WPS5994 Policy Research Working Paper 5994 Governance and Public Service Delivery in Europe and Central Asia Unofficial Payments, Utilization and Satisfaction Mame Fatou Diagne Dena Ringold Salman Zaidi The World Bank Europe and Central Asia Region Poverty Reduction and Economic Management Unit March 2012 Policy Research Working Paper 5994 Abstract Using data from the 2010 Life in Transition Survey, remains lower than in Western European comparator this paper examines the levels of citizens’ satisfaction countries. Although the Life in Transition Survey does with public service delivery in Europe and Central not provide specific objective measures of service delivery Asia and identifies some factors that may help explain quality and efficiency, the data provide three important variation in utilization and levels of satisfaction with clues that may help explain why satisfaction is lower in service delivery. It finds satisfaction with public service transition countries than in western comparators: (i) delivery in Europe and Central Asia to be relatively relatively higher utilization of public services in Eastern high, and, despite the adverse economic and social European and Central Asian countries, (ii) relatively impact of the recent global economic crisis, to have risen higher reported prevalence of unofficial payments, and since 2006 in most countries in the region. However, (iii) relatively underdeveloped mechanisms for grievance the level of satisfaction with public service delivery in redress. Eastern European and Central Asian countries in 2010 This paper is a product of the Poverty Reduction and Economic Management Unit, Europe and Central Asia Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank. org. The author may be contacted at szaidi5@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Governance and Public Service Delivery in Europe and Central Asia: Unofficial Payments, Utilization and Satisfaction Mame Fatou Diagne, Dena Ringold, and Salman Zaidi* JEL Classification Codes: H40, I00, K42, I18, I28 Keywords: Public service delivery, health, education, irregular payments World Bank Sector: POV * World Bank, Washington DC. Correspondence: Salman Zaidi, World Bank, 1818 H Street NW, Washington DC 20433, USA. E. mail: szaidi5@worldbank.org. Phone: (202) 473 1976. Fax: (202) 522 2753. Martin Cumpa and Santhosh Srinivasan provided invaluable research assistance. The authors are grateful to Paolo Belli, Rekha Menon, Lars Sondergaard, and Owen Smith for their suggestions and comments on previous versions of the paper. This paper has also benefited from the comments of numerous colleagues at the World Bank. The opinions and views presented here are those of the authors and should not be attributed to the Executive Directors of The World Bank. 1 1. INTRODUCTION There is growing recognition among policy-makers that measuring inputs and outputs alone is not enough to understand how service delivery works in practice. Rather, the traditions and institutions under which authority is exercised for the common good – i.e. ―governance‖ – also play a crucial role. Better governance is an essential ingredient of reforms targeted at improving service delivery outcomes (e.g. better health status, enhanced learning outcomes, etc.).2 The 2004 World Development Report Making Services Work for Poor People (World Bank, 2003) defined a framework for analyzing the accountability relationships between a triangle of policy makers, providers and citizens (Figure 1). Within this framework, policies can either be implemented through a ―long route of accountability,‖ whereby citizens elect policy-makers who in turn influence service delivery through providers, or a ―short route of accountability,‖ through which citizens may directly influence, participate in, and/or supervise service delivery by providers. In order for both accountability channels to work effectively, citizens‘ levels of satisfaction with the quality and efficiency of public services are very important feedback mechanisms. Figure 1: Accountability Relationships Source: World Bank, 2003. The importance of regular citizen feedback on the quality and efficiency of public service delivery has particular resonance in transition countries, where the relationships between policy-makers, service providers and citizens have transformed dramatically over the past two decades of political and economic transition (and EU accession for some countries). Countries have undertaken wide- ranging service delivery reforms that have influenced institutional relationships, altered incentives for service providers at the institutional and individual levels, and changed the ways in which citizens participate in and experience service delivery. All of these developments have led to changes in the 2 For the purposes of this paper, we define better governance as improvements in incentives for performance and institutional arrangements for holding service providers accountable. 2 ways in which people interact with the state, as well as their perceptions and levels of trust. These are themes which LITS2 is uniquely positioned to address: the survey includes an expanded module on governance and service delivery which enables new analysis on the links between the ways in which people use, experience and interact with public services, their perceptions and trust in providers and institutions, and the quality of service delivery itself. There are obviously a number of caveats in the interpretation of satisfaction data. The presence of externalities (e.g. additional social benefits of education, public health, etc.), may mean that users‘ satisfaction does not capture the full extent of desired social outcomes. Second, even if individuals and society share the same goal – such as more rapid student learning – some persons may have been conditioned to have low expectations, leading them to report high satisfaction despite poor service delivery outcomes. A third reason for caution is information asymmetries: this is particularly evident in health services, where providers have technical information and expertise that patients lack and cannot therefore evaluate properly (Fiszbein, Rogers and Ringold, 2011; Das and Hammer, 2007). Still, subjective data on peoples‘ experiences and perceptions of service delivery quality can help complement more objective measures of service delivery (e.g. from facilities surveys) to help policy makers and development practitioners track progress of country programs and strategies. Using data from the 2010 Life in Transition Survey Round 2 (LITS2), this paper examines the levels of citizens‘ satisfaction with public service delivery in Europe and Central Asia and identifies some factors that may help explain how levels of satisfaction with service delivery vary across different population sub-groups and change over time. Section 2 describes the coverage of governance and service delivery in LITS2 and presents the key findings on the prevalence of unofficial payments, and utilization of and satisfaction with public services. Section 3 analyzes differences across individuals and across countries (over time) in utilization and satisfaction with public services. We conclude by discussing a few measures that governments can implement to further increase citizens‘ satisfaction with public service delivery. 2. UTILIZATION AND SATISFACTION WITH PUBLIC SERVICE DELIVERY 2.1. The data: Government performance and service delivery in the 2010 LITS The Life in Transition Survey was conducted jointly by the European Bank for Reconstruction and Development and the World Bank in 2006 and 2010. Between 1,000 and 1,500 households were interviewed in every country. In total, in 2010, almost 39,000 households were surveyed in 34 countries of Europe and Central Asia to assess public attitudes, well-being and the impacts of economic and political change. The governance and service delivery module focuses on eight different public services or interactions: traffic police, official documents (such as passport or birth certificate), civil courts, primary or secondary education, vocational education, the public health system, unemployment benefits and other social security benefits. For each service, respondents were asked their 3 perceptions of the prevalence of unofficial payments3 and then about their household‘s utilization, satisfaction and actual unofficial payments. Utilization: Respondents were asked if any household member had interacted with, or used, each of the public services. Satisfaction with service delivery: All respondents who indicated that a household member had used the service during the past 12 months were asked if they were satisfied with the quality and efficiency of the service interaction4. Satisfaction data can be a proxy for measuring actual quality of services, as well as an indicator of the extent to which services are responsive to the needs and preferences of clients. They can also help to assess the effects of service delivery reforms (see section 3.2). Unofficial payments: Respondents were asked if they had to make unofficial payments and why – specifically whether the payment was requested, expected, offered or given as a gift – allowing for a more detailed analysis of the incidence and causes of informal payments than is usually possible from household surveys. For education and public services, the 2010 LITS survey provides more in-depth information on perceptions of service quality and grievance redress mechanisms. Regarding education, the survey inquired about any lack of textbooks and supplies, poor teaching, teacher absenteeism, overcrowded classrooms and poor conditions of facilities. Similarly for health, it asked about doctor absenteeism, treatment by staff, availability of medicines, waiting times and cleanliness of facilities. The answers to these questions provide a snapshot of people‘s experiences with services and can act as another measure of service quality. Another new feature of the 2010 LITS is a set of questions related to grievance redress mechanisms in health and education. The survey asks whether people know where to file a complaint if they were unsatisfied with education and health services, whether they filed a complaint, received a response, or were satisfied with the response. Finally, respondents were asked to rate the overall performance of their national, regional, and local governments using a five-point scale (ranging from 1: very bad to 5: very good), and whether they felt the overall performance of these three administrative tiers had worsened, stayed the same or improved in the previous three years. 2.2. Descriptive statistics Utilization. The proportion of respondents who interacted with service providers varied considerably according to the type of service (see figure 2). By far the most often accessed service was the public health system (70 percent in the previous 12 months). Utilization rates for other services were much lower. The next highest usage rates were for primary/secondary public education (23 percent), interaction with traffic police (20 percent) and requests for official documents (14 percent). 3 “In your opinion, how often is it necessary for people like you to have to make unofficial payments/gifts in these situations�, with responses recorded on a five-point scale: 1= never, 2=seldom, 3=sometimes, 4=usually and 5=always. 4 Respondents were asked: “How satisfied were you with the quality and the efficiency of the service/interaction?� and responses were recorded on a five-point scale: 1=very dissatisfied, 2=dissatisfied, 3=indifferent, 4=satisfied and 5=very satisfied. 4 Figure 2 Utilization of public services Source: LiTS2 (2010). Satisfaction Of all the public services covered in the LITS2 questionnaire, the public education system (both primary/secondary and vocational) received the most favorable ratings: over 70 percent of respondents using these public services were either satisfied or very satisfied with the quality and efficiency of the services (Figure 3). Similarly, 60 percent or more were satisfied when making requests for official documents, social security, or receiving medical treatment in the public health system. By contrast, civil courts and traffic police received the lowest service satisfaction ratings, with about 40 percent of respondents in Europe and Central Asia dissatisfied with the quality and efficiency of the service interaction. Figure 3: Satisfaction with service delivery Source: LiTS2 (2010). Among transition countries, overall satisfaction levels tended to be somewhat higher among the New EU member states and Turkey (see Table 1), and generally lower among the Commonwealth of Independent States (CIS), although there was considerable variation evident within individual 5 sub-regions—for example, between Estonia and Romania, Georgia and Azerbaijan or Russia and Ukraine. Interestingly, some relatively poor transition countries (such as Georgia and Moldova) rate quite well for overall satisfaction with the quality and efficiency of public services. Another noteworthy finding is that while satisfaction with service delivery in transition countries is fairly high overall, it is generally considerably lower than prevailing levels in the western European comparator countries. Possible reasons for these differences at country level are discussed in section 3. For several public services there is a negative correlation at the country level between usage rates and the levels of satisfaction with public service delivery (Figure 4)—i.e. countries where a relatively high proportion of respondents report using public services during the past 12 months tend to have lower satisfaction rates than countries where the reverse is true. It could be that high utilization rates place a heavy burden on availability (in terms, for example, of staff time, medicines or teaching materials), thereby reducing the capacity to provide high-quality services and compromising perceptions of quality. This gap between utilization and satisfaction could signal deficits in the quality and availability of certain types of services. For example, health reforms in many transition countries aim to strengthen the supply of primary care and referral systems to higher levels of care. Where these systems are not in place, there may be over-utilization of hospital or emergency care at the expense of more efficient preventative and public health services. In the case of education, dissatisfaction may also reflect concerns about quality and the ability of schools to prepare students for the labor market. Figure 4: Satisfaction vs. utilization rates by public service type Source: LiTS2 (2010). It is important to note that reported satisfaction may be disconnected from actual quality of services if people do not have a basis for comparison, or information, about what level of quality and service they should be expecting. For example, the high satisfaction rates in LiTS2 contrast strikingly with recent findings on education quality outcomes from the Programme for International Student Assessment (PISA) survey. While 58 percent of respondents in the Kyrgyz Republic report 6 satisfaction with education in LiTS2, the 2009 PISA survey found that 83 percent of 15-year olds were functionally illiterate (they scored below PISA level 2). Complaints and grievance redress About 67 percent of respondents who had used health services during the previous 12 months reported at least one problem with service delivery, and for education services the figure was 50 percent. The main problems reported for health were long waiting times, lack of drugs and payment for services that should be free (Figure 5); in the case of education, the main problems were lack of textbooks/supplies, payments for services that should be free, and poor conditions of facilities. Figure 5: Problems with public health and education provision in the previous 12 months reported by respondents Public health clinic/hospital Public education Long waiting times No textbooks or other supplies P ayments required for services that should No drugs available be free P ayments required for services that should Facilities in poor condition be free Treated disrespectfully by staff P oor teaching Frequent / unjustified absence of doctors Overcrowded classrooms Facilities not clean Frequent / unjustified absence of teachers 0 10 20 30 40 0 5 10 15 20 25 Per cent of respondents Per cent of respondents Source: LiTS2 (2010). Grievance redress or complaints mechanisms refer to the institutions and channels that people can use to express their preferences and hold providers to account5. There are a number of specific objectives associated with the use of complaints handling in the public sector: promoting accountability in the delivery of services; measuring the quality of service provision and the effectiveness of policy; and harnessing experience to improve service delivery (Lister et al., 2008). Grievance redress mechanisms have long traditions in Western Europe, but they are relatively underdeveloped in transition countries6. While 16 percent and 11 percent of respondents, respectively, reported that they had filed complaints for education and health in the comparator countries, only 5 percent did so in transition countries for either service. Some countries in the region have begun introducing complaints-handling mechanisms to improve service delivery. For 5 Complaints handling systems take diverse forms, including customer complaints procedures; administrative appeals and tribunals systems; independent complaints handlers or ombudsmen; and judicial review (and other forms of legal action). 6 Sweden first installed an independent ombudsman function in the late 19th century. More recently, the adoption of complaints handling systems spread in the countries of the Organisation for Economic Co-operation and Development in the 1990s as governments looked increasingly to private sector practice to improve the standards of public service provision. Studies on the private sector highlighted that the presence of a complaints system in businesses was associated with better performance (Johnson and Mehra, 2002); 7 example, as part of its health reform efforts, the Turkish government introduced Patients‘ Rights Units in all hospitals which allow citizens to submit complaints related to access to care. High numbers of complaints may not necessarily reflect poor quality of services, but rather accessibility of complaints systems and greater interactions between providers and citizens. Fewer than half of transition respondents know where to file a complaint about health and education services (the highest level of awareness being in Turkey and the lowest in Central Asia), compared to 60 percent for education and 50 percent for health in western Europe. Most people in the transition region who filed a complaint did get a response (although to a lesser extent in Central Asia and south-eastern Europe), and about two-thirds were satisfied with the response that they received (and more so in Russia). Unofficial payments When LiTS2 respondents were asked how often it is necessary for people to make unofficial payments to access public services, a large majority said that such payments were never needed. Nevertheless, the proportion reporting that payments were usually or always needed was notably higher for the public health system than for other services (see Figure 6). The reported prevalence of unofficial payments in the transition countries is higher than in most western comparators. Within the transition region, it is generally lower in the EU countries and south-eastern Europe and higher in Central Asia and the south Caucasus, although once again there is variation within sub-regions (see Table 2). Reported prevalence of unofficial payments in ECA countries is higher than in most Western European comparator countries. LiTS2 respondents were also asked why they had made unofficial payments for services that should have been free: 43 percent of respondents in western comparator countries said they had made payments to express their gratitude to service providers, compared to only 19 percent in transition countries (see Figure 7); in contrast, 41 percent in the western comparators said they made the payments either because they were ―asked to pay‖ or they ―were not asked to pay, but knew that an unofficial payment was expected‖, compared to 60 percent in transition countries. Figure 6 Figure 7 Prevalence of unofficial payments Reason for making unofficial payments 100 19 To express my gratitude 80 43 21 To get things done 60 quicker/better Per cent 15 31 Not asked, but knew 40 13 payment was expected 20 I was asked to pay 29 28 0 Transition countries Western comparators Source: LiTS2 (2010). 8 3. SATISFACTION WITH SERVICE DELIVERY AND GOVERNMENT PERFORMANCE 3.1. Utilization and satisfaction with public services We analyze satisfaction with public health and public primary/secondary education services (the two most accessed services), using probit regressions (where the dependent variable is binary, taking value 1 if the user is satisfied or very satisfied, and 0 otherwise) and ordered probit regressions (where the dependent variable is the satisfaction level, taking values 1 [very dissatisfied] to 5 [very satisfied]). Taking into account the fact that survey responses regarding satisfaction only apply to usage within the previous 12 months, we use a two-stage Heckman sample selection correction. The results are presented in Tables 5 and 6. Public health Relatively wealthier households, those with more children and/or elderly people and those in urban areas are more likely to access public health services than poorer ones which have fewer children or elderly members, or are located in rural areas. Reported levels of satisfaction with public health service provision are positively associated with self-assessed good health, satisfaction with life and greater age (in so far as older respondents were more likely to approve the quality of services received). Richer and better-educated respondents were less likely to be satisfied with the quality and efficiency of treatment received. Personal experience of specific problems in local public health provision (as identified in Figure 5) also has a negative impact on satisfaction. The issue associated with the largest reduction in satisfaction levels is "payments required for services that should be free", followed by being "treated disrespectfully by staff", "long waiting times", "frequent and unjustified absence of doctors" and having "no drugs available" (see Table 5). The perception that unofficial payments are necessary for access is an important factor causing dissatisfaction. Users of the public health system who felt that unofficial payments are always needed are more than four times more likely to report being highly dissatisfied with services received than those who that felt that such payments are never needed. Public education Differences in utilization rates across urban, rural and metropolitan regions are not statistically significant. The reported level of satisfaction with primary and secondary education is positively associated with satisfaction with life; however, other factors such as the socio-economic background of the household and education level of the respondent do not appear to have any discernible impact on satisfaction levels. Personal experience of specific problems in the education system has a negative impact on satisfaction. ―Poor teaching‖ is associated with the largest reduction in satisfaction levels, followed by being "frequent and unjustified absence of teachers‖, ―crowded classrooms‖ and ―payments required for services that should be provided free‖ (see Table 6). Reported satisfaction levels are significantly higher among those respondents who say unofficial payments are never needed. Primary and secondary school users who felt that unofficial payments are always needed were nearly six times more likely to report being highly dissatisfied with the service delivery than those who considered such payments unnecessary. 9 3.2. Changes in satisfaction levels and prevalence of unofficial payments The LiTS2 data show encouraging progress in recent years with regard to public perceptions of the quality and efficiency of service delivery. In most countries, prevailing levels of satisfaction with the various public services covered in the LiTS questionnaire are higher in 2010 compared to 2006, especially regarding requests for official documents, unemployment benefits and other social security payments (see Table 3). Moldova and Russia stand out in this respect (as, to a lesser extent, do Turkey, Tajikistan and then Estonia). Azerbaijan is the only country for which satisfaction with service delivery for most public services in 2010 is lower than in 2006. In the case of public health, for instance, between 2006 and 2010 nearly all transition countries show increases in the proportion of respondents who were either satisfied or highly satisfied with the quality and efficiency of services received (see Figure 8).7 Figure 8: Satisfaction with public health services (current and changes over time) Source: LiTS2 (2010). Turkey, Moldova and Tajikistan are the three countries where satisfaction with public health services increased most. These changes in satisfaction levels can be explained by recent reforms.  In Moldova, recent health sector reforms have included: the creation of an independent mandatory social health insurance structure; increased hospital autonomy; the separation of primary and secondary care financing; steps towards performance-based contracting; and the development of clear accreditation and quality standards. Health provision has recovered to pre-transition levels and spending increased to 6.4 percent of GDP in 2009; however, anecdotal evidence suggests that the introduction of insurance has not replaced unofficial payments as a means of ensuring access to care. 7 It is important to note, that data on satisfaction are not fully comparable between the 2006 and 2010 LiTS surveys. The 2006 survey asked respondents if “you personally� used a service (and, conditional on use, whether you were satisfied); while the 2010 survey asked if “you or anyone in your household� used the service. 10  In Tajikistan, the government‘s 2005 health financing strategy aimed to improve equity, efficiency and cost-effectiveness of the health system by increasing public funding (in particular to primary health care) and introducing a basic-benefits-package (BBP). The BBP provides free services for vulnerable population groups and provides a legal framework for developing the policy for co-payments by patients for selected health services in hospitals.  Finally, in Turkey, the Turkish government launched a major reform program in 2003 to make the health system more effective by improving governance, efficiency, user- and provider-satisfaction and long-term sustainability. The main elements of the program were to establish a single purchaser in the health system, make the public sector health services delivery network autonomous and strengthen human resources management and information systems. The program has had important effects on access to care, especially for the poor. In addition to improved health insurance coverage for the poor, productivity of health personnel and availability of services have increased. There has also been a rise in the immunization of under-five-year-old children, the use of ante-natal services by pregnant women and the overall utilization of health services (Chakrabourty, 2009). With respect to perceptions of the need for unofficial payments when interacting with public officials, the LiTS2 data give a mixed picture on their evolution (see Table 4). In Albania, Bulgaria, Macedonia and Russia, fewer respondents in 2010 felt that unofficial payments are usually, or always, needed compared to 2006, while the opposite was true in Azerbaijan, the Kyrgyz Republic, Moldova and Turkey. 3.3. Perceptions of overall government performance Average perceptions of improvements or worsening in overall government performance appear to have been strongly colored by the severity of impact of the recent global economic crisis (Figure 9). When asked to assess the overall performance of the government, respondents in Tajikistan, Georgia, and Azerbaijan rated the performance of their respective national governments much more highly than did respondents in Romania, Croatia, and Lithuania. At the individual level, we estimate a probit regression where the dependent variable is the rating of government performance (a binary variable equal to 1 if respondent thinks government performance is worse than three years ago) and the main explanatory variable is the respondent‘s subjective assessments of the extent to which he or she was affected by the crisis. Using country fixed-effects and controlling for income and other demographic factors, those reporting being affected ―a great deal‖ by the crisis were on average 14 percent more likely to respond that the performance of the national government worsened in the past 3 years compared to those who were not affected by the crisis. This coefficient is significant at the 5 percent level and this impact of the crisis on government ratings is also found (albeit to a lesser extent) for ratings of local and regional performance. 11 Figure 9: Performance Ratings of National Government and Changes over Time Overall Ratings of the Performance of National Government Uzbekistan Tajikistan Sweden Azerbaijan Georgia Turkey Belarus Montenegro Kazakhstan Hungary Albania Estonia Mongolia Bulgaria Moldova Great Britain France Macedonia Slovenia Kyrgyzstan Russia Armenia Germany Ukraine Czech Republic Poland Bosnia-Herzegovina Serbia Italy Latvia Slovakia Lithuania Croatia Romania 0 20 40 60 80 100 Percent of respondents Very Good Good Neither Bad Very Bad Perceptions of change in performance of national government Georgia 40 Uzbekistan Azerbaijan Tajikistan 30 Turkey Bulgaria 20 Armenia Hungary Sweden Kazakhstan Belarus Mongolia Macedonia Montenegro Albania Moldova Kosovo Estonia Poland Slovenia Great Britain Ukraine 10 Czech Republic Kyrgyzstan SlovakiaSerbia Russia Latvia Bosnia-Herzegovina Italy France Lithuania Germany Croatia Romania 0 -10 -5 0 5 10 GDP growth 2009-2010 Source: LITS 2006 and 2010, IMF. 2010 growth numbers are estimates (IMF WEO October 2010 Update). Source: LiTS2 (2010). 12 4. CONCLUSION Despite the impact of the crisis, LiTS2 data indicate that satisfaction with public service delivery has risen over time in most countries in the region. Prevalence of unofficial payments is quite low: when respondents were asked how often it is necessary for people to make unofficial payments/gifts when using public services, a large majority reported that such payments are never needed. Nevertheless, the data show that the level of satisfaction with public service delivery in most transition countries tends to be lower than prevailing levels in western European comparator countries. Conversely, the perceived frequency of unofficial payments is higher than in comparator countries. We find that satisfaction data on peoples‘ experiences and perceptions of service delivery quality can usefully complement more objective measures of service delivery (e.g. from facilities surveys) to help policy makers and development practitioners track progress of country programs and strategies. The data provide insights on the linkages between utilization, satisfaction and the prevalence of unofficial services, as well as the availability of grievance redress mechanisms. An important limitation is, however, that comparisons across countries or individuals must be interpreted with caution, as people may not share the same basis for comparison or have information about the level of quality and service they should be expecting. Still, perceptions data can serve as a ―call to action‖ for governments if the data reveal outcomes that fall short of expectations (or conversely indicate low expectations). It can also serve a diagnostic function: while there is increased recognition that the quality of governance and service delivery affects service delivery outcomes, the empirical evidence on those links remains limited. How can governments in transition countries further increase citizens‘ satisfaction with service delivery? The insights gained from the LiTS2 provide clues as to how this might be achieved. First, LiTS2 data show utilization of public facilities in transition countries to be generally higher than in the western comparators; this suggests that further analysis may shed more light on the potential for efficiency and quality improvements, and the extent to which better demand management practices could free up public resources to improve quality of services. Second, the prevalence of unofficial payments in transition countries is higher than in western Europe, and is an important factor in explaining dissatisfaction with public service delivery.8 Lastly, the data show that mechanisms for grievance redress in the transition regions are still relatively underdeveloped in comparison to those in the comparator countries, and should be strengthened to help provide citizen feedback to policy- makers on the main problems faced when interacting with public service providers. 8 The inverse correlation between satisfaction on the one hand and usage rates and perceived prevalence of unofficial payments is confirmed by an Ordinary Least Squares (OLS) regression of country-level average satisfaction rates for each of the eight public services covered in the LiTS2 questionnaire. This uses as explanatory variables (i) average usage rates (that is, the percentage of respondents in the country that use that particular public service) and (ii) unofficial payments prevalence rates (the percentage who say that unofficial payments are usually or always needed for that particular service). The derived coefficients for the two explanatory variables in the regression are statistically significant at the 1 percent level. 13 Table 1: Satisfaction with Public Service Delivery by Country Percent of respondents satisfied with quality and efficiency of public service delivery Education Education Social Country Public health (prim. Traffic Official (vocational security Unemp. system /sec.) police documents ) benefits benefits Civil courts Estonia 78 83 61 90 84 73 70 76 Georgia 75 82 75 92 83 28 - 36 Turkey 79 72 39 75 72 38 55 49 Latvia 70 83 53 80 79 74 74 35 Moldova 67 80 39 82 79 59 39 65 Slovenia 70 71 50 75 70 70 67 65 Croatia 70 76 64 72 65 69 54 35 Lithuania 64 69 56 72 79 67 78 44 Poland 62 85 52 81 83 49 39 57 Belarus 59 74 50 73 79 60 66 54 Hungary 59 74 64 73 70 60 61 38 Montenegro 60 73 51 71 72 53 28 42 Slovak Rep. 64 78 49 70 76 42 37 45 Uzbekistan 64 74 43 54 65 48 14 29 Russia 58 75 41 68 79 74 59 51 Czech Rep. 66 73 45 70 68 43 47 44 Bosnia- Herz. 55 70 56 75 63 31 15 69 Armenia 61 72 44 70 70 43 39 30 Bulgaria 61 75 26 52 84 59 48 17 Serbia 60 68 43 61 66 50 47 26 Romania 57 76 49 63 67 49 53 33 Tajikistan 57 62 36 62 51 42 38 29 Kazakhstan 54 69 40 50 62 50 43 27 Mongolia 53 67 14 48 66 34 36 17 Ukraine 45 72 22 57 73 57 41 34 Macedonia 45 66 41 44 53 30 33 42 Kyrgyz Rep. 53 58 21 36 56 46 44 5 Albania 45 67 29 51 49 31 29 47 Western Europe 80 76 55 80 78 67 52 40 Note: Countries ranked in order of average satisfaction rates across all eight public services. 14 Table 2: Reported Prevalence of Unofficial Payments by Country Percent of respondents that report unofficial payments are usually or always needed Public Social Country health Education Traffic Official Vocational security Unemp. Civil system (prim/sec) police docs. Education benefits benefits courts Estonia 4 1 1 0 1 1 0 1 Poland 8 1 4 1 2 1 1 2 Latvia 12 1 5 2 1 1 1 3 Georgia 7 5 1 1 5 2 4 3 Slovenia 8 3 3 3 3 3 3 5 Czech Rep. 11 5 5 3 7 3 3 3 Montenegro 13 6 7 4 6 4 4 3 Lithuania 22 1 6 2 3 7 4 2 Croatia 17 4 8 2 8 3 3 6 Macedonia 13 3 5 5 8 8 7 7 Bulgaria 17 3 13 3 7 3 3 8 Belarus 21 6 14 3 9 2 3 4 Serbia 23 3 11 6 7 5 4 6 Russia 19 10 13 3 12 2 3 6 Bosnia- Herz. 22 9 11 6 11 6 6 8 Hungary 42 3 11 6 6 3 4 7 Romania 44 7 12 5 9 4 4 10 Slovak Rep. 22 11 12 5 19 6 8 12 Mongolia 19 9 12 11 16 8 9 15 Kazakhstan 18 9 21 12 16 8 7 13 Turkey 17 21 16 15 14 11 11 11 Uzbekistan 20 20 18 12 25 11 9 8 Armenia 28 15 22 19 18 13 13 12 Ukraine 43 17 28 14 26 7 8 17 Tajikistan 39 18 29 16 32 10 10 13 Albania 39 24 19 22 23 15 14 14 Moldova 45 18 32 23 24 11 9 19 Kyrgyz Rep. 49 38 43 37 51 29 32 33 Azerbaijan 72 65 66 60 65 63 65 61 Western Europe 3 1 1 1 1 1 1 1 Note: Countries ranked in reverse order of average prevalence rates across all eight public services. 15 Table 3: Changes in satisfaction with service delivery in transition countries, 2006-10 Percent of Respondents Satisfied With Quality and Efficiency of Public Service Delivery Public Health Road Official Unemployment Other Social Civil System Police Documents Benefits Security Courts Benefits 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. Moldova 43 67 25 27 39 13 49 82 33 13 39 26 4 59 55 43 65 22 Russia 42 58 16 21 41 21 33 68 35 28 59 31 25 74 49 18 51 33 Turkey 51 79 28 41 39 -2 60 75 15 21 55 34 42 38 -5 35 49 13 Tajikistan 37 57 20 21 36 15 34 62 28 11 38 27 26 42 15 38 29 -9 Estonia 60 78 18 64 61 -4 73 90 17 49 70 21 63 73 10 48 76 28 Montenegro 49 60 11 28 51 23 54 71 17 20 28 8 22 53 31 22 42 20 Romania 46 57 11 44 49 5 36 63 26 24 53 29 16 49 33 27 33 6 Belarus 49 59 9 43 50 7 42 73 30 17 66 50 29 60 31 49 54 5 Kazakhstan 39 54 15 24 40 16 36 50 14 42 43 1 38 50 13 35 27 -8 Bosnia- 50 55 5 30 56 25 54 75 22 33 15 -18 9 31 22 32 69 37 Herz Serbia 46 60 14 33 43 10 51 61 9 32 47 15 6 50 44 17 26 9 Ukraine 34 45 11 25 22 -3 42 57 15 11 41 30 20 57 37 28 34 5 Georgia 63 75 12 71 75 4 59 92 33 27 0 -27 21 28 7 43 36 -7 Poland 46 62 16 50 52 2 67 81 14 32 39 8 41 49 8 51 57 6 Hungary 50 59 9 47 64 17 62 73 12 27 61 34 51 60 8 46 38 -9 Slovenia 65 70 5 36 50 14 62 75 13 50 67 17 36 70 34 46 65 19 Czech Rep 60 66 6 27 45 17 51 70 18 35 47 11 42 43 1 28 44 16 Croatia 65 70 6 55 64 9 54 72 17 29 54 25 30 69 39 41 35 -6 Latvia 56 70 14 50 53 3 65 80 14 76 74 -3 66 74 7 59 35 -24 Uzbekistan 48 64 16 41 43 2 48 54 6 14 14 -1 53 48 -5 25 29 4 Lithuania 56 64 8 49 56 7 56 72 16 53 78 25 60 67 7 53 44 -9 Albania 40 45 5 19 29 10 38 51 13 14 29 15 37 31 -6 29 47 18 Slovakia 58 64 7 43 49 6 56 70 15 12 37 24 37 42 5 39 45 6 Mongolia 48 53 4 19 14 -5 35 48 13 13 36 23 54 34 -20 18 17 -1 Armenia 64 61 -2 32 44 13 62 70 8 27 39 12 50 43 -6 11 30 19 Kyrgyzstan 50 53 3 17 21 4 35 36 1 19 44 25 43 46 3 23 5 -18 Bulgaria 60 61 1 23 26 2 72 52 -20 41 48 7 29 59 29 31 17 -15 Macedonia 50 45 -5 38 41 3 53 44 -8 16 33 17 18 30 12 24 42 17 Azerbaijan 56 45 -10 24 16 -8 51 61 10 15 5 -10 21 12 -9 47 26 -21 Notes: Ch. denotes change in satisfaction between 2006 and 2010. Countries sorted in order of average increase across all seven services. Data on satisfaction are not fully comparable between the 2006 and 2010 LiTS surveys. The 2006 survey asked respondents if ―you personally‖ used a service (and, conditional on use, whether you were satisfied); while the 2010 survey asked if ―you or anyone in your household‖ used the service. 16 Table 4: Changes in reported prevalence of unofficial payments in transition countries, 2006-10 Per cent o f respondents tha t report unofficial paym ents are usually/always needed Public he alth Traffic O fficial Education Unem plo yme nt O the r social Civil system police docum e nts (Tertiary) benefits security bene fits courts 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. 2006 2010 Ch. Macedonia 23 13 -10 10 5 -5 11 5 -6 17 8 -9 11 8 -4 11 7 -4 13 7 -7 Bulgaria 24 17 -7 21 13 -8 9 3 -6 11 7 -4 6 3 -3 6 3 -3 16 8 -8 Uzbekistan 33 20 -13 22 18 -3 15 12 -2 30 25 -5 12 9 -2 14 11 -3 12 8 -3 Russia 24 19 -5 19 13 -6 9 3 -5 20 12 -8 4 2 -2 6 3 -3 7 6 0 Uk raine 40 43 3 28 28 -1 20 14 -6 33 26 -8 11 7 -4 13 8 -5 18 17 -1 Albania 48 39 -8 23 19 -4 28 22 -5 17 23 6 15 15 1 15 14 -1 21 14 -7 Poland 15 8 -7 4 4 -1 1 1 0 4 2 -3 1 1 0 1 1 0 1 2 1 Latvia 13 12 -1 6 5 -2 3 2 -1 3 1 -1 1 1 0 1 1 0 2 3 1 Mongolia 17 19 3 18 12 -5 16 11 -5 15 16 1 6 8 2 7 9 1 15 15 0 Montenegro 18 13 -5 4 7 3 3 4 1 8 6 -2 3 4 0 4 4 0 4 3 0 Estonia 3 4 0 0 1 1 1 0 0 1 1 0 0 1 0 1 0 0 1 1 0 Ge orgia 8 7 0 1 1 1 2 1 -1 4 5 1 4 2 -2 4 4 0 2 3 1 Lithuania 23 22 -1 6 6 -1 4 2 -2 4 3 -1 0 7 7 2 4 2 5 2 -3 Croatia 14 17 3 6 8 2 3 2 -1 5 8 3 4 3 -1 6 3 -4 5 6 1 Bosnia-Herz. 19 22 3 16 11 -5 7 6 0 8 11 2 4 6 2 5 6 1 7 8 1 Cze ch R ep. 10 11 1 6 5 -1 2 3 1 7 7 -1 1 3 2 1 3 2 3 3 0 Slovenia 7 8 1 2 3 1 1 3 2 5 3 -2 2 3 1 1 3 2 2 5 3 Ro ma nia 31 44 13 10 12 2 8 5 -2 9 9 0 6 4 -2 7 4 -2 7 10 2 Serbia 17 23 6 7 11 4 4 6 2 5 7 2 3 5 2 3 4 1 6 6 1 Slovak Re p. 21 22 1 12 12 -1 3 5 2 13 19 6 2 6 4 2 8 5 12 12 1 Kaza khstan 11 18 7 16 21 5 10 12 1 16 16 0 4 8 4 6 7 1 9 13 4 Belarus 12 21 9 9 14 5 4 3 -1 5 9 4 1 2 1 2 3 1 1 4 3 Hungary 33 42 10 6 11 5 5 6 1 5 6 1 0 3 3 1 4 3 5 7 2 Arm enia 19 28 9 18 22 4 11 19 8 16 18 2 8 13 5 10 13 3 7 12 5 Tajik istan 33 39 6 14 29 15 15 16 0 23 32 9 8 10 2 8 10 2 8 13 5 Turkey 11 17 6 8 16 7 6 15 9 6 14 8 4 11 7 5 11 7 4 11 7 Moldo va 32 45 14 18 32 14 12 23 11 17 24 7 8 11 3 9 9 0 12 19 7 Kyrgyz Rep. 28 49 21 24 43 18 22 37 15 29 51 22 12 29 17 14 32 18 14 33 20 Azerba ijan 30 72 42 11 66 56 15 60 45 16 65 49 7 63 55 7 65 57 8 61 53 Note: Ch. denotes change in prevalence of unofficial payments between 2006 and 2010. Countries sorted in reverse order of average prevalence rates across all seven services. 17 Table 5: Utilization and satisfaction: public health system (1) (2) (3) (4) (5) Satisfacti Satisfacti Satisfied Satisfied on level on level Utilization (0-1) (0-1) (1-5) (1-5) Ordered Ordered Probit Probit Probit probit probit House hold-le ve l va ria ble s Consumption group Lower -0.456*** 0.146*** 0.0908* 0.0947** 0.0413 (0.0384) (0.0565) (0.0545) (0.0452) (0.0452) Middle -0.137*** 0.164*** 0.112** 0.163*** 0.111*** (0.0380) (0.0420) (0.0435) (0.0357) (0.0358) Upper Reference category Locality Rural -0.0628* 0.0650 0.0585 0.0163 0.00467 (0.0351) (0.0409) (0.0415) (0.0348) (0.0343) Metropolitan -0.146*** -0.0422 -0.000152 -0.105** -0.0591 (0.0559) (0.0622) (0.0631) (0.0529) (0.0527) Urban Reference category Number of children (0-14) 0.181*** (0.0230) Number of women aged 15-44 0.0769*** (0.0235) Number of men aged 15-44 0.137*** (0.0246) Number of women aged 45-64 0.0102 (0.0367) Number of men aged 45-64 0.212*** (0.0380) Number of women aged 65+ 0.130** (0.0506) Number of men aged 65+ 0.272*** (0.0483) Individua l-le ve l va ria ble s Health status of respondent Very good -0.232*** 0.423*** 0.363*** 0.360*** 0.284*** (0.0535) (0.0681) (0.0721) (0.0602) (0.0616) Good -0.224*** 0.216*** 0.142*** 0.181*** 0.111*** (0.0373) (0.0472) (0.0482) (0.0398) (0.0400) Medium Reference category Bad 0.195*** -0.180*** -0.141** -0.180*** -0.143*** (0.0545) (0.0593) (0.0612) (0.0508) (0.0506) Very bad 0.258** -0.235** -0.141 -0.200** -0.115 (0.106) (0.0978) (0.0951) (0.0910) (0.0917) Education of respondent Lower secondary or below Reference category Upper secondary -0.0502 -0.0725 -0.0649 -0.101** -0.0906** (0.0413) (0.0502) (0.0512) (0.0422) (0.0418) Post secondary or higher -0.0551 -0.0906* -0.0406 -0.109** -0.0659 (0.0431) (0.0513) (0.0532) (0.0425) (0.0425) 18 Table 5 (continued) (1) (2) (3) (4) (5) Satisfacti Satisfacti Satisfied Satisfied on level on level Utilization (0-1) (0-1) (1-5) (1-5) Ordered Ordered Probit Probit Probit probit probit How often are unofficial payments needed Never 0.254*** 0.215*** (0.0504) (0.0410) Seldom 0.0444 0.0502 (0.0583) (0.0451) Sometimes Reference category Usually -0.215*** -0.202*** (0.0539) (0.0440) Always -0.342*** -0.423*** (0.0629) (0.0587) Female 0.0624* 0.0731** 0.0176 0.0237 (0.0367) (0.0372) (0.0305) (0.0303) 18-34 -0.108** -0.0720 -0.105** -0.0657 (0.0522) (0.0545) (0.0434) (0.0441) 35-44 Reference category 45-54 0.137** 0.0936 0.129*** 0.0902* (0.0599) (0.0624) (0.0484) (0.0487) 55-64 0.121** 0.0492 0.133*** 0.0723 (0.0605) (0.0622) (0.0506) (0.0506) 65+ 0.323*** 0.240*** 0.295*** 0.219*** (0.0680) (0.0705) (0.0586) (0.0579) Satisfied with life Strongly disagree -0.292*** -0.239*** -0.300*** -0.258*** (0.0567) (0.0591) (0.0493) (0.0501) Disagree -0.142*** -0.108** -0.147*** -0.111*** (0.0480) (0.0488) (0.0400) (0.0394) Neither disagree nor agree Reference category Agree 0.0952 0.113* 0.110** 0.129*** (0.0586) (0.0609) (0.0473) (0.0474) Strongly agree 0.278** 0.260** 0.192** 0.163* (0.114) (0.115) (0.0970) (0.0941) Prob lems encountered in local pub lic health clinic or hospital in past 12 months Frequent and unjustified absence of doctors -0.321*** -0.246*** (0.0630) (0.0531) Treated disrespectfully by staff -0.482*** -0.455*** (0.0532) (0.0430) No drugs available -0.255*** -0.234*** (0.0489) (0.0400) Long waiting times -0.353*** -0.355*** (0.0404) (0.0324) Facilities not clean -0.115 -0.0809 (0.0728) (0.0569) Payments required for services that should be free -0.513*** -0.477*** (0.0500) (0.0414) Inverse Mills Ratio (from selection equation) -0.0719 -0.0761 (0.162) (0.158) Constant 0.320*** -0.204 0.192 (0.0771) (0.147) (0.125) Cut 1 -1.402*** -1.895*** (0.114) (0.114) Cut 2 -0.331*** -0.725*** (0.113) (0.111) Cut 3 0.0914 -0.261** (0.113) (0.111) Cut 4 2.057*** 1.821*** (0.116) (0.113) Observations 28,920 28,248 28,248 19,510 19,510 Pseudo R-squared 0.0720 0.0567 0.105 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Country-level dummies not shown Columns (2)-(5): with Heckman sample selection correction 19 Table 6 : Utilization and satisfaction: education (1) (2) (3) (4) (5) S atisfacti S atisfacti S atisfied S atisfied on lev el on lev el Utilization (0-1) (0-1) (1-5) (1-5) Ordered Ordered P robit P robit P robit probit probit Household-level variables Consumption group Lower -0.304*** 0.0477 -0.0133 * -0.305* * -0.0522 (0.0468) (0.0830) (0.0841) (0.0474) (0.0685) Middle -0.0973* * 0.158** 0.125 -0.101** 0.0811 (0.0418) (0.0775) (0.0792) (0.0422) (0.0674) Upper Reference category Locality Rural 0.0548 0.0688 0.0138 0.0493 0.0465 (0.0419) (0.0728) (0.0745) (0.0423) (0.0597) Metropolitan -0.0472 0.0817 0.0489 -0.0433 0.0774 (0.0632) (0.115) (0.120) (0.0638) (0.103) Urban Number of children (0-18) 0.821*** 0.819*** (0.0233) (0.0235) Individual-level variables (respondent) E ducation Lower secondary and below Reference category Upper secondary 0.165*** -0.0613 -0.0113 0.165*** 0.0385 (0.0503) (0.0931) (0.0949) (0.0509) (0.0772) P ost secondary or higher 0.132*** -0.103 -0.0378 0.133** -0.0717 (0.0512) (0.0997) (0.103) (0.0519) (0.0804) Female 0.0970 0.0758 0.00776 (0.0677) (0.0690) (0.0591) A ge of respondent 18-34 -0.113 -0.134* -0.111* (0.0755) (0.0759) (0.0621) 35-44 Reference category 45-54 0.102 0.0612 0.0615 (0.0953) (0.0996) (0.0829) 55-64 0.104 0.0134 0.0140 (0.139) (0.141) (0.101) 65+ -0.258 -0.324 -0.314* (0.201) (0.213) (0.167) S atisfied with life S trongly disagree -0.423*** -0.417*** * -0.368* * (0.109) (0.111) (0.0950) Disagree -0.185** -0.186** * -0.180* * (0.0879) (0.0890) (0.0689) Neither disagree nor agree Reference category A gree -0.0151 -0.0168 0.107 (0.100) (0.102) (0.0839) S trongly agree 0.318* 0.270 0.206* (0.182) (0.169) (0.119) P roblems experienced w ith local public schools in past 12 months No textbooks or other supplies that should be prov ided free of charge -0.0729 -0.0876 (0.0827) (0.0725) P oor teaching -0.657*** * -0.667* * (0.0889) (0.0762) Frequent and unjustified absence of teachers -0.389*** -0.251** (0.133) (0.128) Overcrowded classrooms -0.381*** * -0.290* * (0.0984) (0.0829) Facilities in poor condition -0.0209 -0.0640 (0.0890) (0.0697) P ayments required for services that should be free -0.271*** * -0.306* * (0.0857) (0.0750) How often are unofficial payments needed Nev er 0.261*** (0.0930) S eldom -0.0946 (0.107) S ometimes Reference category Usually -0.166 (0.102) A lways -0.103 (0.114) erse mills ratio (from selection equation) Inv -0.0351 -0.0378 (0.0582) (0.0580) Constant -0.912*** 0.448** 0.730*** * -0.959* * (0.0702) (0.181) (0.178) (0.0718) Cut 1 * -2.361* * (0.162) Cut 2 * -1.446* * (0.151) Cut 3 * -0.939* * (0.150) Cut 4 * 1.313* * (0.152) Observations 28,823 28,495 28,495 28,495 6,892 P seudo R-squared 0.273 0.0728 Robust standard errors in parentheses ** * p<0.01, ** p<0.05, * p<0.1 el Country-lev dummies not shown Columns (2)-(5): with Heckman sample selection correction 20 Figure 10. Satisfaction rates vs. prevalence of unofficial payments by country, 2010 21 References M. Aldons (2001), ―Responsible, representative and accountable government‖, Australian Journal of Public Administration, Vol. 60 (1), pp. 34–42. S. Amin, J. Das and M. Goldstein, eds. (2008), . ―Are you being served? New tools for measuring service delivery�, World Bank, Washington, D.C. S. Chakraborty (2009), ―Health systems strengthening: lessons from the Turkish experience,‖ Europe Central Asia Knowledge Brief. December 2009, Vol. 12. J. Das and J. Hammer (2007), ―Money for nothing: the dire straits of medical practice in Delhi, India‖, Journal of Development Economics, Vol. 83, No. 1, May 2007, pp. 1-36 B. Dasgupta, A. Narayan and E. Skoufias (2009), ―Measuring the quality of education and health services: the use of perception data from Indonesia‖, Policy Research Working Paper 5033, World Bank. L. de Las Casas (2005), ―Complaint and redress mechanisms in international organizations: background research for the Complaint and Redress Dimension‖, Global Accountability Project. A. Fiszbein, D. Ringold and H. Rogers (2011), ―Making services work: indicators, assessments, and benchmarking of the quality of public service delivery‖, Policy Research Working Paper No. 5690, The World Bank. R. Johnson and S. Mehra. (2002), ‗Best-Practice Complaint Management‘, The Academy of Management Executive (1993-2005), Vol. 16, No. 4, (Nov., 2002), pp. 145-154. M. Lewis and G. Pettersson (2009a), "Governance in health care delivery: raising performance", World Bank, Washington, D.C. M. Lewis and G. Pettersson (2009b), "Governance in education: raising performance", World Bank, Washington, D.C. G. Lister, F. Rosleff, M. Boudioni, F. Dekkers, E. Jakubowski and H. Favelle (2008), ‗Handling Complaints in Health and Social Care: International Lessons for England‘, Report prepared for the UK National Audit Office. F. Rogers and M. Koziol (forthcoming, 2011), ―Provider absence surveys: a guidance note‖, World Bank, Washington, D.C. World Bank (2003), World Development Report 2004: Making services work for poor people. Washington, D.C, Oxford University Press for the World Bank. 22