Jensen, J., T. Rutherford, and D. Tarr. 2003. "Economy-wide and Sector Effects of Russia's Hartunian, N. S., C. N. Smart, and M. Thompson. Accession to the WTO." Mimeo. 1981. The Incidence and Economic Costs of Major Health Impairments: A Comparative Johnson, A. M. 1988. "Social and Behavioural Analysis of Cancer, Motor Vehicle Injuries, Aspects of the HIV Epidemic-A Review." Journal Coronary Heart Disease, and Stroke. Lexington, of the Royal Statistical Society. Series A Mass: Lexington Books. (Statistics in Society) 151(1): 99-119. Healy, M. J. R., and H. E. Tillett. 1988. "Short-Term Kaplan, E. H., and M. L. Brandeau. 1994. Modeling Extrapolation of the AIDS Epidemic." Journal of the AIDS Epidemic: Planning, Policy, and the Royal Statistical Society. Series A (Statistics Prediction. New York, NY: Raven Press. in Society) 151(1): 50-65. 110 Lewis, M. A. 1989. AIDS in Developing Countries: Over, M. 2002. "The Macroeconomic Impact of Cost Issues and Policy Tradeoffs. Washington, HIV/AIDS in Sub-Saharan Africa." African DC: Urban Institute Press; distributed by Technical Working Paper Number 3, Population University Press of America. Health and Nutrition Division, Africa Technical Department, World Bank. Markusen, J.R., T. Rutherford, and D. Tarr. 2002. "Foreign Direct Investment in Services and the Pavel, F. 2001. Success and Failure of Post- Domestic Market for Expertise." Policy and Communist Transition. Achen: Shaker Verlag. Research Working Paper 2413. Washington, DC: Pyatt, G., and J.I. Round. 1985. Social Accounting The World Bank. Matrices: A Basis for Planning. Washington, DC. Mathers, C., T. Vos, and C. Stephenson. 1999. The Robalino, D. A., C. Jenkins, and K. E. Maroufi. 2002. Burden of Disease and Injury in Australia. "The Risks and Macroeconomic Impact of Catatalog Number PHE 18. Australian Institute for HIV/AIDS in the Middle East and North Africa: Health and Welfare, Canberra. Why Waiting to Intervene Can Be Costly." Policy Mavrov, G., and G. Bondarenko 2002. "The Evolution Research Working Paper Series: 2874, World Bank, of Sexually Transmitted Infections in the Ukraine," Washington, DC. Sexually Transmitted Infections 78: 219-21. Robalino, D. A., A. Voetberg, and O. Picazo. 2002. Ministry of Economy and European Integration. "The Macroeconomic Impacts of AIDS in Kenya: 2004. Monitoring of Macroeconomic and Estimating Optimal Reduction Targets for the Sectoral Indicators. Edition 2 (58): 5-88, Kiev, in HIV/AIDS Incidence Rate." Journal of Policy Ukrainian. Modeling 24(2): 195-218. Murray, C. J. L., and A. K. Acharya. 1997. Ruehl, C., V. Pokrovsky and V. Vinogradov. 2002. "Understanding DALYs." Journal of Health "The Economic Consequences of HIV in Russia." Economics 16(6): 703-30. World Bank, Moscow, Russia. Murray, C. J. L., et al. 2000. "Development of WHO Rutherford, T.F. 1999. "Applied General Equilibrium Guidelines on Generalized Cost-Effectiveness Modeling with MPSGE as a GAMS Subsystem: An Analysis." Health Economics 9(3): 235-51. Overview of the Modeling Framework and Syntax." Computational Economics 14(1/2): Murray, C. J. L., and A. D. Lopez, 2000. "Progress and October. Directions in Refining the Global Burden of Disease Approach: A Response to Williams." Rutherford, T., and S. Paltsev. 1999. "From an Input- Health Economics 9(1): 69-82. Output Table to a General Equilibrium Model: Assessing the Excess Burden of Indirect Taxes in Murray, C. J. L., A. D. Lopez, and WHO (World Health Russia." Draft, University of Colorado. Organization). 1994. Global Comparative Assessments in the Health Sector: Disease Schwartlander, B. A., K. A. B. Stanecki, T. C. Brown, Burden, Expenditures and Intervention P. O. B. Way, R. D. Monasch, J. E. Chin, D. F. Packages. Geneva: WHO. Tarantola, and N. A. Walker. 1999. "Country-specif- ic Estimates and Models of HIV and AIDS: NBU (National Bank of Ukraine). 2004. NBU Methods and Limitations." AIDS 13(17): 2445-58. Bulletin 1: 70. Sharp, S. 2004a. "Informing Policy through Modeling: OECD (Organisation for Economic Co-operation and A Case Study of the Socio-economic Implications Development). 2003. "Achieving Ukraine's of AIDS in Russia." Presentation, United Nations Agricultural Potential: Stimulating Agricultural Development Programme (UNDP). Growth and Improving Rural Life." Executive Summary, Chapters 1-9, and Annexes 1-5. Working ------. 2004b. "Modelling the Macroeconomic Party on Agricultural Policies and Markets. Implications of a Generalised AIDS Epidemic in the Russian Federation." In Reversing the Epidemic: Facts and Policy Options, ed. Bratislava, UNDP: 85-90. 111 Shoven, J.B., and J. Whalley. 1992. Applying General UNDP (United Nations Development Program). 2004. Equilibrium. Cambridge University Press. "Reversing the Epidemic: Facts and Policy Options." UNDP, New York. SIFYA (State Institute on Family and Youth Affairs. 2004. Family in Independent Ukraine (1991- WHO (World Health Organization) Commission on 2003). Official Report on Family Issues in Macroeconomics and Health, Ed. 2001. Ukraine. Kiev, in Ukrainian. "Macroeconomics and Health: Investing in Health for Economic Development." WHO. Stouthard, M., M. Essink-Bot, G. Bonsel, J. Barendregt, and P. Kramers. 1997. "Disability Wilkie, A. D. 1988. "An Actuarial Model for AIDS." Weights for Diseases in the Netherlands." Journal of the Royal Statistical Society. Series Department of Public Health, Erasmus University, A (Statistics in Society) 151(1): 35-39. Rotterdam. The World Bank. 2003. "The Major Problems of UISS (Ukrainian Institute of Social Studies). 2002. Development of the Coal Sector and the Donbas "Gender Parity in the Emerging Modern Ukrainian Region." Concise Report, Kiev. Society." Kiev, in Ukrainian. ------. 2004. "Ukraine: Building Foundations for UNAIDS (Joint United Nations Programme on Sustainable Growth." A Country Economic HIV/AIDS). 2002. "Improved Methods and Memorandum, Report No. 30928-UA, Poverty Assumptions for Estimation of the HIV/AIDS Reduction and Economic Management Unit, Epidemic and Its Impact: Recommendations of Europe and Central Asia Region, The World Bank, the UNAIDS Reference Group on Estimates, Washington, DC, 98 pages. http://www-wds.world- Modelling and Projections: The UNAIDS bank.org/servlet/WDSContentServer/WDSP/IB/200 Reference Group on Estimates, Modelling and 5/01/11/000160016_20050111155332/Rendered/PDF Projections [Report]." AIDS 16(9): W1-W14. /309280UA.pdf 112 U kraine, the second largest country in Europe, has achieved significant progress in macroeconomic stabilization and microeconomic reform during its transition toward a market economy over the past fifteen years. However, the gains from its economic recovery are threatened by an HIV/AIDS epidemic that started in 1987 and accelerated dramatically in 1995 when the virus penetrated the subpopulation of injecting drug users. At present, Ukraine's HIV/AIDS epidemic is among the fastest growing in Europe, with officially registered new HIV cases having doubled over 2000-2004. Official data suggest that Ukraine's epidemic may be on the brink of the generalized epidemic phase: by the end of 2004, the share of heterosexual mode of transmission has increased to almost a third of new cases. The spread of HIV/AIDS is superimposed on the adverse demographic situation characterized by both depopulation and deteriorating health status. Socioeconomic Impact of HIV/AIDS in Ukraine was prompted by the need to assess the potential long- term impact of the rapidly growing HIV/AIDS epidemic in the country. It constructs a baseline demographic projection of the Ukrainian population for 1994-2014 and three epidemic scenarios for HIV/AIDS (medium, optimistic, and pessimistic) using low-high estimates of the size of and prevalence in most-at-risk populations. Model parameters include the rate of mother-to child transmission, the rate of progressions from HIV to AIDS, and availability of antiretroviral therapy. Based on the epidemic scenario, the study estimates the effect of HIV/AIDS on the labor force, both at the nation- al level and for the worst-affected regions. It also calculates direct budgetary impact due to disability, foregone revenue, and increased health care expenditure. Three macroeconomic models are constructed and applied to analyze macroeconomic costs of the epidemic in Ukraine: a simple growth model, a macroeconometric model, and a computable general equilibrium model. The latter two are multisectoral models evaluating differential effects of the epidemic on various sectors of economy. The tools to curb the epidemic and its consequences both in human toll and economic hardship are at hand. Socioeconomic Impact of HIV/AIDS in Ukraine sets out clearly the reasons why those tools must be implemented now.