WPS8586 Policy Research Working Paper 8586 Efficiency of Public Spending in Education, Health, and Infrastructure An International Benchmarking Exercise Santiago Herrera Abdoulaye Ouedraogo Macroeconomics, Trade and Investment Global Practice September 2018 Policy Research Working Paper 8586 Abstract Governments of developing countries typically spend data from 2006–16 on education, health, and infra- between 20 and 30 percent of gross domestic product. structure. The paper verifies empirical regularities of the Hence, small changes in the efficiency of public spending cross-country variation in efficiency, showing a negative could have a major impact on aggregate productivity growth association between efficiency and spending levels and the and gross domestic product levels. Therefore, measuring ratio of public-to-private financing of the service provision. efficiency and comparing input-output combinations of Other variables, such as inequality, urbanization, and aid different decision-making units becomes a central chal- dependency, show mixed results. The efficiency of capi- lenge. This paper gauges efficiency as the distance between tal spending is correlated with the quality of governance observed input-output combinations and an efficiency indicators, especially regulatory quality (positively) and per- frontier estimated by means of the Free Disposal Hull and ception of corruption (negatively). Although no causality Data Envelopment Analysis techniques. Input-inefficiency may be inferred from this exercise, it points at different (excess input consumption to achieve a level of output) factors to understand why some countries might need more and output-inefficiency (output shortfall for a given level resources than others to achieve similar education, health, of inputs) are scored in a sample of 175 countries using and infrastructure outcomes. This paper is a product of the Macroeconomics, Trade and Investment Global Practice. 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://www.worldbank.org/research. The authors may be contacted at sherrera@worldbank.org and aouedrago3@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 Efficiency of Public Spending in Education, Health, and Infrastructure: An International Benchmarking Exercise1 Santiago Herrera Abdoulaye Ouedraogo JEL Classification: C14, H5, H51, H52, H54, I1, I2, E62 Keywords: efficiency frontiers, efficiency of public spending, education expenditure, health expenditure, infrastructure spending. 1 The paper carries the names of the authors and should be cited accordingly. It follows closely the presentation of Herrera-Pang (2005) and uses the data and STATA codes available in the WPS website. The findings, interpretations and conclusions are the authors’ responsibility and do not necessarily represent the view of the World Bank or its Executive Directors. I. Introduction Governments of developing countries typically spend resources equivalent to between 15 and 30 percent of GDP. Hence, small changes in the efficiency of public spending could have a significant impact on GDP and on the attainment of the government’s objectives. The first challenge faced by stakeholders is measuring efficiency. This paper attempts such quantification and verifies empirical regularities in the cross country-variation in the efficiency scores. The paper has four chapters. The first one presents the methodology that defines efficiency as the distance from the observed input-output combinations to an efficient frontier. This unobservable frontier is estimated using the Free Disposable Hull (FDH) and Data Envelopment Analysis (DEA) techniques. The exercise focuses on health and education expenditure because they absorb the largest share of most countries’ budgets, and because it facilitates comparison of results with existing literature.2 We extend the analysis by measuring efficiency of capital spending in achieving several infrastructure output indicators. The second chapter estimates the efficiency frontiers for 14 education output indicators and seven health output indicators, based on a sample of data for 175 countries for the period 2006-2016, and six infrastructure output indicators based on a sample of 150 countries using 2007-2015 data. Both input-efficiency (excess input consumption to achieve a level of output) and output-efficiency (output shortfall for a given level of inputs) are scored. The chapter presents both the single input-single-output and the multiple-inputs multiple-outputs frameworks. The third chapter explores empirical regularities in the cross-country variation in the efficiency scores. Using a Tobit panel approach, this chapter explores the relationship between efficiency scores and expenditure levels, the wage bill as a share of the total budget, the degree of urbanization, and aid-dependency. The fourth and last chapter summarizes the conclusions. II. Measuring Efficiency: Methodologies and Overview of the Literature This chapter describes the empirical methods used in this paper to measure efficiency and surveys the literature more directly related to the analysis of public expenditure efficiency. Empirical and theoretical measures of efficiency are based on ratios of observed output levels to the maximum that could have been obtained given the inputs utilized. This maximum constitutes the efficient frontier which will be the benchmark for gauging the relative efficiency of the observations. There are multiple techniques to estimate this frontier and they have been applied to examine the efficiency of public spending in different contexts. These are the topics of the next two sections. 2 This paper follows closely the Herrera-Pang (2005) presentation to facilitate intertemporal comparisons. 2 II.1. Methods for Measuring Efficiency The origin of the discussion of efficiency measurement dates to Farell (1957), who identified two different ways in which productive agents could be inefficient: one, they could use more inputs than technically required to obtain a given level of output, or two, they could use a sub-optimal input combination given the input prices and their marginal productivities. The first type of inefficiency is termed technical inefficiency while the second one is known as allocative inefficiency. These two types of inefficiency can be represented graphically by means of the unit isoquant curve in Figure 1. The set of minimum inputs required for a unit of output lies on the isoquant curve YY’. An agent’s input-output combination defined by bundle P produces one unit of output using input quantities X1 and X2. Since the same output can be achieved by consuming less of both inputs along the radial back to bundle R, the segment RP represents the inefficiency in resource utilization. The technical efficiency (TE), input-oriented, is therefore defined as TE = OR/OP. Furthermore, the producer could achieve additional cost reduction by choosing a different input combination. The least cost combination of inputs that produces one unit of output is given by point T, where the marginal rate of technical substitution is equal to the input price ratio. To achieve this cost level implicit in the optimal combination of inputs, input use needs to be contracted to bundle S. The input allocative efficiency (AE) is defined as AE = OS/OR. X2/Y Y T P R S Y’ O X1/Y Figure 1 Technical and Allocative Inefficiency This paper focuses on technical efficiency, due to the lack of comparable input prices across the countries. This concept of efficiency is narrower than the one implicit in social welfare analysis. That is, countries may be producing the wrong output very efficiently (at low cost). We abstract from this consideration (discussed by Tanzi 2004), focusing on the narrow concept of efficiency. 3 Numerous techniques have been developed to estimate the unobservable efficient frontier (in this case the isoquant YY”). These may be classified using several taxonomies. The two most widely used catalog methods into parametric or non-parametric, and into stochastic or deterministic. The parametric approach assumes a specific functional form for the relationship between the inputs and the outputs as well as for the inefficiency term incorporated in the deviation of the observed values from the frontier. The non- parametric approach calculates the frontier directly from the data without imposing specific functional restrictions. The first approach is based on econometric methods, while the second one uses mathematical programming techniques. The deterministic approach considers all deviations from the frontier explained by inefficiency, while the stochastic focus considers those deviations a combination of inefficiency and random shocks outside the control of the decision maker. This paper uses non-parametric methods to avoid assuming specific functional forms for the relationship between inputs and outputs or for the inefficiency terms. Other papers use the parametric approach (Greene (2003) Grigio (2013, 2014). The remainder of the section briefly describes the two methods: Free Disposable Hull (FDH) and Data Envelopment Analysis (DEA). The FDH method imposes the least amount of restrictions on the data, as it only assumes free-disposability of resources. Figure 2 illustrates the single-input single-output case of the FDH production possibility frontier. Countries A and B use inputs XA and XB to produce outputs YA and YB, respectively. The input efficiency score for country B is defined as the quotient XA/XB. The output efficiency score is given by the quotient YB/YA. A score of one implies that the country is on the frontier. An input efficiency score of 0.75 indicates that this country uses inputs in excess of the most efficient producer to achieve the same output level. An output efficiency score of 0.75 indicates that the inefficient producer attains 75 percent of the output obtained by the most efficient producer with the same input intake. Multiple input and output efficiency tests can be defined in an analogous way. Output D E C A YA YB B O XA XB Input Figure 2 Free Disposal Hull (FDH) production possibility frontier 4 The second approach, Data Envelopment Analysis (DEA), assumes that linear combinations of the observed input-output bundles are feasible. Hence it assumes convexity of the production set to construct an envelope around the observed combinations. Figure 3 illustrates the single input- single output DEA production possibility frontier. In contrast to the vertical step-ups of the FDH frontier, the DEA frontier is a piecewise linear locus connecting all the efficient decision-making units (DMU). The feasibility assumption, displayed by the piecewise linearity, implies that the efficiency of C, for instance, is not only ranked against the real performers A and D, called the peers of C in the literature, but also evaluated with a virtual decision maker, V, which employs a weighted collection of A and D inputs to yield a virtual output. DMU C, which would have been efficient by FDH, now lies below the variable returns to scale (VRS, further defined below) efficiency frontier, XADF, by DEA ranking. This example shows that FDH efficiency scores are higher than DEA ones. The input-oriented technical efficiency of C is now defined by TE = YV/YC. CRS F D Output N VRS Y V C A B VRS O X Input Figure 3 DEA production possibility frontier If constant returns to scale (CRS) characterize the production set, the frontier may be represented by a ray extending from the origin through the efficient DMU (ray OA). By this standard, only A would be rated efficient. The important feature of the XADF frontier is that this frontier reflects variable returns to scale. The segment XA reflects locally increasing returns to scale (IRS), that is, an increase in the inputs results in a greater than proportionate increase in output. Segments AD and DF reflect decreasing returns to scale. It is worth noticing that constant returns to scale technical efficiency (CRSTE) is equal to the product of variable returns to scale technical efficiency (VRSTE) and scale efficiency (SE). Accordingly, DMU D is technically efficient but scale inefficient, while DMU C is neither technically efficient nor scale efficient. The scale efficiency of C is calculated as YN/YV. For more detailed exploration of returns to scale, readers are referred to Charnes, Cooper, and Rhodes (1978) and Banker, Charnes, and Cooper (1984), among others. 5 The limitations of the non-parametric method derive mostly from the sensitivity of the results to sampling variability, to the quality of the data and to the presence of outliers. This led the literature to explore the relationship between statistical analysis and non- parametric methods (Simar and Wilson, 2000) and propose solutions like constructing confidence intervals for the efficiency scores using asymptotic theory in the single input case (for input-efficiency estimators) or single-output (in the output efficiency) case, given these are shown to be maximum likelihood estimators (Banker, 1993 and Goskpoff, 1996). For multiple input-output cases the distribution of the efficiency estimators is unknown or quite complicated and analysts recommend constructing the empirical distribution of the scores by means of bootstrapping methods (Simar and Wilson, 2000). Other solutions to the outlier or noisy data consist in constructing a frontier that does not envelop all the data points, building an expected minimum input function or expected maximum output functions (Cazals, Florens and Simar, 2002, and Wheelock and Wilson, 2003). Another limitation of the method, at least in the context of this paper, is the inadequate treatment of dynamics, given the lag between input consumption (public expenditure) and output production (health and education outcomes). II.2. Overview of Precursor Papers There is abundant literature measuring productive efficiency of diverse types of decision- making units. For instance, there are papers measuring efficiency of museums (Bishop and Brand, 2003), container terminals (Cullinane and Song, 2003; Herrera-Pang 2006), electric generation plants (Cherchye and Post 2001), banks (Wheelock and Wilson, 2003), schools (Worthington, 2001) and hospitals (Bergess and Wilson, 1998), among others. Few papers, however, analyze aggregate public-sector spending efficiency using cross-country data. These are the direct precursors of this paper and have been surveyed elsewhere.3 The more recent directly related papers are Herrera-Pang (2005), which use non- parametric methods (FDH and DEA) to gauge efficiency for numerous health and education output indicators. Using a large set of developing countries for the period 1996-2002, the paper estimates efficiency scores in a first stage and in a second stage they use panel methods to explore the correlates of efficiency variation across countries. One of the common results of that paper and previous literature is that countries with low attainment in results in health or education show up with high efficiency scores. The apparently counterintuitive result is due to some countries’ low spending levels and low education attainment results, which are considered as the “origin” of the efficiency frontier: though the country appears as efficient because the peer countries spend more (in the OECD context), its output levels are also higher and probably more desirable from a broader perspective. The next chapters discuss this topic and report similar results for other countries. 3 Gupta and Verhoeven (2001), Evans and Tandon (2000), Jarasuriya and Woodon (2002) Greene (2003a), Afonso, Schuknecht and Tanzi (2003), Afonso and St. Aubyn (2004) were surveyed in Herrera-Pang (2005). 6 Afonso, Romero and Monsalve (2013) focus exclusively on Latin American and Caribbean countries, and define aggregate measures of public sector output, which include education, health infrastructure, macro stability, and economic performance. The aggregations are weighted and unweighted. Given the complexity of the aggregations, their results are not directly comparable to Herrera-Pang or Grigoli’s results described below. But their findings in general show output efficiency scores higher than input efficiency scores, like the Herrera-Pang results for LAC, especially in health. Grigoli and Kapsoli (2013) use stochastic frontier analysis (SFA) to examine efficiency in health spending in a sample of 80 countries. Using a single indicator for health outcomes, the health adjusted life expectancy (HALE) average for 2006-2010, and lagged values of spending and other correlates (average of 2001-2005), the authors compute efficiency scores. This is an innovative way to capture the dynamics in the relationship between spending and health outcomes, but still imposes restrictions on the data. The reported scores are high, and the authors acknowledge the fact when reporting average efficiency scores of .94 compared to significantly lower scores of .81 of other studies. That average also seems high compared to results presented in this paper. The efficiency rankings of Grigoli and Kapsoli (2013) are also strange, as the efficiency scores of Trinidad and Tobago, for instance, ranked within the most efficient group in Herrera- Pang (2005) and Afonso et.al (2013), show up in the third quartile of the efficiency distribution. Grigoli (2014) uses an envelopment methodology, variant of the DEA, designed by Wagstaff and Wang (2011), to examine efficiency of education spending. One of the advantages of this approach is that it allows dealing with heterogeneity across groups, as different frontiers are constructed for different groups of countries. It also uses a LOWESS method, which helps dealing with the measurement error, data outliers, and stochastic nature of the problem at hand. The added complexity comes at a cost, as the study uses only one education outcome variable, net secondary enrollment. The efficiency scores have a larger dispersion and little information is given on the distribution of the scores, but many countries about 15% of the total, are on the efficiency frontier. Perelman et al. (2017) use two concepts, performance and productivity, to compare Latin American countries’ efficiency in education and health spending. The performance concept compares outcomes without considering resource constraints. Their concept of productivity links the outcomes to the resources, like the efficiency concept used in this paper and other literature in this review. For education outcomes, the authors use the PISA scores exclusively, while the health outcome is a human development index which is an aggregation of several health outcomes.4 The education efficiency scores of the Perelman et al. study are, on average, .63 for input-efficiency and .75 for output- 4 The education outcome considers only six LAC countries. The health outcomes used to compute the human development index are Disability-Adjusted Life Years (DALY), the probability of dying between 30 and 70 years old from cardiovascular, cancer, diabetes, or chronic respiratory diseases (PROB); the percentage of out-of-pocket expenditure in total health expenditure as an indicator of financial protection (FIPR); and an indicator of equity in health EQTY. 7 efficiency. In health, the average input-efficiency is .59 and output-efficiency is .90. These results match the pattern found in other studies, including the present one, of output-efficiency exceeding input-efficiency, and that of health efficiency tending to be higher than education, especially on the output measure.5 III. Empirical Results III.1. Input and Output Indicators: Description, Assumptions and Limitations Cross-country comparisons assume some homogeneity across the world in the production technology of health, education, and infrastructure.6 There are two particular aspects in which the homogeneity assumption is important. First, the comparison assumes that there is a small number of factors of production that are the same across countries. Any omission of an important factor will yield as a result a high efficiency ranking of the country that uses more of the omitted input. Second, the comparison requires that the quality of the inputs is the same, with the efficiency scores biased in favor of countries where the quality is of higher grade. Factor heterogeneity will not be a problem if it is evenly distributed across countries. It will be problematic if there are differences between countries in the average quality of a factor (Farrell, 1957). The exercise that we present suffers from this limitation, given that the main input in both production technologies is used more intensively in richer countries (with higher per-capita GDP). The main input is public spending per capita on education, health and infrastructure, and it shows a clear positive association with per- capita GDP (Figures 4A, 4B and 4C).7 This positive association may be explained by several factors. One of them is the Balassa-Samuelson effect, according to which price levels in wealthier countries tend to be higher than in poorer countries.8 This applies to both final goods and factor prices. Thus, the price of the same service (health or education, for instance) will be higher in the country with higher GDP. Similarly, wages in the relatively richer counties are higher, given the higher marginal productivity of labor, which will tend to increase costs, especially in labor-intensive activities such as health and education. 5 Averages of Perelman scores come from Tables 6 and 19. The order of magnitude of the averages is very similar to those reported in this study for the PISA scores of the same LAC country sample. The average reported in this study for education input efficiency is 0.67 and the output efficiency is .80. The LAC region health input efficiency score is .48 and output efficiency is .95, presented in tables in the next section. 6 See Table A.9 in Appendix A for the list of countries included in the study. 7 Education and health spending are measured in constant 2011 US dollars in PPP terms, and is an update of the figures presented in Herrera-Pang (2005), and the source is WDI from the World Bank. Capital spending comes from a different source, namely the IMF WEO. 8 The Balassa-Samuelson effect refers to the fact that price levels are higher in richer countries than in poorer countries. It can be shown that relative wages and relative prices are a function of the marginal productivity of labor in the traded goods. Given higher capital abundance in the richer countries, the productivity of labor tends to be higher in these countries, and hence will be wages and prices. 8 The positive association (Figures 4A, 4B and 4C) can also be interpreted as evidence of the validity of Wagner’s hypothesis at the cross-country level. This hypothesis, postulates that there is a tendency for governments to increase their activities as economic activity increases. Since 1890 Wagner postulated that economic development implied rising complexities that required more governmental activity, or that the elasticity of demand for publicly provided services, in particular education, was greater than one. This hypothesis has been tested econometrically (Chang, 2002; Afonso et. al. 2017) in time series and cross-country settings, showing that this is not particular to the series used for the present study. Figure 4. Public Expenditure on Education, Health and Capital and GDP (all per capita and in logs) 2009-2015 Figure 4A. Public Expenditure on Education and Figure 4B. Public Expenditure on Health and GDP GDP 2009-2015 2009-2015 Education Spending vs GDP Per Capita Health Spending vs GDP Per Capita 10 NOR KWT DNK LUX 8 ISL SWE SAU CHE FIN USA NLD AUT CYP NZL AUS CAN FRA DEU GBR OMN LUX SVN ITA ESP TTO NOR BWA EST MYS HUN KOR BHR DNKCHE USA LTU CZE POL LVA GRC SWE AUT DEU FRA NLD CAN ISL 8 BRB HRV SVK KNA NZLGBR FIN BLR MEX ARG VEN RUS ITAAUS ledu/Linear prediction lhea/Linear prediction CRI BRA CHL ESP VCT TUNZAF URY IRN SVN CZE KWT NAM DZAMUSROMATG TUR GRC ARE BLZ LCA ECU BGR THA GAB PAN SVK UKR SWZJAM COL MKD KAZ HRV HUN SAU GRD EST BHR 6 BOL FJI TKM CYP LTU KOR OMN MARMNG CPV EGY AZE LBN URY POL CRIPAN RUS BHSTTO TUR ROM LVA ATG MDA HND GUYPRY PER DOM ARG BRB BLR BGR CHL VNM WSM SLVIDN ALB CHN SLB LSO GHAYEM TON GTM JOR COL MKD BRA BWA LBN KNA ZAF DZA MEX IRNKAZMYS 6 DJI VUT ARM LKA SWZ LCA VCT THA NAM GRD TUN KGZ KEN AGOGEO UKR DMA SLV MUS VEN IND COG PHL WSM BLZ GUYPRY PER GAB CIV ALB JAM FJI DOM ZWESEN NIC MRT PAKNGA MDA TONCPV MNG BOL CHNECU EGY TZA CMR LAO LSO HND NIC GTM DJI LKA TKM AZE COMBEN NPL TJK VNM AGOARM MAR BFA MLI RWA SLB VUT UZB COGGEO PNG KGZ GHA 4 ETH TGOERI SDN ZMB PHL IDN MWI BGD MOZ UGA MDG ZMB YEM BDI TCD KHM MWI SDN KEN MRT 4 SLE GMB GMB TZA SEN NGA NER GIN HTI BFA MLI NPL BEN CIV IND GNB RWA ZWE KHM LAO PAK MDG UGA COM CMR MOZ BDI ETHSLE TCDTJK TGOERI HTI BGD NER GNB CAF GIN ZAR CAF ZAR 2 2 6 8 10 12 6 8 10 12 Log GDP Per Capita PPP 2011 Log GDP Per Capita PPP 2011 9 Figure 4C. Public Capital Spending and GDP 2007-2015 Public Investment Spending vs GDP per Capita ARE LUX 8 Public Investment Spending Per Capita OMN JPN AUS CHE NZL SGPNLD USA CAN FIN FRA SAU KWT ISLIRL VEN AUT TTO BHR KOR ITA BEL GBR MYS BRB CZE DEU EST PRTGRC HRV SYC DZA BWA GAB TUR POLSVK ISR MUS LTU 6 ROM LVA URY MEX RUS GUY ECU COL NAM KAZ THAMNE BIH CPVAZE PAN SUR CHL BLZ BGR CRI BTN AGO TUN SWZ IRN ARG ALB JOR DOM PER ZAF BRA BOL GEO LBN LSO PRY SRB VNM MAR ARM GHA NGA IND LKA EGYIDN SLV TCD NIC HND UKR GTM TJK SENZMB MDAPHL 4 ETH RWA MMR KHM MLI HTI KEN CIV MOZ BFA UGABEN YEM BGD CMR GMBTZA PAK ZWE MDG MWI SLE GIN NPL BDI 2 ZAR 4 6 8 10 12 lgdp Data Source: World Economic Outlook-IMF Previous studies that measured the efficiency of public spending recognized the positive association and suggested alternative solutions. One possibility is to split the sample by groups of countries (Gupta and Verhoeven, 2001). We may partially control for this factor by excluding the industrialized nations from the sample, or by presenting the results clustered regionally (Africa (AFR), East Asia and Pacific (EAP), Eastern Europe and Central Asia (ECA), Latin America and Caribbean (LAC), Middle-East and North Africa (MNA) and South Asia (SAS)). A second alternative incorporates directly the per- capita GDP as a factor of production, jointly with expenditure and other inputs (Jarasuriya and Woodon 2002). The problem with this approach is that it combines variables derived from a production function approach, and hence with clear interpretation, with others (GDP per capita) that are difficult to interpret from any viewpoint. When the two types of variables are combined, their effects cannot be disentangled. We attempt to control for this factor by including GDP per capita as an explanatory variable in the “second stage” regressions presented in the last section. A third option consists in using as an input the orthogonal component of public expenditure to GDP.9 We estimated the efficiency scores using as input both the original and the orthogonalized variables. The goodness-of-fit of each model was gauged based 9 The orthogonalized expenditure variable is the residual of the linear regression between pubic expenditure and GDP per capita. Since residuals may take positive and negative values, the variable was right-shifted to avoid negative values to facilitate graphical presentation of the frontiers. The orthogonalization of capital spending was done using a log-log regression, as this transformation of the variables resulted in a distribution of the efficiency scores that produced a better fitness observed by comparing Fig. 5B (log-log) and the figure reported in the Appendix without the log transformation. 10 on the frequency distribution of the inefficiency measures, as suggested by Farrell (1957) and Varian (1990). Comparing the distributions of the efficiency scores computed alternatively with the public expenditure or the orthogonalized value (Figures 5A and 5B), we see that the orthogonalization reduces the bias towards inefficiency (in the case of education, Fig. 5A) or towards efficiency (capital spending, Fig. 5B).10 Given this mitigation of the skewedness, the remainder of the paper considers the orthogonal component of expenditure as the input in all the efficiency score calculations. FDH FDH 15 8 6 10 Density Density 4 5 2 0 0 0 .2 .4 .6 .8 1 .4 .6 .8 1 score score Public Expenditure Orthogonalized Public Expenditure Figure 5A. Density of Efficiency Scores: Education-Gross Primary School Enrollment Efficiency Scores: Quality of the Overall Infrastruct Efficiency Scores: Quality of the Overall Infrastructure Unorthogonalized Public Investment Orthogonalized Public Investment 3 3 2 2 Density Density 1 1 0 0 .4 .6 .8 .2 .4 .6 .8 1 score score Public Expenditure Orthogonalized Public Expenditure Figure 5B. Density of Efficiency Scores – Capital Spending-Overall Quality of Infrastructure 10 The efficiency scores depicted in Figs. 5A and 5B are FDH efficiency scores, one for input efficiency (5A) and the other for output efficiency (5B). 11 This paper uses 14 indicators of education output, seven indicators of health output, and six indicators of infrastructure output.11 The education indicators are: primary school enrollment (gross and net), secondary school enrollment (gross and net), gross tertiary school enrollment, literacy of youth, average years of school, second level complete, PISA learning scores for math, reading and science and the WEF Indices for Quality of Match and Science Education, Quality of Primary Education, and Quality of the education system. Though the PISA country sample is limited, it is a numerical score. The WEF indices are computed for a significantly larger sample but have a limited range from 1-7. The correlation between the PISA learning scores and other output variables is high (.80 with average years of school and .78 with net primary school enrollment), as shown in Figure 6.12 The health output indicators are: life expectancy at birth, immunization (DPT13 and measles), and the disability-adjusted life expectancy (DALE), maternal survival rate, infant survival rates and Tuberculosis Free Population.14 For infrastructure output, we considered six indexes produced by the World Economic Forum: quality of overall infrastructure, quality of roads, quality of railroad infrastructure, quality of port infrastructure, quality of transport infrastructure, and quality of electricity supply. The cross-country comparisons with this set of indicators assume some form of data homogeneity, which might be problematic given the diversity of countries in the sample considered. Even for a more homogeneous group of countries, such as the OECD, there is call for caution when comparing expenditure levels in member countries (Jounard, et al. 2003). There is very little to do to overcome this limitation, except subdivide the sample into different groups. Probably a regional aggregation can be useful, but even at that level there may be extreme heterogeneity. 11 The data sources are: The World Bank World Development Indicators (WDI), Barro-Lee database, and Crouch and Fasih (2004), PISA, and the World Health Organization (Mathers et al, 2000). A complete list of variables and data sources can be found in Table A.10 of Appendix A. For capital spending the source is the WEO data base from the IMF and the infrastructure output indicators come from the World Economic Forum (WEF). 12 The correlation coefficients and Figure 6 exclude developed nations for the Crouch and Fasih (2004) sample. 13 DPT is Diphtheria-Pertussis and Tetanus 14 Maternal survival rates and infant survival rates are transformations of the mortality rates as follows: Maternal Survival Rate= (100000-Maternal Mortality Rate)/100000. The infant survival rate (ISR)=1000- Infant Mortality rate)/1000. Tuberculosis free population is a transformation of the incidence of tuberculosis variable (per 100,000 inhabitants). 12 Correlation: PISA Science Scores and Net Primary Enrollment Correlation: PISA Science and Average Years of School 550 550 FIN FIN ERI CAN CAN VNM KOR VNM KOR NZL NZL AUS NLD CHN CHN NLDSVNAUS DEU DEU CHE SVN GBR GBR CHE 500 500 POL POL USA DNK NOR DNK NOR CZE USA FRA SWE AUT FRA SWE LVA ESP LVA ESP HUN PISA Science Scores HUN PISA Science Scores LUX ITA LTU RUS ISL ITA LUX LTUISL RUS HRV HRV SVK GRC GRC 450 450 ARE ARE CHL CHL BGR TUR BGR TUR ARG ARG ROM CYP URY URY CYP ROM MDA MDA THA THA TTO MEX CRI CRI TTO MEX JOR JOR COL GEO ALB COL ALB 400 400 BRA BRA KAZ KAZ IDN TUN TUN IDN MKD LBN PER PER DZAPAN DZA PAN AZE 350 350 DOM KGZ DOM KGZ 40 60 80 100 0 5 10 15 Net Primary Enrollment Average Years of School Data Source: World Bank Indicators Data Source: World Bank Indicators & Barro Lee Figure 6. Correlation between Learning Scores and Other Education Indicators There are four other limitations arising from the variables and data selected for the analysis. The first one refers to the level of aggregation. While the exercises use aggregate public spending on health, education and capital spending as inputs, they use disaggregate measures of output, such as primary enrollment or secondary enrollment. Ideally, the input should have differentiated between primary and secondary education expenditures. Similarly, health care spending could be disaggregated into primary care level care and secondary level. The data can be disaggregated even further, by analyzing efficiency at the school or hospital levels. Second, there are omitted factors of production. This is especially true in education, as the paper did not consider private spending due to data constraints for developing nations. If this factor were used more intensively in a group of countries, then the efficiency scores (reported in the next section) would be biased favoring efficiency in that group. The third limitation arising from the data is the combination of monetary and non- monetary factors of production and outputs. The paper uses together with public expenditure, other non-monetary factors of production such as the ratio of teachers to students, in the case of education, or literacy of adults in the case of health and education. Other factors of production that could have been used were the physical number of teaching hours (in education) or the number of doctors or in-patient beds, as Afonso and St. Aubyn did for the OECD countries. However, inexistent data for many developing countries constrained the options. In the case of infrastructure output, the physical measures such as those used in Herrera-Pang for container ports are not available, hence we are limited to using indexes. A fourth limitation arising from the selected indicators is that these do not allow for a good differentiation between outputs and outcomes. For instance, most of the indicators of education, such as completion and enrollment rates, do not measure how much learning is taking place in a country. In education, this paper advances by considering the learning scores as one of the indicators. In health, other outcomes such as the number of sick-day leaves or the number of missed-school days because of health-related causes 13 could be better reflections of outcomes. Two of the selected health output indicators, DPT and measles immunization, are delivered in vertical programs, that is, in campaigns that are relatively independent of basic health systems and therefore may not be good indicators of the actual quality of the health system. Additionally, in most countries, these two activities (immunization) account for very small fractions of the health budgets. Finally, the fact that life expectancy is influenced by diet, lifestyle, and a clean environment, that to the extent that are not included as factors of production may bias the efficiency scores. III.2. Single Input-Output Results III.2.1. FDH and DEA Education Frontiers Figures 7a-d show both FDH and DEA efficiency frontiers for four of the 14 output indicators: gross primary school enrollment, secondary gross enrollment, quality of math and science education index (from WEF) and the PISA Science scores. Individual country efficiency scores for the four indicators are reported in Table A.1-2 of Appendix A.15 The graphical efficiency frontiers for other output indicators can be found in Appendix A (Figure A.1). Several results may be highlighted: a. The input efficiency rankings (education) are robust to the output indicator selected, as implied by the Spearman rank-correlation coefficient (see Tables A.1 and A.3 in Appendix A), which is positive, significant and high. The range oscillates between .35 and .97. This result implies that countries ranked as efficient (or inefficient) per one indicator, are ranked similarly when other output indicators are used. b. Despite the orthogonalization by GDP, the relatively rich countries tend to be in the more efficient group, especially when output efficiency is considered. c. In general, output efficiency scores are higher than input efficiency scores. 15 The efficiency scores of all the indicators can be found at the MFM website indicated in footnote 1. 14 Free Disposable Hull (FDH) Data Envelopment Analysis (DEA) Gross Primary Enrollment vs Education Expenditure Gross Primary Enrollment vs Education Expenditure MDG MDG 140 140 MWI MWI NPL NPL RWA RWA Gross Primary School Enrollment Gross Primary School Enrollment BDI BDI SLE SLE TGO TGO NIC NIC 120 120 BRA KHM BEN BRA KHM BEN VUT LAOCOL VUT LAOCOL AGO AGO SLV SLV UGA UGA ECU MAR SLB GEO ECU MAR ARG SLBCRI GEO MNG GNBARGKEN HNDSWZ CRI PHLMNG GNB KEN CPVHND COG SWZ BLZ PHL URY COG BGD CPV BLZ BGD GTM URY KAZ CHNGTM CMR LSO TUN NAM BWA SWE KAZ CHN CMR LSO IND TUN NAM BWA SWE IRN IND IRN OMN IDN GHA OMN IDN VNM GHA NLD FRA GBR VNM TTO NLD GBR FRA PAN ESP PER AUS EGY ALB FJI TTO MOZ PAN ESP DOM PER AUS EGY ALB FJI MOZ COM UKR DOM DEU PRY COM VCT MEX KGZ UKR PRY MEX VCT CHE LVA SAU DEU SAU SVK MUS 100 CHE MUS LVA KGZ ITA TUR VEN CHL MYS 100 ITA SVK BGR TUR CHL CZE HUN KOR LTU EST BLR ZWE BOL AUT CYP DNK CZE HUN KOR LTU ARMUSAVEN EST BLR MYSZWE AUTBOL CYP FIN DNK ATGAZE RUS LBN LKA BGR ARMUSA POL CAN TJK ZAR NZL FIN NOR ATGAZE RUS LBN LKA THA POL CAN ZAR TJKSVN NZL ISL NOR THA MRTETH SVNBRB ZAF ISL MRTETH BRB ZAF ROM HRV ROM HRV MDA PAK CAF MDA PAK CAF TZA TZA TCD TKM NGA TCD TKM NGA MKD GIN KNA MKD GIN KNA GUY GMB GUY GMB CIV CIV BFA BFA SEN 80 SEN 80 MLI MLI SDN SDN NER NER DJI DJI 60 60 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI Data Source: World Bank WDI (a.1) (a.2) Gross Secondary Enrollment vs Education Expenditure Gross Secondary Enrollment vs Education Expenditure 100 100 CHN CHN ESPKOR BLR FRA NZL BLR FRA NZL ITA HRV LTU SWE NOR ITA ESPKOR HRVLTU SWE NOR KAZ GBR GBR HUN EST SVN CYP FIN KAZ HUN EST SVN CYP FIN POL Gross Secondary School Enrollment OMN OMN POL Gross Secondary School Enrollment GEOARM NLD LVA BRB DNK GEO NLD LVA BRB DNK USA CHL ISL ARM USA CHL ISL BGR AUS ARG VCT UKR VCT UKR ATG ROM LKAMUS ALB BGR AUS ARG KNA CHE TUR FJI TJK ATG KNA ROMLKAMUS ALB EGY LCA CHE TUR FJI TJK EGYLCA 80 THA KGZ BRA SAU 80 THA KGZ SAU IRN PER GUY ECU COL MDA IRN PER GUY ECU BRA COL CRI MDACRI URY BOL JAM BOL JAM IDN VEN URY IDN VEN LBNPAN CPV BLZ LBNPAN SLV MEX MYS CPV BLZ SLV MEX MYS DOM PHL PRY DOM PHL PRY IND 60 IND 60 NPL NPL MAR MAR VUT KEN GHA VUT KEN BGD NIC HND BGD GHA NIC HND GTM LAO GTM COM ZWE BEN LAO COM ZWE CMR BEN 40 CMR 40 PAK PAK SWZ LSO SWZ MLI LSO GIN MWI ETH MLI GIN MWI MDG ETH RWA MDG RWA MRT MRT 20 BFA 20 BDI BFA MOZ BDI MOZ NER AGO CAF NER AGO CAF 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI Data Source: World Bank WDI (b.1) (b.2) Quality of math and science vs Education Expenditure Quality of math and science vs Education Expenditure FIN FIN 6 6 CHE CHE LBN LBN BRB NZL BRB NZL NLD Quality of math and science NLD Quality of math and science CAN CYP CAN CYP FRA SVN FRA SVN KOR KOR LTU ESTMYS TUN LTU ESTMYS TUN ISL 5 5 ISL HRV HRV DEU AUS DNK DEU AUS DNK TTO SWE TTO UKR SWE IRN IND AUTUKR ROM IRN CHN IND AUT ROM CHN HUN LKA HUN LKA IDN LVA IDNPOL ALB USA LVA GBR CRI RUS CZE ALBPOL USA GBR CRI RUS CZE NOR MUS SAU NOR MUS MKD VNM CIV SAU MKD MNG VNM CIV BGRTHA MNG RWA BGRTHA RWAZWE ITA ZWE ITA SVK ARMGUY MAR SVK ARMGUY MAR BENSEN MDA KEN 4 BEN SEN MDA KEN 4 CMR KAZ CMR KAZ OMN OMN LAO MDG GHA BWA LAO MDG BFAGHA BWA BFA GMB GMB CPV SWZ PHL ESP TUR PAK NPL CPV SWZ ESP PHL TUR PAK NPL ETH GEO ETH AZE GEO COLECU TJKBDILSO AZE COLECUTJK BDILSO MWI KHM MWI BLZ KHM BGD GIN BLZ KGZ BGD GIN UGA ARG KGZ UGA ARG URY JAM CHL URY JAM BOL GAB CHL NGA TCD BOL GAB NGA TCD 3 VEN 3 PAN VEN SLVMRT MLI PAN SLVMRT MLI NAM HTI NAM HTI MEX MOZ MEX MOZ HND HND TZABRA NIC TZABRA NIC SLE SLE GTM EGY GTM EGY PER PER PRY PRY DOM DOM 2 ZAF 2 ZAF AGO AGO 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/World Economic Forum Data Source: World Bank WDI/World Economic Forum (c.1) (c.2) 15 Pisa Science Literacy score vs Education Expenditure Pisa Science Literacy score vs Education Expenditure 550 550 FIN FIN KORCAN VNM KORCAN VNM NZL NZL AUS AUS DEU CHN NLD SVN DEU CHN NLD SVN GBR CHE GBR CHE 500 Pisa Science Literacy score 500 Pisa Science Literacy score POL POL USA NOR DNK USA NOR DNK CZE FRA AUT SWE CZE FRA AUT SWE ESP HUN LVA ESP HUN LVA ITA RUS ISL ITA RUS ISL HRVLTU HRVLTU SVK SVK 450 450 CHL CHL BGR TUR BGR TUR ARG ARG ROM URY CYP ROM URY CYP MDA MDA THA THA CRI TTO CRI TTO MEX MEX GEO GEO ALB COL ALB COL 400 400 BRA KAZ BRA KAZ IDN TUN IDN TUN LBN LBN MKD PER MKD PER PAN PAN AZE AZE 350 350 DOM DOM KGZ KGZ 1000 1500 2000 2500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/NCES Data Source: World Bank WDI/NCES (d.1) (d.2) Figure 7 Education Efficiency Frontier: Single Input and Single Output d. Output-efficiency rankings also vary with the selected output indicators. The Spearman correlation coefficient of the output-efficiency scores (see Tables A.2 and A.4 in Appendix) show that these are robust to the selected indicator, though the range is higher (.16 to .90). e. To grasp the order of magnitudes of the deviations from the efficiency frontier, and the disparity in efficiency across countries, we computed an average for all indicators for the groups of more efficient and least efficient countries (Table 1).16 The input-efficiency estimations indicate that a country in the bottom of the distribution could reach the same educational attainment levels by spending approximately 50 percent less. The output efficiency estimators indicate that, on average, with the prevailing expenditure level a typical country of this group could reach educational attainment levels more than twice as high (Table 1A). Table 1. Education Attainment: Single Input, Single Output Country Clustering Input-Efficient Output Efficient More efficient Italy, Kazakhstan, Lebanon, Austria, Finland, Kazakhstan, Czech Republic, Slovak Italy, Canada, Lebanon, Republic, China, Republic of Czech Republic, Great Korea Britain, Sweden, Republic of Korea Least efficient Denmark, Norway, Sweden, Angola, Mozambique, Saudi Arabia, Namibia, Burundi, Mauritania, Botswana, Moldova, Costa Tanzania, Peru Rica 16 These are simple averages of the top 10 and bottom 10 countries in each ranking. 16 Table 1A. Education Attainment: Single Input, Single Output Efficiency Scores Input-Efficient Output Efficient More efficient .89 .92 Least efficient .47 .39 f. This clustering exercise reveals (Table 1) that several developed economies, such as Denmark, Norway and Sweden, appear within the least input-efficient category. This is due to the amount of spending these countries do. Resource-rich countries also appear in the inefficient group, such as Saudi Arabia, Botswana, Angola, and Mozambique. Mostly African countries populate this group. Among the more efficient group, Lebanon, Republic of Korea, China, Czech Republic, and Kazakhstan appear as both input and output efficient. Curiously, Sweden appears within the most output-efficient category. Explaining why these countries appear in each cluster requires more in-depth analysis, undertaken in the last section, which explores the association of efficiency scores with other variables. g. It is critical to note that even if a country appears as efficient, there might still be a significant discrepancy between the observed output level and the desired or target output level. For instance, Azerbaijan, Lebanon and Romania appear as efficient countries on the FDH efficiency frontier or very close to it (Figure 7 a.1) because they spend relatively little on education; they are close to the origin of the frontier with low spending and low learning levels. But their PISA scores are far below those of countries that spend similar amounts but have significantly higher PISA scores, such as Italy or the Czech Republic. Within the LAC countries, Panama and Chile have the highest input efficiency scores, though Panama’s case is due to the low spending level effect discussed above. The important point is that the country moves along the efficiency frontier to the higher target output level. Countries can even improve efficiency by exploiting scale economies if they are operating in the increasing returns to scale zone of the production possibility frontier (output levels smaller than that of point A, Figure 3). h. The regional aggregation of the efficiency scores shows that input-oriented scores (Table 2A) are lower than output-oriented scores (Table 2B).17 This is especially true in ECA, EAP, and MNA. In general, scores are higher for primary enrollment and decrease with the level of education (secondary enrollment), especially when output-oriented measures are considered. Across regions, ECA, EAP, and MENA have similar levels of input-inefficiency of about 65-70 percent, while LAC and SAS average around 60 percent, and AFR about 55 percent. 17 The regional aggregate is the simple average of the individual country scores obtained for the whole sample. The scores were not computed by constructing separate efficiency frontiers for each region. The regional classification exercise excludes the developed economies in all the regional aggregation exercises. 17 i. Output efficiency in ECA and EAP is in the range of 75 to 80 percent, in LAC and MNA around 70 percent, in SAS about 60 percent and in AFR 50 percent.18 In EAP, ECA, and LAC the output efficiency scores are about 15 percent higher than the input efficiency, while in SAS they are about the same and in AFR 7 percent lower. This indicates that the first three regions are closer to the frontier on the output side than on the input (spending) side. Table 2A. Educational Attainment: Input-Efficiency scores by regions across the world - Single Input, Single Output AFR EAP ECA LAC MNA SAS Gross primary enrollment .61 .65 .67 .65 .65 .70 Net primary enrollment .58 .72 .73 .70 .75 .69 Gross Secondary enrollment .57 .68 .66 .63 .70 .63 Net secondary enrollment .56 .61 .66 .63 .65 .62 Average years of school .56 .64 .69 .60 .59 .63 Second level complete .56 .63 .68 .60 .59 .63 PISA Science .82 .79 .68 .74 PISA Reading .78 .78 .68 .74 PISA Math .80 .79 .67 .74 Quality of Math and Science .57 .64 .69 .60 .69 .64 Table 2B. Educational Attainment: Output-Efficiency scores by regions across the world - Single Input-Single Output AFR EAP ECA LAC MNA SAS Gross primary enrollment .72 .79 .75 .77 .76 .79 Net primary enrollment .81 .93 .94 .94 .92 .90 Gross Secondary enrollment .34 .73 .89 .74 .77 .59 Net secondary enrollment .33 .59 .75 .68 .63 .49 Average years of school .42 .65 .86 .65 .59 .51 Second level complete .20 .39 .76 .41 .35 .32 PISA Science .90 .87 .77 .79 PISA Reading .88 .85 .79 .74 PISA Math .88 .86 .74 .78 Quality of Math and Science .57 .72 .72 .53 .73 .66 j. Two groups of indicators capture the quality of educational attainment: the PISA scores (Math, Science and Reading) and the World Economic Forum (WEF) various indicators, of which we focus on one, the Quality of Math and Science 18 These are simple averages of the indicators by region. In output efficiency the primary and secondary level completed indicators were omitted due to problems with the data. 18 education. Both variables have limitations: the PISA country sample is smaller and with relatively greater weight of developed nations, while the WEF variable is an index constructed to reflect the quality in a large sample of countries. III.2.2. FDH and DEA Health Frontiers This section presents the single input (public expenditure on health per capita in PPP terms)-single output efficiency frontiers using both the FDH and DEA methodologies; seven alternative output indicators are used: life expectancy at birth, the disability adjusted life expectancy (DALE), DPT immunization, measles immunization, tuberculosis free population, maternal survival, and infant survival rates.19 Figures 8a-d show the FDH and DEA efficiency frontiers for four indicators. The specific country rankings for four of the health indicators are listed in Table A.4-5 of Appendix A. The graphical efficiency frontiers for other output indicators can be found in Appendix A (Figure A.2). Free Disposable Hull (FDH) Data Envelopment Maternal Survival Rate vs Health Expenditure 100000 Maternal Survival Rate vs Health Expenditure 100000 SAU OMNBHR CYPKOR RUS KAZ LVA POL BLR EST LTU LBN BGR TUR MKD HUN SVK LUX URY GRC CZE AUS HRVESP SVN FIN ISL ITA CHE CAN GBR AUT DEU FRA USANOR NZL SWE DNK NLD SAU OMNBHR CYPKOR RUS KAZ AZE LVA CHL IRN LBNPOL BLR THAEST LTU BGR TUR MKD HUN ARM SVK LUX URY UKR GRC CZE AUS HRV MDA ESP SVN CRI FIN ISL ITA CHE CAN GBR AUT NOR DEU FRA USANZL SWE DNK NLD MYSAZE CHL IRN TKM THA MEX ARM CHN GRD LKAALB ROM EGY BRB FJI GEO UKR BLZ TJK UZB CPV MDACRI MYS TKMMEX CHN GRD LKAALB ROM EGY GEO BRABRB FJI MNG VCT LCABLZ UZB CPVTJK TTO MUS ARG BRA MNG VCT LCA TUN ECU SLV VNM COLWSM JOR TTO MUS ARG TUN ECU PER SLV VNM COLWSM JOR BHS DOM PER GTM JAM KGZ VUT BHS DOM VEN GTM JAMPAN KGZ VUT VEN PHL PAN SLB PHL MAR TON SLB IDNBWAMAR ZAF DZA TON PRYHND IDNBWAZAF DZA PRYHND NIC KHM NIC KHM IND PAK BGD IND PAK LAOBGD LAO BOLPNG BOL ZMB GUYPNG DJI ZMB GUY DJI NPL NAM NAMNPL GAB RWA GAB SDN RWA SDNSEN GHA SEN GHA COM COM UGA ETH MDG UGA ETH MDG HTI HTI TGO BFA TGO BFA Maternal Survival Rate SWZ Maternal Survival Rate TZASWZ TZA BEN 99500 BEN 99500 COG ZWE COG ZWE AGO MOZ AGO MOZ ERILSO KEN ERILSO KEN GNB NER GNB NER MLI CMR MRT MLI CMR MRT CIV MWI CIV MWI GIN GIN ZAR ZAR GMB GMB BDI BDI NGA NGA TCD TCD CAF CAF 99000 99000 SLE SLE 98500 98500 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI, World Health Organization Data Source: World Bank WDI, World Health Organization   (a.1) (a.2) 19 Tuberculosis free, maternal survival rates, and infant survival rates are transformations of the original variables, tuberculosis incidence, maternal mortality, and infant mortality, respectively. Transformations are described in footnote 15. 19 1 Infant Survival Rate vs Health Expenditure Infant Survival Rate vs Health Expenditure 1 CYP KOR EST LUX CZE GRC SVN ESP ISL FIN ITA DEUNOR AUT SWE CYP KOR EST LUX CZE GRC SVN ESP ISL FIN ITA NOR AUT SWE BLR LTU POL HUN SVK AUS HRV CHE GBR CAN FRA D NL BLR LTU POL HUN SVK AUS HRV CHE GBR CAN DEU FRA DNK NLD BHR MYS LVA ATG CHL USA NZL BHR MYS LVA ATG CHL USA NZL KNA RUS LBN LKABGR ROM URY CRI UKR KNA RUS LBN LKABGR ROM URY CRI UKR OMN BHS ARG MKD OMN BHS ARG MKD SAU MUS MEXGRDCHN THA BRB GEOLCA SAU THA MUS MEXGRDCHN BRB GEOLCA ALB PER TUN MDA ALB PER TUN KAZ VEN IRN TUR ARM BRA JAM COL BLZTON SLV PAN WSM JOR KAZ VEN IRN TUR ARM BRA JAM TON COL BLZ SLV PAN MDA WSM JOR VCT VCT .98 .98 TTO VNMHND TTO VNMHND MNG FJI PRY ECU CPV NIC MNG FJI PRY ECU CPV NIC EGY PHL DZA KGZSLB EGY PHL DZA KGZSLB IDN DMA VUT IDN DMA VUT DOM MAR GTM UZB DOM MAR GTM UZB Infant Survival Rate Infant Survival Rate GUY GUY AZE AZE KHM BOL NPL KHM BOL NPL BGD BGD BWAZAF NAM ERI RWA BWAZAF NAM ERI RWA .96 .96 GAB SEN MDG KEN GAB SEN MDG KEN IND COG TJK IND COG TJK YEM YEM TZA GMB TZA GMB UGA GHA PNG UGA GHA PNG TKM SDN ETH MWI TKM SDN ETH MWI ZMBZWE ZMBZWE LAO TGOBDI LAO TGOBDI NER .94 .94 MRT DJI SWZ MRT NER DJI SWZ HTI BFA COM HTI BFA COM CMR MOZ CMR MOZ GIN GNB GIN GNB AGO BEN AGO BEN PAK PAK CIV CIV NGA MLI LSO NGA MLI LSO .92 .92 ZAR TCD ZAR TCD CAF SLE CAF SLE .9 .9 1000 2000 3000 4000 1000 2000 3000 4000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI, World Health Organization Data Source: World Bank WDI, World Health Organization   (b.1) (b.2) Free Tuberculosis vs Health Expenditure Free Tuberculosis vs Health Expenditure BRB BRB 1 ISL USA 1 OMNBHR SAU CYP TTO CHL IRN MUS LBN MEX ARG ALB EGY TUR HUN POL LUX JAM MKDSVK URYJOR AUS GRC CZE SVN CRI HRVESPITA FIN CAN CHE GBR DEU NOR AUT FRA NZL SWE DNK NLD OMNBHR SAU CYP TTO CHL IRN MUS LBN MEX ARG ALB EGY TUR HUN POL LUX JAM MKDSVK JOR AUS URYGRC CZE SVN CRI HRVESP ISL ITA FIN CAN CHE GBR USA DEU NOR AUTNZL FRA NLD DNK SWE VEN LVA EST BGR BRA TUN SLV COL BLZ PAN PRY YEMNIC VEN LVA EST BGR BRA TUN SLV COL BLZ PRY PAN YEMNIC LKA DOM LTU GTM ARM ECU DZA HND LKA DOM LTU GTM ARM ECU DZA HND KOR RUS CHN MYS AZE MAR GEO ROM PER UKR GUY BEN KOR RUS CHN MYS AZE MAR GEO ROM PER UKR GUY BEN KAZ THA GHA BOL CPV BFA KGZ KAZ THA GHA BOL CPV BFA KGZ IND NPL VNM MDA MLI IND NPL VNM MDA MLI IDNMNG LAO CMRRWA TJK TZA IDNMNG LAO RWA TJK TZA CMR NGAPAK GIN SEN HTI BGD TCDBDI UGA NGAPAK GIN SEN HTI BGD TCDBDI MWI MDG GMB UGA MWI MDG GMB PHL CIV ETH PHL CIV ETH AGOMRTKEN AGO KEN MRT KHM KHM .995 ZMB ZMB .995 GAB GAB Free Tuberculosis Free Tuberculosis MOZ MOZ BWA BWA SLE SLE ZWE ZWE NAM LSO NAM LSO ZAF ZAF .99 .99 SWZ SWZ .985 .985 1000 2000 3000 4000 500 1000 2000 3000 4000 5000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI, World Health Organization Data Source: World Bank WDI, World Health Organization   (c.1) (c.2) Disability Adjusted Life Expectancy vs Health Expenditure Disability Adjusted Life Expectancy vs Health Expenditure AUS ESP ITA LUX ISLCHE FRA SWE AUS ESP ITA LUX ISLCHE FRA SWE CAN NZLNOR NLD CAN NZLNOR NLD CRI FIN GRC AUT CRI FIN AUT 80 GRC 80 GBR DEU GBR DEU CYP KOR CHL SVN DNK CYP KOR CHL SVN DNK LBN PER PAN LBN PER PAN TUR USA TUR USA CZE Disability Adjusted Life Expectancy CZE Disability Adjusted Life Expectancy NIC NIC OMN ATG POLECUTUN BRB COL DZA ALB URY HRV OMN ATG POL TUN ECU BRB COL DZA URY HRV MEXTHA ARG LKA THA ARG MEXLKAALB SAU BHR EST LCASVK SAU BHR EST LCASVK MYS VENIRN DOM CHN JAM HUN SLV PRY JOR MYS VENIRN DOM CHNJAM HUN SLV PRY JOR BRA ARM ROM BGR MKD MAR BRA ARM ROM BGR MKD MAR BHS LTU MUS VNM BHS LTU MUS VNM TTO LVA GEO DMA CPV TTO LVA GEODMA CPV BLR GRD GTM VCTBOL HND WSM BLR GRDGTM VCTBOL HND WSM EGY BLZ EGY BLZ BGDMDA BGD MDA 70 70 AZE IDN UKRTON TJK AZE IDN UKRTON TJK PHL PHL KAZ RUSTKM UZB MRT KGZ NPL KAZ RUSTKM UZB MRT KGZ NPL INDSDN GUY KHM INDSDN GUY KHM MNG YEM PAK COM GMB MNG YEM PAK COM GMB FJI FJI LAOSEN DJI LAOSEN DJI GAB RWA GHA GAB RWA GHA VUT KEN VUT KEN ERI BEN SLB ERI BEN SLB NGA MDG ETH HTI NGA MDG ETH HTI AGO AGO 60 60 COG TZA TGO COG TZA TGO MLI PNG NER MLI PNG NAM ZAR GIN NAM NER ZAR GIN BFA UGA BFA UGA BWA CMR TCDBDI BWA CMR BDI ZAF CIV MOZ GNB TCD ZAF CIV MOZ SLE GNB SLE MWI MWI ZMB ZWE ZMB ZWE 50 50 SWZ SWZ CAF CAF LSO LSO 1000 2000 3000 4000 500 1000 2000 3000 4000 5000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI, Institute for Health Metrics and Evaluation (IHME) Data Source: World Bank WDI, Institute for Health Metrics and Evaluation (IHME)   (d.1) (d.2) Figure 8: Health Efficiency Frontiers Single Input-Single Output 20 Several results may be highlighted: a. The input-efficiency rankings of the health indicators are positively and significantly correlated, with the Spearman rank correlation coefficient oscillating between .19 and .93 (see Tables A.1 and A.3 in Appendix A). This indicates that the efficiency ranking is similar regardless of the output indicator being used. b. Despite the orthogonalization by GDP, the richer countries tend to be more efficient (Table 3). The inefficient group has two types of countries: one small group of rich countries like the USA, Great Britain, and Switzerland which have high expenditure levels and not extremely high output (input inefficiency) and another much larger group of countries with relatively low spending but their output indicators could be substantially larger, like Lesotho, Malawi, and Sierra Leone. Other large African countries, Nigeria and Mozambique, appear in this group as well. Table 3. Health Attainment: Single Input, Single Output Input-Efficient Output Efficient More efficient Bahrain, Saudi Arabia, Oman, Oman, Saudi Arabia, Bahrain, Luxembourg, Trinidad and Luxembourg, Switzerland, Tobago, Republic of Korea, Czech Republic, Finland, Italy Australia Least efficient Burundi, Mozambique, Central African Republic, Democratic republic of Congo, Sierra Leone, Nigeria, Lesotho, Malawi, Switzerland, Lesotho, Malawi, United States, Great Britain Mozambique, Democratic Republic of Congo, Zimbabwe Table 3A. Health Attainment: Single Input, Single Output Efficiency Scores Input-Efficient Output Efficient More efficient .85 1.0 Least efficient .38 .77 c. The health efficiency rankings are closely correlated: the Spearman coefficient of output-efficiency ranking is .34 and the output one is .93 (see Tables A.2 and A.4 in Appendix A). d. To examine the efficiency differences across groups of countries and the potential efficiency gains of moving closer to the frontier, countries were clustered into the least and most efficient (Tables 3 and 3A). The typical country of the least efficient group could maintain the same output levels with about 50% of the expenditure level. Or on the output side, the least efficient group could increase output levels by 35% for a given expenditure level (Table 3A) The divergence 21 between least and most efficient groups in output-efficiency is not as large as in education. e. Output-efficiency scores are notoriously higher than the input-efficiency ones (Tables 4A and 4B). While the input-efficiency regional average oscillates between 40 and 50 percent, the output-efficiency average fluctuates between 90 and 95 percent, except in AFR where it is 85 percent. The large difference between the health input and output efficiency did not happen in the case of education, nor in previous studies based on data from the early 2000s. Comparing these results with efficiency scores of the early 2000s (Herrera-Pang, 2005), there seems to be a stagnation of input efficiency while output efficiency increased significantly, especially in AFR and SAS, which had the lowest achievement levels in the earlier studies. Table 4A. Health Input-Efficiency scores by regions across the world - Single Input, Single Output AFR  EAP  ECA  LAC  MNA  SAS  Immunization, measles   .42  .46  .46  .46  .58  .44  Immunization  .42  .55  .50  .47  .63  .47  Life Expectancy at birth  .42  .48  .48  .48  .60  .44  DALE  .49  .54  .54  .56  .65  .50  Free Tuberculosis Cases  .42  .48  .48  .48  .61  .44  Infant Survival Rate  .42  .48  .56  .47  .59  .44  Maternal Survival Rate  .42  .46  .46  .46  .59  .44  Table 4B. Health Output-Efficiency scores by regions across the world - Single Input-Single Output   AFR  EAP  ECA  LAC  MNA  SAS  Life Expectancy at birth  .72  .87  .90  .90  .91  .85  Immunization DPT  .82  .87  .95  .93  .94  .89  Immunization Measles  .80  .87  .96  .93  .93  .85  DALE  .73  .85  .90  .91  .92  .85  Tuberculosis free  .99  1.0  1.0  1.0  1.0  1.0  Infant survival rate  .95  .98  .99  .98  .98  .96  Maternal Survival rate  1.0  1.0  1.0  1.0  1.0  1.0  22 III.2.3. FDH and DEA Infrastructure Efficiency frontiers This section presents the single input (public capital expenditure per capita)- single output frontiers, using both the FDH and DEA methodologies, with six alternative output indicators: quality of overall infrastructure, quality of roads, quality of railroad infrastructure, quality of port infrastructure, quality of transport infrastructure, and quality of electricity supply. Figures 9a-c show the FDH efficiency frontier for three indicators. The specific country scores for all the infrastructure indicators and the DEA frontiers are can be found in Appendix A.7. Free Disposable Hull (FDH) Data Envelopment Quality of Overall Infrastructure vs Public Investment Quality of Overall Infrastructure vs Public Investment 7 7 CHE CHE SGP SGP FIN FIN FRA AUT FRA DEU AUT ARE DEU ARE ISL Quality of Overall Infrastructure ISL Quality of Overall Infrastructure NLD 6 NLD LUX JPN 6 LUX JPN PRT CAN USA PRT CAN USA BEL BEL BRB KOR BRB KOR MYS MYS BHR BHR GBR OMN GBR OMN CHL NAM EST AUS SAU CHL NAM EST AUS SAU 5 JOR LTU CZE 5 JOR LTU CZE ISR HRV TUR SYC NZL ISR HRV NZL TUN MUS KWTTHA PAN TUR TUN SYC THA SLV ZAF IRL LVA LKA RWA BTN SLV IRL MUS KWT PAN GRC AZE GRC ZAF LVA LKA RWA BTN BWATTO AZE GTM GMB GTM BWATTO MAR SVK SWZ GEO SUR MAR GMB IRN SWZ ITA 4 SVK GEO SUR URY MEX KAZ IRN ITA UKR ARMCIV 4 URY MEX KAZ UKR ARMCIV EGY KEN IDN HND RUS GHA KHM IND CPV EGY KEN CPV DOM BLZ SEN POL ALB TJKGUY ECU DZA IDN HND RUS GHA KHM IND MDA DOM BLZ SEN POL ALB TJKGUY ECU DZA PHLPAK CRI MLI MNE COL PAK MDA BRA ZMB ARG PER UGA ETH PHL CRI MLI MNE COL ZWE BGR VNM LSO BRA ZMB PER UGA ETH MWI NIC GAB 3 ZWE ARG BGR SRB TZA BOL MWI NIC VNM GAB LSO MDG CMR SLE BEN ROM 3 SRB TZA BOL YEM MOZ VEN MDG CMR SLE BEN ROM NGA BGD BFA YEM MOZ VEN NPL NGA BGD BFA BDI BIH NPL PRY MMR TCD BDI BIH PRY GIN AGO MMR TCD HTI 2 GIN AGO HTI 2 0 .5 1 1.5 2 2.5 0 .5 1 1.5 2 2.5 Log-Log Form: Orthogonalized Public Investment Per Capita Log-Log Form: Orthogonalized Public Investment Per Capita Data Source: World Economic Outlook-IMF/World Economic Forum Data Source: World Economic Outlook-IMF/World Economic Forum     (a.1) (a.2) Transport Infrastructure vs Public Investment Transport Infrastructure vs Public Investment 7 7 ARE SGP ARE SGP FRA DEU FRA DEU NLD NLD JPN 6 JPN 6 CAN KOR CAN KOR USA USA GBR CHE Transport Infrastructure GBR CHE Transport Infrastructure MYS FIN MYS FIN AUS AUS BEL BEL PRT AUT PRT AUT 5 ISL 5 ISL BHR THA SAU BHR THA SAU BRB LUX TUR OMN BRB LUX TUR OMN ITA ITA ZAF IND NZL PAN ZAF IND NZL PAN IRL IRL CHL MEX CHL MEX IDN RUS NAM IDN RUS NAM LTU CZE LTU CZE ISR MARLKA MUS EST 4 MUS EST SYC ISR MARLKA GRC LVA HRV 4 GRC LVA HRV SYC KWT GMB AZE KWT AZE SLV DOM JOR TUN TTO SLV DOM JOR GMB TUN TTO EGY GEO EGY GEO BRA KEN SWZ RWA ECU BRA KEN SWZ RWA ECU PAK IRN CIV SVK GTM VNM CPV BWA ETH GTM PAK IRN CIV SVK UKRPHL URY KAZ SUR UKRPHL URY KAZ VNM CPV BWA ETH SUR HND ARMARG SEN POL MNE PER ALB HND SEN POL MNE GHA BLZ BGR KHM GUY BTN ARG ALB 3 ARM PER GHA BLZ BGR KHM COL GUY BTN ZWE ZMB NIC MLI TJK DZA 3 ZWE NIC MLI COL TJK DZA CRI CMR BGD ZMB CRI CMR BGD TZA BOL TZA BEN UGA ROM BOL MDG NGA MWI BEN UGA ROM GAB MOZ NGA MWI MOZ YEM SRB MDA BDI BFA LSO MDG MDA GAB SLE VEN YEM SRB SLE BDI BFA LSO NPL PRY VEN TCD AGO NPL PRY TCD AGO GIN MMR BIH 2 GIN MMR HTI BIH 2 HTI 0 .5 1 1.5 2 2.5 0 .5 1 1.5 2 2.5 Log-Log Form: Orthogonalized Public Investment Per Capita Log-Log Form: Orthogonalized Public Investment Per Capita Data Source: World Economic Outlook-IMF/World Economic Forum Data Source: World Economic Outlook-IMF/World Economic Forum     (b.1) (b.2) 23 8 Quality of Electricity Supply vs Public Investment Quality of Electricity Supply vs Public Investment 8 CHE ISL FIN NLD SGP CHE ISL FIN NLD DEU BEL GBR AUT FRA CAN ARE DEU BEL GBR AUT FRA SGP CAN CZELUX JPN JPN ARE CZELUX Quality of Electricity Supply BRB USA Quality of Electricity Supply IRL PRT IRL PRT BRB USA SVK KOR AUS SAU OMN SVK KOR AUS SAU OMN 6 6 ISR BTN MYS ISR BTN ITA URY JOR BHR NZL URY BHR NZL MYS CHL CRI LTU TUN EST ITA CRI JOR HRV CHL LTU HRV TUN EST LVA MAR POLBIHTHA NAM MUS TTO LVA POLBIHTHA NAM MUS TTO GTM GRC PAN COL MAR PAN IRN KWT SYC GTM GRC IRN KWT COL SLV GEO SLV SYC GEO BRA SRB LKA PER BRA LKA PER UKR ARM RUS TUR SRBARM EGY KAZ AZEDZA UKR RUS TUR AZEDZA MDA MEX ROM EGY KAZ MEX ROM BLZ RWA MDA GMB BLZ RWA 4 CIV SWZ MNE BOL GMB 4 PHL IDN HND BGR PHL IDN HND CIV SWZ MNE BGR BOL ALB SUR BWAECU ALB SUR ECU ZAF KEN VNM LSO ZAF KEN VNM BWA LSO ZMB ARG MLI MOZ ZMB MLI NIC IND ETH ARG IND MOZ PRY KHM NIC ETH GHA GUY PRY GHA KHM CMR MMR MMR GUY CMR MWI BEN UGA BFA GAB VEN BEN UGA VEN PAK MDG TZA SEN TJK MWITZA SENBFA GAB SLE BDI CPV PAK MDG BDI TJK SLE CPV 2 DOM BGD 2 ZWE ZWE DOM BGD NPL HTI HTI NGA TCD AGO NPL NGA AGO GIN YEM GIN YEM TCD 0 0 0 .5 1 1.5 2 2.5 0 .5 1 1.5 2 2.5 Log-Log Form: Orthogonalized Public Investment Per Capita Log-Log Form: Orthogonalized Public Investment Per Capita Data Source: World Economic Outlook-IMF/World Economic Forum Data Source: World Economic Outlook-IMF/World Economic Forum     (c.1) (c.2)   Figure 9: Infrastructure Frontiers Single Input-Single Output As in previous cases, countries were clustered into the more efficient and least efficient groups (Table 5). The more efficient countries are the richer economies, with Chile included in that group. Within the inefficient group we find mostly African countries and a few LAC countries like Ecuador and Trinidad and Tobago. The discrepancy in efficiency between both groups is extraordinary (Table 5A) and much larger than in education or health. Table 5. Infrastructure Single Input, Single Output) Input-Efficient Output Efficient More efficient Germany, Belgium, Israel, United Araba Emirates, Switzerland, Singapore, United Germany, Slovenia, Kingdom, Austria, Chile, Switzerland, Netherlands, Finland, France. Great Britain, Iceland. Least efficient Ecuador, Etiopia, Lesotho, Haiti, Tchad, Gabon, Uganda, Venezuela RB, Trinidad and Romania, Vietnam Tobago, Angola, Botswana Table 5A. Infrastructure Single Input, Single Output Efficiency Scores Input-Efficient Output Efficient More efficient .68 .96 Least efficient .10 .32 Input-efficiency is extremely low, with the average regional input-efficiency score between 18% and 30% (Table 6A); AFR is the lower bound at 18% and EAP at 29%. Output-efficiency is significantly higher (Table 5B), indicating that capital spending 24 deviation from the frontier is larger than the output deviation from the frontier. This result could be explained due to the index of quality being used as output indicator (which is limited by nature of the construction from 1-7). Also, the computation of the frontiers including both developed and developing countries could be biasing the results, though when other inputs are included the problem is mitigated significantly. The efficiency scores are closely correlated with indicators of the regulatory quality, or the effectiveness of government (Table 7). This is especially true when output- efficiency indicators are used.20 These relationships are explored econometrically in the last section, when other factors are controlled for. Table 6A Infrastructure: Input-Efficiency scores by regions across the world - Single Input, Single Output AFR  EAP  ECA  LAC  MNA  SAS  Quality of electricity supply  0.17  0.24  0.19  0.18  0.20  0.19  Transport infrastructure  0.17  0.33  0.18  0.19  0.25  0.20  Quality of air transport infrastructure  0.18  0.31  0.17  0.20  0.22  0.19  Quality of port infrastructure  0.19  0.32  0.19  0.22  0.24  0.20  Quality of railroad infrastructure  0.20  0.25  0.23  0.19  0.24  0.28  Quality of roads  0.18  0.30  0.18  0.20  0.24  0.19  Quality of overall infrastructure  0.17  0.26  0.18  0.19  0.22  0.19  Table 6B Infrastructure: Output-Efficiency scores by regions across the world - Single Input, Single Output AFR  EAP  ECA  LAC  MNA  SAS  Quality of electricity supply  0.44  0.68  0.70  0.63  0.75  0.49  Transport infrastructure  0.48  0.67  0.53  0.54  0.63  0.54  Quality of air transport infrastructure  0.57  0.71  0.62  0.66  0.71  0.60  Quality of port infrastructure  0.56  0.65  0.55  0.60  0.67  0.53  Quality of railroad infrastructure  0.35  0.52  0.51  0.30  0.48  0.47  Quality of roads  0.52  0.66  0.51  0.57  0.70  0.56  Quality of overall infrastructure  0.52  0.65  0.59  0.57  0.69  0.54  20 The overall infrastructure index is weakly correlated with ranking on the quality of the public investment management systems (Dabla-Norris, et. al.), probably due to the limited sample in the PIMS study. 25 Table 7 Correlation of Efficiency Rankings and Governance indicators. Spearman Rank Correlation    Control of  Government  Political Stability   Regulatory  Rule of  Voice and  Corruption  Effectiveness  and Absence of  Quality  Law  Accountability  Violence/Terrorism  Input  0.358***  0.374***  0.291***  0.439***  0.381***  0.378***  Efficiency  (113)  (113)  (113)  (113)  (113)  (113)  Output  0.678***  0.752***  0.488***  0.720***  0.686***  0.420***  Efficiency  (113)  (113)  (113)  (113)  (113)  (113)  III.3. Multiple Inputs and Multiple Outputs Education, health, and infrastructure attainment are not solely determined by public spending. Other inputs, such as private spending also affect the output indicators. For both health and capital spending, we could get data on private spending, but not for education.21 Hence the education production technology will have multiple indicators of educational attainment, and three possible inputs (public spending, teachers per pupil, and adult literacy rate). In health, besides public spending, three other inputs were included: private spending, improved sanitation condition, gross secondary school enrollment and the literacy of adults. The analysis was limited to include up to two inputs and two outputs to avoid the curse of dimensionality: increasing the number of output or input indicators biases efficiency scores towards one, increases the variance of the estimators, and reduces the speed of convergence to the true efficiency estimators (Simar and Wilson, 2000; Groskopff, 1996). III.3.1. DEA Education Frontiers The multi-input multi-output efficiency scores produce a ranking of countries like the single input-single output case (Table 8). Lebanon and Kazakhstan appear as efficient countries, close to the origin due to their low spending levels but low attainment levels. However, new countries appear within the more efficient group, such as Finland, the Islamic Republic of Iran and Spain. Within the least efficient countries, Norway appears as input-inefficient, as well as Brazil and Saudi Arabia. The least output efficient countries are mostly African counties, with Pakistan and Peru. The discrepancy in efficiency between the more efficient group and least efficient is lower than in the single input case, but still the typical country of the least efficient group could achieve the same output with 40% lower spending. 21 The source of health data is the WDI while for capital spending it is WEO. 26 Table 8. Educational Attainment: Multiple Inputs, Multiple Outputs sample countries Input-Efficient Output Efficient More efficient Lebanon, Panama, Bahrain, Kazakhstan, Spain, Islamic Kazakhstan, Finland, China, Republic of Iran, Switzerland, Spain, Islamic Republic of Iran Bahrain, Lebanon, Russian Federation Least efficient Cyprus, Costa Rica, Norway, Mali, Senegal, Dominican Iceland, Moldova, Brazil, Republic, Côte d’Ivoire, Saudi Arabia Angola, Mozambique, Sierra Leone, Pakistan, Peru. Table 8A. Educational Attainment: Multiple Inputs, Multiple Outputs efficiency scores Input-Efficient Output Efficient More efficient .99 .99 Least efficient .61 .56 When multiple inputs are considered, efficiency scores tend to rise. The regional aggregation for input and output efficiency scores shows (Tables 9A and 9B) that, as the model becomes more complex (adding inputs or outputs), scores rise due to the increasing uniqueness of the input-output bundle, and the relatively higher difficulty of finding peers. The first four rows of Table 8A allow gauging the impact on efficiency of adding literacy of adults as an additional input, by comparing scores with the single input case (Table 2A). In Africa and SAS, the change is dramatic, increasing from around 60 percent to the low 80s in AFR and low 90s in SAS. In EAP, ECA and MNA changes are of relatively lower importance, and in LAC the addition of the input is insignificant. This differential response to the inclusion of the additional input is due to the low levels of adult literacy in AFR and SAS, which yield gains in efficiency. In output efficiency there is no change in EAP and ECA, but in MENA and LAC the change is relevant with MNA being on the efficiency frontier (Table 8B, row 1 and Table 2B row 8). Individual country efficiency scores for the multiple inputs for education are reported in Table A.3 of Appendix A. Table 9A. Education Attainment: Input-Efficiency scores by regions across the world - Multiple Inputs, Multiple Outputs AFR EAP ECA LAC MNA SAS 2 inputs (public expenditure, literacy of adult) – 1 outputs (Net primary enroll.) 0.84 0.83 0.78 0.76 0.82 0.91 2 inputs (public expenditure, literacy of adult) – 1 outputs (Net Secondary enroll.) 0.84 0.81 0.83 0.82 0.88 0.91 2 inputs (public expenditure, literacy of adult) – 1 outputs (Average Year of 0.83 0.82 0.92 0.78 0.81 0.90 Schooling) 2 inputs (public expenditure, literacy of 0.96 0.91 0.91 1.00 adult) – 1 outputs (Pisa Science) 2 inputs (public expenditure, teachers per pupil) – 2 outputs (Pisa Science & net 0.92 0.86 0.89 0.88 primary enroll.) 27 Table 9B. Education Attainment: Output-Efficiency scores by regions across the world - Multiple Inputs, Multiple Outputs AFR EAP ECA LAC MNA SAS 2 inputs (public expenditure, literacy of 0.91 0.87 0.80 1.00 adult) – 1 outputs (Pisa Science) 2 inputs (public expenditure, literacy of 0.85 0.96 0.95 0.93 0.98 0.91 adult) – 1 outputs (Net primary enroll.) 2 inputs (public expenditure, literacy of 0.57 0.71 0.81 0.74 0.81 0.76 adult) – 1 outputs (Net Secondary enroll.) 2 inputs (public expenditure, literacy of adult) – 1 outputs (Average Year of 0.69 0.73 0.92 0.74 0.79 0.82 Schooling) 2 inputs (public expenditure, teachers per pupil) – 2 outputs (Pisa Science & net 0.98 0.95 0.96 0.98 primary enroll.) III.3.2. DEA Health Multi Input-Output In health there are multiple possible combinations of inputs (public expenditure, private expenditure, and literacy of adults) and outputs (life expectancy at birth, immunization DPT, immunization measles, and Disability Adjusted life expectancy (DALE)). Of the additional inputs, private sector health spending deserves special attention given its magnitude and pattern of use across countries. In many countries, private spending in health is larger than public spending, with the ratio of private to public spending being larger than one and decreasing with the level of GDP per capita (Fig 10). Hence, richer countries use this input less intensively than poorer countries, or, that poorer countries use more public financing relative to private to achieve health outcomes. There is a clear positive relationship between private spending in health and GDP per capita (Fig. 11), like the one displayed by public spending. Hence, we will use the orthogonalized component in the inputs. 28 Ratio of Private to Public Health Spending as a Share of GDP vs GDP Per Capita 5 shareprivatepublicgdp /Linear prediction HTI SLE GEO 4 AZE KHM 3 GNB YEM UGA CMR SDN TJK CIV IND NGA BGD COM PAK PHL 2 MLI ZWE MAR RWA GTM ZAR VEN IRN IDN EGY TGO NER PRY KNA GIN NPL LBN ERI LAO ARM BRA TCDTZA MRT ECU BHS CYP CAF HND MDA UZB LKAGRD ALBZAF MUS CHL USA VNM LCA 1 MOZ BENSEN MEX MDG KEN ZMB UKR RUS TTO BDI MWIETHBFA MNG JAM TUNCHN PER BGR MYS KAZ KOR KGZ NIC NAM GMB DJI DOM BWA ARG LVA AGO GUYSLV BLZ TKM MKD BRBURY HUN CHE GHA FJI BOL SWZ JORDMA DZAGAB VCT ATG BLR PAN POL GRC SVK LTUSVN BHR AUS CANSAU LSO CPV COG CRI COL THA ESP ITA AUT ARE FIN DEU PNG TON TUR ROM EST FRA ISL VUT HRV CZENZL GBR SWE OMN NLD NOR DNK LUX KWT SLB WSM 0 6 8 10 12 Log GDP Per Capita PPP 2011 Data Source: World Bank Indicators Figure 10 Ratio of Private to Public Health Spending and GDP per Capita Private Health Spending vs GDP Per Capita USA 8 CHE CYP CAN AUS AUT DEU LUX KOR BHS TTO FRA FIN NOR lhea/Linear prediction CHLGRC RUS IRN KNA SVN ESP ITA SWE DNK AZE BRA ISL SVK NZL NLD ARE LBN HUN GBRBHRSAU GEO ZAF VEN BGRURY MEX LTU PRY LCA MUSLVA ARG POL 6 GRD ECU PAN BRB KAZ MYS CRI CZE GTM TUN BWA ATG EGY NAM UKR ALB HRV EST KWT MDA MAR MKD BLR CHN SDN JORPERDZA ROM TUR HND JAMMNG SWZ DMA SLV COL DOM OMN YEM IND VNMPHLARM LKA GAB BLZIDN VCT SLE UZB NIC NGAGUY THA KHM CIV HTI BOL FJI UGA TJKCMR DJI RWA MLI NPL KGZZMB PAK CPV AGO TKM GNB ZWE LSO TZA COM KEN MRT LAO 4 BGD GHA COG MWI BENSEN TON TGO BFA WSM GIN GMB TCD NER MOZETH BDI ERI MDG PNG ZAR CAF VUT SLB 2 6 8 10 12 Log GDP Per Capita PPP 2011 Data Source: World Bank Indicators Figure 11 Private Health Spending and GDP per Capita 29 Table 10. Health Attainment Country Clusters: Multiple Inputs-Multiple Outputs Sample of countries Input-Efficient Output Efficient More efficient Oman, Bahrain, Cyprus, Oman, Luxembourg, Cyprus, Luxembourg, Saudi Arabia, Iceland, Norway, Saudi Lebanon Arabia, Bangladesh, Bahrain Least efficient South Africa, Brazil, Georgia, South Africa, Lesotho, Paraguay, Namibia, St. Lucia Zimbabwe, Angola, Sierra Leone Table 10A. Health Attainment Country Clusters: Multiple Inputs-Multiple Outputs Efficiency scores Input-Efficient Output Efficient More efficient 1.0 1.0 Least efficient .64 .84 As in previous cases, the countries were clustered into the most efficient and least efficient groups. In this more complex model, the more efficient group is directly on the frontier (score of 1) while the least efficient group on average scores .64 and .84 in input and output efficiency, respectively. The discrepancy between the groups narrows, but still the scope for expenditure reduction is about 30%. This is lower than in the education case. 30 Table 11A. Health Attainment: Input-Efficiency scores by regions across the world - Multiple Inputs-Multiple Outputs   AFR  EAP  ECA  LAC  MNA  SAS  2 inputs (public expenditure, Private expenditure) – 1 output  0.41  0.52  0.46  0.40  0.59  0.44  (Immunization DPT)  2 inputs (public expenditure, Private expenditure) – 1 output  0.41  0.43  0.47  0.39  0.53  0.42  (Maternal Survival Rate)  2 inputs (public expenditure, Private expenditure) – 1 output  0.45  0.51  0.48  0.49  0.59  0.46  (Dale)  2 inputs (public expenditure, Improved Sanitation) – 1 output  0.85  0.83  0.83  0.82  0.90  0.84  (Maternal Survival Rate)  2 inputs (public expenditure, Improved Sanitation) – 1 output  0.86  0.79  0.72  0.82  0.88  0.80  (Dale)  2 inputs (public expenditure, Adult Literacy Rate) – 1 output  0.80  0.76  0.65  0.64  0.83  0.86  (Immunization DPT)  2 inputs (public expenditure, Adult Literacy Rate) – 1 output  0.78  0.79  0.86  0.76  0.94  0.89  (Maternal Survival Rate)  2 inputs (public expenditure, Adult Literacy Rate) – 1 output  0.81  0.75  0.70  0.82  0.92  0.87  (Dale)  2 inputs (public expenditure, Improved Sanitation) – 2 outputs  0.88  0.86  0.89  0.85  0.86  0.84  (Dale, Maternal Survival Rate)  2 inputs (public expenditure, Improved Sanitation) – 2 outputs  0.87  0.82  0.84  0.84  0.84  0.83  (Dale, Infant Survival Rate)  2 inputs (public expenditure, Adult Literacy Rate) – 2 outputs  0.82  0.81  0.86  0.85  0.97  0.93  (Dale, Maternal Survival Rate)  2 inputs (public expenditure, private expenditure) – 2 outputs  0.94  0.98  0.99  0.98  0.98  0.96  (Dale, Infant Survival Rate)  Table 11B. Health Attainment: Output-Efficiency scores by regions across the world - Multiple Inputs, Multiple Outputs AFR EAP ECA LAC MNA SAS 2 inputs (public expenditure, Private expenditure) – 1 outputs 0.81 0.87 0.95 .093 0.94 0.88 (Immunization, DPT) 2 inputs (public expenditure, Improved Sanitation) – 1 outputs 0.86 0.93 0.91 0.93 0.94 0.93 (Life Expectancy at Birth) 2 inputs (public expenditure, Improved Sanitation) – 1 outputs 0.86 0.88 0.95 0.92 0.97 0.90 (Immunization, DPT) 2 inputs (public expenditure, Improved Sanitation) – 1 outputs 0.88 0.91 0.91 0.94 0.95 0.93 (Dale) 2 inputs (public expenditure, Adult Literacy Rate) – 1 outputs 0.85 0.88 0.93 0.92 0.97 0.90 (Immunization, DPT) 2 inputs (public expenditure, Adult Literacy Rate) – 1 outputs 1.0 1.0 1.0 1.0 1.0 1.0 (Maternal Survival Rate) 2 inputs (public expenditure, Adult Literacy Rate) – 1 outputs 0.84 0.90 0.91 0.95 0.98 0.94 (Dale) 2 inputs (public expenditure, Improved Sanitation) – 2 outputs 1.0 1.0 1.0 1.0 1.0 1.0 (Dale, Maternal Survival Rate) 2 inputs (public expenditure, Adult Literacy Rate) – 2 outputs 0.97 0.99 0.99 0.99 1.0 0.99 (Dale, Infant Survival Rate) 2 inputs (public expenditure, private expenditure) – 2 outputs 0.94 0.98 0.99 0.98 0.98 0.96 (Dale, Infant Survival Rate) 31 The impact of private spending on efficiency scores is puzzling. The input efficiency scores remain unchanged (Table 11A, rows 1-3) compared to the single output case (Table 4A), though output efficiency increases (Tables 11B vs. 4B). But all the other combinations of inputs increase efficiency. For instance, when the adult literacy rate is included as an input, the AFR DALE efficiency increases from .49 to .81 and MNA efficiency increases from .65 to .92. Output efficiency increase as well, though not in the same proportion; from .73 to .84 in AFR and from .92 to .98 in MNA. Other additional inputs, such as improved sanitation conditions, have similar effects on efficiency measures. Similar results are obtained when the two input-two output models are considered, but efficiency scores increase when private health spending is considered. Table 11B shows that, on average, developing nations score between .84 and .89 in output efficiency in the multiple input-output framework. These figures imply that developing countries could raise their output levels by an average of 15 percent with the same input consumption, if they were as efficient as the comparable benchmark countries. This figure is simply indicative, as the estimate varies with the country and with the selected indicator and has a large variance across countries: for instance, the bottom decile of (input) efficiency scores is about .64, implying that the scope for increasing health and education attainment levels is between 3 or 4 times higher than for the whole sample average. Individual country efficiency scores for the multiple inputs for health are reported in Table A.6 of Appendix A. III.3.3. DEA Infrastructure Multi Input-Output In infrastructure we will focus on adding private capital spending as an input.to the existing production function of one input (quality of overall infrastructure). The private spending is generally larger than public capital spending, with the ratio being larger than one (Figure 12). However, there are cases where private spending is extremely low compared with public capital spending, such as Trinidad and Tobago, Venezuela, and Mozambique. The trend is for the ratio to increase with the level of GDP. Private capital spending is also positively correlated with GDP (Fig. 13). Hence, we will use the orthogonalized component. Countries were clustered into the most and least efficient (Table 12) and the efficiency scores of both groups show a marked increase in efficiency (Table 12A). The most efficient group lies on the frontier, while the least efficient group averages .28. 32 Ratio of Private to Public Investment Spending vs GDP per Capita 15 LBN privpub10 /Linear prediction ISR 10 IDN BEL CHL BRA DEU CHE ZMB ARM NPL HND GRC AUT IRL PHL MAR IRN SUR ITA ZAF LVA EGY SLV UKR DOM CRI SRB SGP GTM KOR 5 CZE SVK MDA PRT FIN GBRCAN FRA TZA LKA PAN RUS ISLAUS MDG ZWE TUR HRV EST BHR NLD USA BGD PER BGR KAZ ARGSYC LTU LUX NIC NGA BTN ALB JOR ROM MEX URY JPN ZAR UGA PAK IND CPV THA MNEMUS POL BEN SEN TUN GAB KENCMR PRY BLZ NAMCOL KWT NZL TCD VNM GEO BIH BWA BRB SLE MWIGMBMLI KHM GHA HTI BFA YEM MMRCIV MYS ARE GIN SWZ DZA ECU SAU BDI MOZ RWA ETH LSO BOL GUY OMN TJK AZE VEN AGO TTO 0 6 8 10 12 lgdp Data Source: World Economic Outlook-IMF Figure 12 Ratio of Private to Public Capital Spending Private Investment Spending vs GDP per Capita 10 CHE LUX ent Spending Per Capita AUS IRL AUT SGP CAN BEL FIN NLD JPN FRA DEUUSA ISL ITA NZLGBRARE KOR ISR CZE BHR GRCKWT EST PRT 8 SUR LVACHLSVKSAU GAB HRV SYC OMN PAN TUR POL LTU LBN BRA MEX KAZ IRNROM URY BRB RUS CRI BWA MUS BGR ZAF ARG MYS DOM ALB THA MNE VEN CPV BTN IDN COL NAM SRB PER TTO MAR DZA JOR ARMTUN LKA ECU BIH BLZ GEO AZE UKRSLV 6 HND GTM SWZ EGY PHL ZMB IND NIC NGA PRY MDA VNM GUY TCD SENCMR GHA Private Investm BGD TZA LSO KEN BOL HTI UGABEN PAK NPL ZWE AGO KHM CIV MLI MMRYEM MDGGMB 4 BFA TJK RWA GIN ETH SLE MWI ZAR MOZ BDI 2 4 6 8 10 12 lgdp Data Source: World Economic Outlook-IMF Figure 13 Private Capital Spending per capita and GDP per capita The regional aggregation shows that MNA and EAP have the highest efficiency scores, both input and output oriented (Table 13). Individual country efficiency scores for the multiple inputs for infrastructure are reported in Table A.8 of Appendix A. 33 Table 12. Capital Spending Multiple Inputs, Multiple Outputs Input-Efficient Output Efficient More efficient Germany, Israel, Singapore, Germany, Israel, United Arab Belgium, Switzerland, France, Emirates, Barbados, Finland, United Arab Emirates Switzerland, Netherlands, Singapore Least efficient Haiti, Bosnia-Herzegovina, Haiti, Chad, Myanmar, Myanmar, Paraguay, Angola, Bosnia-Herzegovina, Burundi, Sierra Leone, Paraguay, Bangladesh, Venezuela RB Nigeria, Moldova, Lesotho Table 12A. Capital Spending Multiple Inputs, Multiple Outputs Input-Efficient Output Efficient More efficient .99 1.0 Least efficient .28 0.33 Table 13A. Infrastructure: Input-Efficiency scores by regions across the world - Multiple Inputs-Multiple Outputs AFR  EAP  ECA  LAC  MNA  SAS  Quality of electricity supply  0.52  0.57  0.61  0.61  0.69  0.47  Transport infrastructure  0.48  0.59  0.48  0.52  0.59  0.47  Quality of air transport  0.49  0.57  0.49  0.56  0.60  0.46  infrastructure  Quality of port infrastructure  0.47  0.54  0.43  0.51  0.58  0.43  Quality of railroad infrastructure  0.46  0.55  0.56  0.46  0.50  0.54  Quality of roads  0.54  0.58  0.47  0.55  0.67  0.49  Quality of overall infrastructure  0.52  0.56  0.53  0.55  0.66  0.47  34 Table 13B. Infrastructure: Output-Efficiency scores by regions across the world - Multiple Inputs-Multiple Outputs AFR  EAP  ECA  LAC  MNA  SAS  Quality of electricity supply  0.50  0.68  0.71  0.66  0.78  0.49  Transport infrastructure  0.54  0.67  0.55  0.58  0.66  0.55  Quality of air transport  0.63  0.71  0.64  0.69  0.75  0.61  infrastructure  Quality of port infrastructure  0.62  0.66  0.56  0.63  0.71  0.54  Quality of railroad infrastructure  0.42  0.54  0.55  0.36  0.49  0.47  Quality of roads  0.59  0.66  0.52  0.60  0.74  0.57  Quality of overall infrastructure  0.58  0.65  0.61  0.61  0.74  0.55  IV. Covariates of Inefficiency Variation across Countries This chapter seeks to identify correlates of efficiency variation across countries. Efficiency scores were obtained in the first stage described up to this section. This two- stage approach seeks to identify statistically significant regularities common to efficient or inefficient countries. This exercise does not try to identify supply or demand factors that affect health and education outcomes, such as those described by Filmer (2003). The scope is limited to verifying statistical association between the efficiency scores and environmental variables, as done in Herrera and Pang (2005). IV.1. Method, Variables and Data Description The cross-section consists of many countries (varying from 19 to 69 depending on the output indicator) and a single period, as we collapsed the 2009-2015 information into a period average for each variable. The dependent variable is the input efficiency score calculated by the DEA method in the first stage. Given that the dependent variable (the efficiency score) is continuous and distributed over a limited interval (between zero and one), it is appropriate to use a censored (Tobit) regression model to analyze the relationships with other variables. The input-oriented estimator reflects the consideration that input choices are more under the policy maker’s control. The independent variables reflect environmental effects included in precursor papers, as well as suggested by others recently. We included the following independent variables.22 a. The size of government expenditure. Most of the papers surveyed in the previous section explore the relationship between the size of the government (or expenditure as a percentage of GDP) and efficiency levels. The objective is to verify if additional pubic spending is associated with better education and health outcomes. While some papers have found a negative association between efficiency and expenditure levels (Gupta- Verhoeven 2001, Jarasuriya-Woodon 2003, and Afonso et.al. 2003 Herrera-Pang, 2005), 22 The precise definition and sources can be found in Appendix A, Table A.10. 35 others have found a positive association (Evans et.al. 2003) and others have found no significant impact (Filmer and Pritchett, 1999). b. A government budget composition variable. Given that both education and health are labor-intensive activities, the government’s labor policies will determine the efficiency with which outputs are delivered. We chose a budget composition indicator to reflect this, in particular, the ratio of the wage bill to the total budget. A higher ratio is expected to be negatively correlated with efficiency. c. Per-capita GDP. We included the per-capita GDP to control for the Balassa- Samuleson effect in comparing across countries. If richer countries tend to be more inefficient (given higher wages in these countries), a negative sign is expected. However, it must be recalled that to obtain the efficiency scores in the “first stage” we constructed an auxiliary variable (the orthogonalized public expenditure). Hence the inclusion of this variable in the second stage is an attempt to control for any remaining Balassa- Samuleson effects. d. Urbanization. The clustering of agents makes it cheaper to provide services in urbanized areas rather than in rural. Higher degree of urbanization should reflect in higher efficiency, making positive as the expected sign of the coefficient on this variable. e. Prevalence of HIV/AIDS. Based on WHO mappings of the disease, we included a dummy variable in the most severely affected countries to control for the role of this epidemic in the poor health outcomes. Evans et al. (2000) report that AIDS lowers the disability adjusted life expectancy (DALE) by 15 years or more. AIDS also affects education outcomes both directly and indirectly (Drake, et al. 2003): directly because school-age children are affected: UNAIDS estimates that almost 4 million children have been infected since the epidemic began, and two-thirds have died. However, the indirect channel is relatively more important: AIDS leaves orphaned children who are more likely to drop-out of school or repeat. All these factors reflect how AIDS affects the demand for education. But the supply is also affected by the decreasing teacher labor force due to illness or death, or the need to care for family (Pigozzi, 2004). Prevalence of HIV/AIDS should be negatively associated with education and health outcomes. Consequently, efficiency scores should be negatively associated with the dummy variable. f. Income distribution inequality. Ravallion (2003) argues that, besides the mean income, its distribution affects social indicators because their attainment is mostly determined by the income of the poor. Hence, we controlled for the distribution of income by including the Gini coefficient as an explanatory variable. Higher inequality is expected to be associated with lower educational and health attainments, making negative the expected sign of this variable. g. Share of public sector in the provision of service. Services can be provided by both the public and private sectors, and efficiency indicators will differ across countries depending on the relative productivities of both sectors. Previous studies have included this variable to explain differences in outcomes (Le Grand, 1987; Berger and Messer, 36 2002) or efficiency scores (Greene, 2003a). The specific variable we included was the ratio of publicly financed service over the total spending (sum of private and public spending). h. External aid. To the extent that countries do not have to incur the burden of taxation, they may not have the incentive to use resources in the most cost-effective way. Another channel through which aid-financing may affect efficiency is through the volatility and unpredictability of its flows. Given that this financing source is more volatile than other types of fiscal revenue (Bulir and Hamann, 2000), it is difficult to undertake medium- term planning within activities funded with aid resources. If this is the case, we would expect a negative association between aid-dependence and efficiency in those activities funded with aid, mostly health services. To our knowledge there are no previous attempts to establish a relationship between efficiency and the degree to which activities are financed by external aid. There is, however, recent evidence of a negative association between donor financing and some health outcomes (Bokhari, Gottret, and Gai, 2005). i. Institutional Variables. Countries with better institutions, more transparency, and less corruption are expected to have higher efficiency scores. Similarly, countries that have suffered wars or state failures are expected to register lower efficiency scores. To capture these effects, we included the World Bank Governance Indicators. The data on educational and health indicators are not available on a continuous annual basis for many countries. Thus, averages of the variables were computed over the 2009-2016 period for all variables. IV.2. Results The Tobit estimation on panel data is defined as follows. VRSTEit  f (WAGEit , GOVEXPit , PUBTOTit , GDPPCit , URBAN it , AIDSit , GINI it , EXTAIDit , INSTit , CONS ) where VRSTE it = Variable returns to scale DEA input or output efficiency scores. WAGEit = Wages and salaries (% of total public expenditure) GOVEXPit = Total government consumption expenditure (% of GDP) PUBTOTit = Share of expenditures publicly financed (public/total) GDPPCit = GDP per capita in constant 1995 US dollars URBAN it = Urban population (% of total) AIDS it = Dummy variable for HIV/AIDS GINI it = Gini Coefficient EXTAIDit = External aid (% of fiscal revenue) INSTit = Institutional indicators WB Governance Indicators. CONS = Constant 37 Tables 14, 15, and 16 show the results for the single-input single-output education, health and infrastructure scores, with input efficiency scores (a) as dependent variable and output efficiency scores (b), respectively. The findings show: a. Government expenditure (GOVEXP) is negatively associated with efficiency scores in education (Tables 14 a and b). This result is robust to changes in the output indicator selected. In the output efficiency case, the impact is ambiguous specially when the PISA Math and Science scores are the output indicators (Table 14 b). In health (Tables 15 a and b), the negative association is present in both input and output efficiency. In infrastructure, the expenditure variables (GOVEXP and PUBGFC10PC) are negative in the six output indicators that are used (Table 16a).23 There is a robust trade-off between size of expenditure and efficiency. b. The wage bill within total expenditure (WAGE) and efficiency is ambiguous in education; it is positive and significant when the enrollment indicators (Table 14 a) are considered, but negative when the PISA scores are considered (Table 14b). WAGE is insignificant in health and positive in infrastructure. c. The share of public financing within the total (sum of public and private) is robustly associated with lower efficiency scores. In the case of education (EDUCPUBTOT) it is negatively associated when input efficiency is considered (Table 14a) and a similar result is obtained in health (HEAPUBTOT, Table 15a). In health, the share of public financing is positively associated with output efficiency, rendering this impact as ambiguous in health. Further research would be needed to explain why this is the case in health services. In infrastructure it is negative (Table 16A). d. GDP per capita is positively associated with efficiency scores, indicating that efficiency scores are higher in richer countries despite the orthogonalization procedure (Tables 14a, 15a, 16a). e. Urbanization is negatively associated with efficiency scores in education (Tables 13a and b), while it is positively associated in health. However, scores are unrelated to this variable. f. The effect of HIV/AIDS is negative on both education and health (Tables 13b and 14b). g. Income distribution (GINI) is negatively associated with efficiency in education (table 14a) and health (Table 15b). However, there are cases of sign reversals that 23 The govexp variable is total consumption expenditure while PUBGFC10PC is capital spending per capita. Hence the infrastructure equations include both types of spending. 38 are puzzling, with the primary enrollment output efficiency scores being positively associated with GINI (Table 14b); in Health, GINI is positively associated with input efficiency in Measles Immunization and Maternal Survival (Table 15a) but negatively related with the output efficiency measures (Table 15b). h. The external aid dependency ratio (EXTAID) is negatively associated with efficiency in most education indicators, except with PISA output efficiency (Table 14b) where it is positive. In the case of health there is a negative association (Table 15b). Though no causality relationship can be inferred from the exercise, this is one of the results that merit more detailed research. This result might be explained by the volatility of aid as a funding source that limits medium term planning and effective budgeting. Interpretation of these results requires caution due to several limitations. First, education and health outcomes are explained by multiple supply and demand factors (Filmer, 2003) which are not included here. The omission of one of these factors in the health or education production functions in the previous stage could explain some of the cross- country co-variation of the efficiency results (Ravallion, 2003). The goodness-of-fit analysis of the first stage indicated that no important factor seemed to be omitted. Of course, there can always be additional factors that could be included but the curse of dimensionality24 is particularly pressing in non-parametric statistical methods (even if the data were available). The second limitation derives from the intuitive question why the set of explanatory variables used in the second stage were not included in the first stage. The answer lies in that most of these variables are environmental and outside the control of the decision- making unit. The inclusion of these environmental variables would have had little justification from the production function perspective. Additionally, by maintaining the production function as simple as possible the dimensionality curse is avoided. Finally, the third limitation arises from the fact that if the variables used in the first stage to obtain the efficiency estimator are correlated with the second stage explanatory variables, the coefficients will be inconsistent and biased (Simar and Wilson, 2004; Grosskopf, 1996; Ravallion, 2003). To examine the extent of this potential problem, we calculated correlation coefficients between the “first-stage” inputs and the second-stage explanatory variables. The largest correlation coefficients were between GDP per capita and the teachers per pupil ratio and the literacy rate of adults. To examine the sensitivity of the results to the inclusion of GDP per capita, all the estimations were performed without this variable and none of the results changed. 24 As the number of outputs increases, the number of observations must increase exponentially to maintain a given mean-square error of the estimator. See Simar and Wilson (2000). 39 V. Concluding Remarks and Directions for Future Work The paper presented an application of non-parametric methods to analyze the efficiency of public spending. Based on a sample of more than 140 countries, the paper estimated efficiency scores for 14 education output indicators, seven health output indicators, and seven infrastructure attainment indicators. Our results show that countries could achieve substantially higher education, health, and infrastructure quality levels: on average, developing countries reached output efficiency of about .6 in infrastructure, .8 in education, and .9 in health. That means that, for the same spending levels, peer countries achieve significantly higher outputs, with infrastructure being the area with more room for improvement. On the input side, efficiency is lower. Input efficiency scored .53, .75, and .86, for infrastructure, health, and education, respectively. These results indicate that developing countries deviate more from the frontier on the input (spending) side than on the output dimension. This is just an indicative figure, as the figures vary across countries and with the selected output indicator. It is crucial to identify what are the institutional or economic factors that cause some countries to be more efficient than others in service delivery. In terms of policy implications, it is vital to differentiate between the technically efficient level and the optimal or desired spending level. Even if a country is identified as an “efficient” benchmark country, it may very well still need to expand its public spending levels to achieve a target level of education or health attainment indicators. Such is the case of countries with low spending levels and low attainment indicators, close to the origin of the efficient frontier. The important thing is that countries expand their scale of operation along the efficient frontier. The methods used in the paper can be interpreted as tools to identify extreme cases of efficient countries and inefficient ones. Once the cases have been identified, more in- depth analysis is required to explain departures from the benchmark, as proposed and done by Sen (1981). Given that the methods are based on estimating the frontier directly from observed input-output combinations, they are subject to sampling variability and are sensitive to the presence of outliers. In a “second stage” the paper verified statistical association between the efficiency scores and environmental variables that are not under the control of the decision-making units. The Tobit regressions showed that the variables, which are negatively associated with efficiency scores, include the size of public expenditure, the proportion of the service that is publicly financed, the prevalence of the HIV/AIDS epidemic on health efficiency scores, income inequality on education efficiency scores, and external aid-financing on some of the efficiency scores. 40 Table 14A. Correlates of Cross-Country variation in Education Input efficiency Net Gross Gross Net Average Secondary Math Science Literacy of First Level VARIABLES Primary Primary Secondary Secondary Years of Level PISA PISA Youth Complete Enrollment Enrollment Enrollment Enrollment School Complete Scores Scores wage 0.30** 0.35** 0.21** 0.16** 0.12 0.344* 0.12 0.10 -0.037 0.14 govexp -1.502*** -1.534*** -0.619*** -0.733*** -1.172*** 1.043* -1.383*** -0.934*** -1.317 -1.591* educpubtot -0.434 -0.922** -0.968*** -0.630*** -0.24 -1.871*** -0.416 -0.238 -0.717 -0.692 gdppc 8.1e-06* 2.6e-05*** 1.6e-05*** 1.2e-05*** 9.5e-06** -7.8E-06 1.93e-05*** 1.27e-05*** 9.7E-06 8.7E-06 urban -0.002 -0.0035** -0.0001 -0.0012 -0.0025** 0.0056** -0.0041*** -0.0024*** 0.00013 0.0022 aids 0.00491 0.068 0.0238 0.0339 0.0404 0.220*** 0.0554 0.0306 gini -0.0028 7.5E-05 -0.003* -0.003** -0.007*** -0.01*** -0.001 -0.0046*** 0.002 -0.0002 aid 0.000256 0.00428 -0.0203** -0.00687** -0.00738*** -0.00505 -0.00648 -0.00642*** 0.407* 0.362* Constant 0.999*** 0.550** 0.676*** 0.788*** 1.049*** 0.347 0.877*** 0.871*** 0.636 0.482 Observations 56 45 46 57 42 54 53 53 19 19 Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 41 Table 14B. Correlates of Cross-Country variation in Education Output efficiency Net  Gross  Net  Gross  Average  Second  Literacy of First Level  Math PISA  Science  VARIABLES  Primary  Primary  Secondary  Secondary  Years of  Level  Youth  Complete  Scores  PISA Scores  Enrollment  Enrollment  Enrollment  Enrollment  School  Complete  wage  ‐0.121  ‐0.0124  ‐0.133  ‐0.0687  ‐0.219  ‐0.19  ‐0.00286  ‐0.304  ‐0.388***  ‐0.333***  govexp  ‐0.251  ‐0.633**  0.475  0.610*  0.596*  0.509  ‐0.0435  1.091*  0.609**  0.439*  educpubtot  ‐0.184  ‐0.556*  0.0138  ‐0.507  ‐0.216  ‐0.182  ‐0.21  ‐0.534  0.653*  0.331  gdppc  1.91E‐07  3.43E‐06  1.80e‐05***  1.32e‐05***  1.22e‐05**  1.12e‐05**  ‐1.73E‐07  1.19E‐05  1.96e‐05***  1.49e‐05***  urban  ‐0.00017  ‐0.00188*  ‐0.00035  0.000197  0.000425  ‐0.00091  ‐0.00245  0.000623  ‐0.00330**  ‐0.00244*  aids  ‐0.0890***  ‐0.0644*  ‐0.145***  ‐0.173***  ‐0.0867*  ‐0.142***  ‐0.0923  ‐0.041  gini  0.00351***  0.00474***  ‐0.00153  ‐0.00664***  ‐0.00276  0.00415**  0.0066  ‐0.00779**  ‐0.00218  ‐0.00058  aid  0.00700**  0.00438  ‐0.00362  ‐0.0340***  ‐0.0068  ‐0.0047  0.00173  ‐0.0103  0.376***  0.283**  Constant  0.916***  0.856***  0.490***  0.870***  0.624***  0.722***  0.267  0.597***  0.688***  0.739***  Observations  58  61  59  48  54  55  54  54  19  19  Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 42 Table 15A. Correlates of Cross-Country variation in HEALTH INPUT efficiency Disability Life Immunization Immunization Free Infant Maternal VARIABLES Adjusted Life Expectancy DPT Measles Tuberculosis Survival Rate Survival Rate Expectancy wage 0.0455 0.0684 0.0329 -0.0239 0.0406 -0.0794* -0.0226 govexp -0.436*** -0.421** -0.107 -0.231*** -0.0885 -0.190* -0.209*** heapubtot -0.105 0.106 -0.333 -0.296*** -0.160* -0.213** -0.312*** gdppc 6.50e-06*** 5.60e-06* 5.4E-06 6.44e-06*** 4.75e-06*** 1.14e-05*** 6.26e-06*** urban 0.000567 0.000925 -0.00054 -3.9E-05 0.00186*** 0.000492 9.09E-06 aids 0.0114 -0.00872 -0.0424 0.00163 -0.00517 0.000387 0.000924 gini -0.00053 0.000577 0.00129 0.000763* 0.000273 8.77E-05 0.000801** aid -0.00033 -4.2E-05 -0.0012 -0.00088 -0.00149 -0.00162 -0.00078 Constant 0.458*** 0.449*** 0.473*** 0.415*** 0.300*** 0.356*** 0.407*** Observations 67 67 67 67 62 64 67 Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 43 Table 15B. Correlates of Cross-Country variation in HEALTH OUTPUTefficiency Disability  Life  Immunization  Immunization  Free  Infant Survival  Maternal  VARIABLES  Adjusted Life  Expectancy  DPT  Measles  Tuberculosis  Rate  Survival Rate  Expectancy  wage  0.0189  0.0814**  ‐0.0125  ‐0.08  0.00275  ‐0.0134  ‐0.00143  govexp  ‐0.262***  ‐0.402***  0.352  0.369  ‐0.00890**  ‐0.0144  0.000172  heapubtot  0.294***  0.266**  0.0612  0.29  0.00444  0.0918***  0.000682  gdppc  9.48E‐07  1.68E‐06  5.11E‐06  5.56E‐06  ‐9.50E‐08  4.97E‐07  1.03E‐08  urban  0.000585  0.00131***  ‐0.00126  ‐0.00091  3.91e‐05**  6.34E‐05  8.60E‐07  aids  ‐0.111***  ‐0.0948***  ‐0.0788**  ‐0.0689*  ‐0.00128*  ‐0.0209***  ‐0.00261***  gini  ‐0.00158**  ‐0.00226***  0.000412  ‐0.00029  ‐0.000121***  ‐0.00027  ‐4.73E‐06  aid  ‐0.00422***  ‐0.00172  0.00246  0.00109  ‐0.0001  ‐0.00165***  ‐0.000342***  Constant  0.912***  0.892***  0.883***  0.859***  1.003***  0.982***  1.000***  Observations  69  69  69  69  63  69  69  Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 44 Table 16A Correlates of cross-country variation in Capital Spending Input efficiency Quality of air  Quality of overall  Quality of railroad  Quality of port  Quality of  VARIABLES  Quality of roads  transport  infrastructure  infrastructure  infrastructure  electricity supply  infrastructure  wage  18.55*  19.21*  5.137  25.25**  11.31  16.74**  govexp  ‐39.17*  ‐39.59  ‐37.47*  ‐45.27*  ‐26.91  ‐35.98**  pubshare  ‐0.183*  ‐0.163  ‐0.308***  ‐0.168  ‐0.191*  ‐0.235***  pubgfcp10pc  ‐0.000353***  ‐0.000419***  ‐0.000202*  ‐0.000461***  ‐0.000396***  ‐0.000291***  gdppc10  1.73e‐05**  2.86e‐05***  9.76E‐06  2.42e‐05***  1.89e‐05**  1.38e‐05**  urban  0.000526  ‐0.0002  ‐0.00011  0.000439  0.000532  0.000109  gini  0.00218  0.00247  0.00258*  0.00329**  0.00438***  0.00155  aid  0.00155  0.00197  0.00161  0.00158  0.00285  0.000551  Voice and Accountability  ‐0.0191  ‐0.0185  ‐0.0242  ‐0.0228  ‐0.0161  ‐0.00367  Rule of law  0.129**  0.131  0.115*  0.137*  0.149**  0.072  Regulatory Quality  ‐0.0137  ‐0.00223  ‐0.0318  ‐0.00663  ‐0.0251  ‐0.0175  Political Stability and  ‐0.0272  ‐0.0296  ‐0.0115  ‐0.0322  ‐0.0311  ‐0.0227  Absence of Violence  Government  0.00553  0.00565  ‐0.0125  0.0246  0.0378  0.00317  Effectiveness  Control of Corruption  ‐0.0627  ‐0.0605  ‐0.0463  ‐0.0696  ‐0.109**  ‐0.0407  Constant  0.140*  0.142  0.248***  0.0971  0.054  0.190***  Observations  58  58  53  58  58  58  Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 45 Table 16B. Correlates of cross-country variation in Capital Spending Output efficiency Quality of  Quality of air  Quality of overall  Quality of port  Quality of  VARIABLES  Quality of roads  railroad  transport  infrastructure  infrastructure  electricity supply  infrastructure  infrastructure  wage  1.797  20.85*  ‐42.68**  21.80*  10.26  10.96  govexp  4.265  ‐39.83  49.14  ‐28.22  ‐54.46**  53.28*  pubshare  ‐0.291***  ‐0.220*  ‐0.412**  ‐0.191  ‐0.0434  ‐0.155  pubgfcp10pc  5.01E‐05  4.85E‐05  0.000317  ‐5.75E‐05  ‐0.00017  ‐0.0002  gdppc10  ‐1.82E‐06  5.84E‐06  ‐1.81E‐05  2.70E‐06  1.87E‐06  1.23E‐05  urban  0.000218  ‐0.00131  0.000193  0.000241  0.00111  0.00224**  gini  4.27E‐05  0.00165  ‐0.00014  0.00165  0.000841  ‐0.00282  aid  0.00145  0.0037  ‐0.00112  0.000419  ‐0.00536  ‐0.00082  Voice and Accountability  ‐0.100***  ‐0.0914***  ‐0.0708*  ‐0.0331  ‐0.0619**  ‐0.123***  Rule of law  ‐0.0581  ‐0.113  0.0839  0.00519  0.0229  ‐0.126  Regulatory Quality  0.0491  0.019  ‐0.0954*  0.0164  0.113***  0.0905*  Political Stability and  0.0222  0.00201  0.0376  ‐0.0162  ‐0.0152  0.0107  Absence of Violence  Government  0.146**  0.229***  0.141  0.112  0.135*  0.218**  Effectiveness  Control of Corruption  0.056  0.0948  ‐0.0865  ‐0.00461  ‐0.0629  0.085  Constant  0.654***  0.624***  0.577***  0.544***  0.652***  0.546***  Observations  63  63  59  63  63  63  Note: ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 46 References 1. Afonso, A. and M. St Aubyn 2004 Non-parametric approaches to education and health: Expenditure efficiency in OECD countries”. Mimeo. Technical University of Lisbon. 2. Afonso, A., L. Schuknecht and V. Tanzi 2003 Public sector efficiency:an international comparison”. Working paper 242. European Central Bank. 3. Afonso, A., A. Romero, and E. Monsalve 2013. Public Sector Efficiency: Evidence for Latin America. IADB Discussion Paper IDB-DP-279. 4. Afonso, A. and J. Alves 2017. “Reconsidering Wagner’s Law; Evidence from the Functions of Governemnt”, Applied Economics Letters. Vol. 24 No. 5; 346-350. 5. Aigner, D.J., C.A.K. Lovell, and P. Schmidt. 1977. Formulation and estimation of stochastic frontier production function models, Journal of Econometrics, Vol. 6: 21- 37. 6. Alexander, R. and M. Jaforullah, 2004. Explaining efficiency differences in New Zealand secondary schools. Paper presented in the International Conference on Policy Modeling. Paris, June. 7. Banker, R. 1993 “Maximum likelihood, consistency and data envelopmemnt analysis: a statistical foundation” Management Science, Vol. 39, No. 10 8. Berger, M. and J. Messer. 2002. “Public financing of health expenditures, insurance, and health outcomes”, Applied Economics, 34, pp. 2105-2113 9. Birdsall,N., S. Claessens, and I. Diwan 2003. “Policy selectivity forgone: debt and donor behavior in Africa”, The World Bank Economic Review, Vol. 17, No. 3 10. Bishop, P. and S. Brand 2003 “The efficiency of museums: a stochastic frontier production function approach”, Applied Economics, 35, pp.1853-1858. 11. Bokhari, F., P. Gottret and Y. Gai (2005) “Government health expenditures and health outcomes”. Mimeo. The World Bank. 12. Bulir, A. and A. Hamann 2000. “Aid volatility: an empirical assessment”, IMF Staff Papers, vol. 50, No. 1, pp. 64-89. 13. Burgess, J. and P. Wilson 1998 “Variation in inefficiency in U.S. Hospitals” Canadian Journal of Operational Research and Information Processing, 36, pp.84- 102. 14. Bowlin, William F. 1998. Measuring Performance: An Introduction to Data Envelopment Analysis (DEA), Journal of Cost Analysis, pp. 3-27. 15. Cazals, C., Florens, J. and L. Simar 2002 “Non-parametric frontier estimation; A robust approach” Journal of Econometrics, 106(1) pp. 1-25 16. Chang, T. “An econometric test of Wagner’s law for six countries based on cointegration and error-correction modeling techniques”, Applied Econometrics. 34, pp.1157-1169 17. Charnes, A. W. Cooper and E. Rhodes 1978 “measuring efficiency of decision- making units” European Journal of Operational Research, 3 pp.429-444. 18. Cherchye, L. and T. Post 2001. “Methodological advances in DEA: A survey and an application for the Dutch electricity sector”. ERIM Report Series Research in Management, ERS-2001-53-F&A. 19. Crouch, Louis and Tazeen Fasih 2004. Patterns in Educational Development: Implications for Further Efficiency Analysis. mimeo, World Bank. 47 20. Cullinane K.and D.Song 2003 “A stochastic frontier model of the productive efficiency of Korean container terminals”, Applied Economics, 35, pp.251-267 21. Dabla Norris,E., J. Brumby, A. Kyobe, Z. Mills, C. C. Papageorgiu. 2012. Investing in public investment management: an index of public investment effciciency. Journal of Economic Growth. 17 235-266. 22. Drake, L., C. Maier, M. Jukes, A Patrikios, D. Bundy, A. Gardner, C. Dolan (2003) “School-age children: Their nutrition and health 23. Evans, D, A. Tandon, C. JL Murray, and J. A Lauer. 2000. The comparative efficiency of national health systems in producing health: an analysis of 191 countries. World Health Organization GPE Discussion Paper Series No. 29 24. Fare, R. and S. Grosskopf “Non-parametric tests of regularity, Farrell efficiency, and goodness of fit”. Journal of Econometrics, 69. pp.415-425. 25. Farrell, M. 1957 “The measurement of productive efficiency”, Journal of the Royal Statistical Society, Series A. Vol. 120, No. 3, pp. 253-290 26. Filmer, D 2003. “Determinants of health and education outcomes”. Background note for World Development Report 2004. The World Bank. Mimeo. May. 27. Filmer, D. And L.Pritchett, 1999. “The impact of public spending on health: does money matter?” Social Science and Medicine, 49, pp.1309-1323 28. Galiani, S., P. Gertler, and E. Schargrodsky 2005 “Water for life: the impact of the privatization of water services on child mortality”, Journal of Political Economy, vol. 113, no. 1 29. Glaeser, E. and B. Sacerdote. 1999. “Why is there more crime in cities?” Journal of Political Economy, 107, pp 225-258 30. Greene, W. 2003a. Distinguishing between heterogeneity and inefficiency: Stochastic frontier analysis of the World Health Organization’s panel data on national health cares systems”. Mimeo. NYU. www.stern.nyu.edu/~wgreene. 31. Greene, W. 2003b Fixed effects and bias due to the incidental parameters problem in the Tobit model. Mimeo. NYU. www.stern.nyu.edu/~wgreene 32. Grigoli, F. and J. Kapsoli (2013) Waste not, want not: the effciency of health expenditure in emerging and developing economies. IMF Working paper WP/13/187. 33. Grigoli, F. (2014) A hybrid approach to estimating the efficiency of public spending on education in emerging and developing economies. IMF Working Paper WP/14/19. 34. Grosskopf, S. 1996 “Statistical inference and non-parametric efficiency: a selective survey” The Journal of Productivity Analysis, 7 pp.161-176. 35. Gupta, Sanjeev, and Martin Verhoeven. 2001. The efficiency of government expenditure, experiences from Africa. Journal of Policy Modeling, 23, 433-467. 36. Haque, N., H Pesaram and S. Sharma 1999 “Neglected heterogeneity and dynamics in cross-country savings regressions”. IMF working paper. 37. Hanushek, Eric A. 2002. The failure of input-based schooling policies. National Bureau of Economic Research 38. Hanushek, Eric A. and Javier A. Luque. 2002. Efficiency and equity in schools around the world. National Bureau of Economic Research. 39. Herrera, S. and G. Pang (2005). Efficiency of public spending in developing countries: An efficiency frontier approach. Policy Research Working Paper 3645. 40. ICRG International Country Risk Indicators, 2004. http://www.icrgonline.com 48 41. Jayasuriya, Ruwan, and Quentin Wodon. 2002. Measuring and explaining country efficiency in improving health and education indicators. The World Bank 42. Jounard, I. , M. Konsgrud, Y. Nam, R. Price 2004. Enhancing the cost effectiveness of public spending: experience in OECD countries. Economics Department Working paper 380. OECD. February. 43. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton, 1999a. Governance matters, the World Bank Policy Research Working Paper WPS2196 44. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton, 1999b. Aggregating governance indicators, the World Bank Policy Research Working Paper WPS2195 45. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton, 2002. Governance matters II - updated indicators for 2000-01, the World Bank Policy Research Working Paper WPS2772 46. Le Grand, J. 1987. “Inequalities in health: Some international comparisons” European Economic Review, 31, 182-191 47. Liese, B. N. Blanche and G. Dussault. 2003 The human resource crisis in health services in Sub_Saharan Africa. Background paper for the WDR 2004. 48. Maddala, G. S. 1987. “Limited dependent variable models using panel data” The Journal of Human Resources. XXII, 3 49. Mathers, Colin D, Ritu Sadana, Joshua A Salomon Christopher JL Murray, and Alan D Lopez. 2000. Estimates of DALE for 191 countries: methods and results. World Health Organization GPE working paper No. 16. 50. Murillo-Zamorano, L. 2004 “Economic efficiency and frontier techniques” Journal of Economic Surveys, Vol. 18, No. 1, pp.33-77. 51. Murray, Christopher JL, Joshua A Salomon, and Colin Mathers. 1999. A critical examination of summary measures of population health. World Health Organization GPE Discussion Paper No. 2. 52. Nin,A. C. Arndt, and P.Preckel 2003. “Is agricultural productivity in developing countries really shrinking? New evidence using a modified nonparametric approach” Journal of Development Economics, 71, pp.395-415 53. Perelman, S., J. Perelman, and D. Santin 2017. Performance and productivity in education and health: Brazil in the international context. Mimeo. World Bank. Brazil Public Expenditure review. August. 54. Pigozzi,M 2004. . “Quality education and HIV/AIDS”. UNESCO. mimeo 55. Ravallion, M. 2003. “On measuring aggregate social efficiency”. Policy Research Working Paper 3166. The World Bank. Novemeber. 56. Sen, A. 1981. “Public action and the quality of life in developing countries”, Oxford Bulletin of Economics and Statistics, Vol. 43, No. 4 November. 57. Simar, L. and P. Wilson (2000) “Statistical inference in non-parametric frontier models: The state of the art” Journal of Productivity Analysis, 13 pp 49-78. 58. The State Failure Task Force, 2004. Political Instability (aka State Failure) Task Force Data, http://www.cidcm.umd.edu/inscr/stfail/sfdata.htm 59. Tanzi V. 2004. Measuring efficiency in public expenditure. Paper presented in Conference on Public Expenditure Evaluation and Growth. The World Bank. October. 60. Varian, H. 1990 “Goodness-of-fit in optimizing models”, Journal of Econometrics, 46, pp. 125-140. 49 61. Wagstaff, A. and L.C. Wang 2011. A hybrid approach to effciency measurement with empirical illustrations from education and health. World Bank. Policy Research Working Paper 5751. August. 62. Wheelock D. and Wilson, P. 2003. “Robust Non-parametric estimation of efficiency and technical change in U.S. Commercial Banking” Working Paper. Federal Reserve Bank of St. Louis. November. 63. Wilson, P. (2004) “A preliminary non-parametric analysis of public education and health expenditures in developing countries”. Mimeo. The World Bank. 64. Wobst, P.and C. Arndt. “HIV/AIDS and primary school performance in Tanzania”. Presented at the 25th International Conference of Agricultural Economists. Durban, South Africa. 65. World Bank, 2004. Making Services Work for the Poor. World Development Report 2004. 66. Worthington, A. 2001. “An empirical survey of frontier efficiency measurement techniques in education” Education Economics, Vol. 9, No. 3 50 Appendix A. International Benchmarking Exercise Table A.1 Efficiency Score for Selected Education Indicators (Single Input - Output) PISA Math Score  PISA Science Score    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  ALB  0.678  0.678  0.738  0.738  0.773  0.688  0.774  0.774  ARG  0.697  0.629  0.789  0.789  0.697  0.650  0.829  0.824  AUS  0.878  0.825  0.942  0.942  0.947  0.910  0.984  0.981  AUT  0.765  0.703  0.928  0.928  0.715  0.710  0.936  0.929  AZE  0.999  0.915  0.886  0.885  0.876  0.876  0.769  0.769  BGR  0.831  0.766  0.824  0.813  0.831  0.780  0.860  0.841  BRA  0.579  0.579  0.713  0.713  0.660  0.582  0.762  0.756  CAN  0.877  0.862  0.974  0.974  1.000  1.000  1.000  1.000  CHE  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  CHL  0.798  0.720  0.788  0.788  0.798  0.754  0.848  0.847  CHN  0.959  0.924  0.993  0.993  0.959  0.918  0.983  0.981  COL  0.633  0.633  0.720  0.720  0.721  0.642  0.774  0.771  CRI  0.530  0.530  0.748  0.748  0.604  0.547  0.794  0.783  CYP  0.548  0.507  0.817  0.817  0.548  0.506  0.819  0.802  CZE  1.000  0.973  1.000  0.981  1.000  1.000  1.000  1.000  DEU  0.973  0.927  0.966  0.961  1.000  0.989  1.000  0.996  DNK  0.420  0.398  0.948  0.948  0.420  0.400  0.923  0.923  DOM  0.745  0.745  0.621  0.615  0.745  0.745  0.644  0.636  ESP  0.870  0.868  0.918  0.914  0.910  0.889  0.959  0.949  FIN  0.567  0.565  0.983  0.983  1.000  1.000  1.000  1.000  FRA  0.768  0.704  0.925  0.925  0.768  0.719  0.940  0.933  GBR  0.706  0.686  0.920  0.920  0.776  0.756  0.968  0.961  GEO  0.699  0.699  0.755  0.755  0.797  0.712  0.780  0.779  HRV  0.813  0.783  0.864  0.864  0.813  0.808  0.912  0.911  HUN  0.798  0.794  0.903  0.903  0.834  0.814  0.930  0.928  IDN  0.679  0.679  0.708  0.708  0.679  0.679  0.744  0.743  ISL  0.512  0.472  0.930  0.930  0.458  0.458  0.894  0.894  ITA  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  KAZ  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  KGZ  0.571  0.571  0.619  0.619  0.571  0.571  0.625  0.618  KOR  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  LBN  0.913  0.913  0.978  0.860  0.913  0.913  0.966  0.844  LTU  0.772  0.762  0.893  0.893  0.772  0.770  0.915  0.914  LVA  0.665  0.661  0.901  0.901  0.696  0.684  0.931  0.923  MDA  0.614  0.552  0.784  0.784  0.614  0.564  0.810  0.799  MEX  0.707  0.629  0.773  0.773  0.707  0.636  0.787  0.783  51 Table A.1 (continued)  PISA Math Score  PISA Science Score    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  MKD  0.653  0.653  0.694  0.694  0.653  0.653  0.726  0.724  NLD  0.789  0.772  0.970  0.970  0.851  0.805  0.975  0.970  NOR  0.482  0.448  0.934  0.934  0.482  0.457  0.921  0.921  NZL  0.607  0.575  0.948  0.948  0.655  0.642  0.989  0.968  PAN  0.855  0.855  0.740  0.727  0.855  0.855  0.776  0.765  PER  0.704  0.704  0.702  0.702  0.704  0.704  0.727  0.725  POL  0.855  0.794  0.934  0.934  0.855  0.830  0.955  0.954  ROM  0.955  0.882  0.884  0.866  0.955  0.881  0.869  0.869  RUS  0.915  0.907  0.976  0.927  0.915  0.912  0.972  0.953  SVK  0.946  0.945  0.987  0.960  0.946  0.933  0.958  0.953  SVN  0.764  0.720  0.945  0.945  0.785  0.768  0.970  0.963  SWE  0.524  0.479  0.923  0.923  0.490  0.486  0.911  0.911  THA  0.781  0.700  0.780  0.780  0.781  0.711  0.801  0.801  TTO  0.715  0.639  0.777  0.777  0.715  0.645  0.790  0.786  TUN  0.556  0.556  0.689  0.689  0.556  0.556  0.745  0.737  TUR  0.782  0.719  0.809  0.809  0.782  0.731  0.832  0.832  URY  0.716  0.647  0.790  0.790  0.716  0.660  0.816  0.812  USA  0.805  0.795  0.894  0.894  0.900  0.852  0.947  0.946  VNM  0.756  0.692  0.924  0.924  0.815  0.807  0.993  0.985  52 Table A.2 Efficiency Score for Selected Education Indicators (Single Input - Output) Secondary Education Gross School enrollment, primary (% gross) Enrollment Input Efficiency Output Efficiency Input Efficiency Output Efficiency Country FDH DEA FDH DEA FDH DEA FDH DEA AGO 0.985 0.715 0.991 0.840 0.613 0.613 0.170 0.170 ALB 0.645 0.645 0.926 0.776 0.651 0.651 0.639 0.588 ARG 0.910 0.617 0.778 0.778 0.571 0.571 0.644 0.644 ARM 0.642 0.642 0.883 0.738 0.647 0.647 0.740 0.677 ATG 0.944 0.944 0.913 0.890 AUS 0.614 0.614 0.887 0.757 1.000 1.000 1.000 1.000 AUT 0.526 0.526 0.690 0.690 0.648 0.549 0.708 0.708 AZE 0.865 0.865 0.909 0.851 0.867 0.867 0.992 0.892 BDI 0.966 0.776 0.886 0.886 0.550 0.550 0.172 0.172 BEN 0.964 0.696 0.831 0.831 0.559 0.559 0.295 0.295 BFA 0.554 0.554 0.566 0.566 0.559 0.559 0.156 0.156 BGD 0.877 0.613 0.920 0.780 0.596 0.596 0.347 0.347 BGR 0.695 0.695 0.920 0.771 0.701 0.701 0.740 0.721 BLR 0.584 0.584 0.836 0.704 BLZ 0.772 0.555 0.768 0.768 0.527 0.527 0.540 0.540 BOL 0.513 0.513 0.690 0.690 0.519 0.519 0.575 0.575 BRA 0.928 0.671 0.831 0.831 0.756 0.571 0.726 0.726 BRB 0.529 0.529 0.670 0.670 0.750 0.577 0.736 0.736 BWA 0.413 0.413 0.737 0.737 0.421 0.421 0.576 0.576 CAF 0.560 0.560 0.626 0.626 CAN 0.614 0.614 0.835 0.712 0.875 0.663 0.728 0.728 CHE 0.687 0.687 0.943 0.786 0.703 0.703 0.764 0.747 CHL 0.655 0.655 0.900 0.760 0.662 0.662 0.718 0.669 CHN 0.666 0.666 0.960 0.819 0.672 0.672 0.654 0.617 CIV 0.557 0.557 0.580 0.580 0.563 0.563 0.199 0.199 CMR 0.856 0.580 0.907 0.758 0.582 0.582 0.304 0.304 COG 0.791 0.564 0.765 0.765 COL 0.994 0.717 0.988 0.839 0.600 0.600 0.663 0.663 COM 0.551 0.551 0.717 0.717 CPV 0.822 0.588 0.767 0.767 0.560 0.560 0.609 0.609 CRI 0.782 0.529 0.777 0.777 0.592 0.493 0.700 0.700 CYP 0.422 0.422 0.689 0.689 0.431 0.431 0.688 0.688 CZE 0.793 0.793 0.926 0.831 0.801 0.801 0.940 0.816 DEU 0.688 0.688 0.948 0.791 0.979 0.740 0.814 0.794 DJI 0.563 0.563 0.449 0.449 DNK 0.278 0.278 0.689 0.689 0.400 0.377 0.859 0.859 DOM 0.720 0.720 0.958 0.818 0.724 0.724 0.759 0.610 53 Table A.2 (continued) School enrollment, primary (% gross) Secondary Education Gross Enrollment Input Efficiency Output Efficiency Input Efficiency Output Efficiency Country FDH DEA FDH DEA FDH DEA FDH DEA ECU 0.935 0.644 0.950 0.796 0.585 0.585 0.585 0.585 EGY 0.636 0.636 0.927 0.771 0.641 0.641 0.645 0.586 ESP 0.706 0.706 0.963 0.814 1.000 0.989 1.000 0.991 EST 0.585 0.585 0.837 0.706 0.830 0.640 0.737 0.737 ETH 0.551 0.551 0.657 0.657 0.557 0.557 0.234 0.234 FIN 0.384 0.384 0.685 0.685 0.551 0.456 0.778 0.778 FJI 0.600 0.600 0.885 0.745 FRA 0.536 0.536 0.731 0.731 0.764 0.658 0.801 0.801 GAB 0.805 0.805 0.533 0.465 GBR 0.525 0.525 0.731 0.731 0.646 0.546 0.707 0.707 GEO 1.000 0.736 1.000 0.858 0.678 0.678 0.741 0.704 GHA 0.532 0.532 0.736 0.736 0.538 0.538 0.421 0.421 GIN 0.558 0.558 0.597 0.597 0.564 0.564 0.280 0.280 GMB 0.560 0.560 0.586 0.586 0.565 0.565 0.380 0.380 GNB 0.844 0.620 0.922 0.780 GTM 0.943 0.651 0.971 0.808 0.640 0.640 0.483 0.438 GUY 0.616 0.616 0.721 0.616 0.748 0.639 0.712 0.712 HND 0.791 0.575 0.772 0.772 0.539 0.539 0.494 0.494 HRV 0.670 0.670 0.839 0.718 0.677 0.677 0.764 0.725 HTI 0.576 0.576 0.317 0.317 HUN 0.651 0.651 0.890 0.750 0.794 0.671 0.792 0.736 IDN 0.649 0.649 0.945 0.794 0.654 0.654 0.629 0.581 IND 0.592 0.592 0.904 0.767 0.598 0.598 0.444 0.444 IRN 0.752 0.752 0.993 0.868 0.757 0.757 0.851 0.708 ISL 0.344 0.344 0.675 0.675 0.494 0.412 0.782 0.782 ITA 0.821 0.821 0.930 0.849 1.000 0.860 1.000 0.894 JAM 0.533 0.533 0.636 0.636 KAZ 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 KEN 0.812 0.592 0.773 0.773 0.552 0.552 0.426 0.426 KGZ 0.536 0.536 0.700 0.700 0.542 0.542 0.616 0.616 KHM 1.000 0.740 1.000 0.853 0.602 0.602 0.308 0.308 KNA 0.971 0.971 0.806 0.796 KOR 0.668 0.668 0.884 0.756 0.678 0.678 0.779 0.740 LAO 1.000 0.739 1.000 0.854 0.621 0.621 0.340 0.340 LBN 0.913 0.913 0.909 0.872 0.913 0.913 0.814 0.761 LKA 0.757 0.757 0.913 0.800 0.760 0.760 0.925 0.772 LSO 0.521 0.521 0.744 0.744 0.527 0.527 0.330 0.330 LTU 0.626 0.626 0.886 0.731 0.888 0.662 0.806 0.727 LVA 0.532 0.532 0.704 0.704 0.541 0.541 0.694 0.694 54 Table A.2 (continued) School enrollment, primary (% gross) Secondary Education Gross Enrollment Input Efficiency Output Efficiency Input Efficiency Output Efficiency Country FDH DEA FDH DEA FDH DEA FDH DEA MAR 0.914 0.623 0.927 0.781 0.571 0.571 0.440 0.440 MDA 0.500 0.500 0.638 0.638 0.507 0.507 0.628 0.628 MDG 1.000 1.000 1.000 1.000 0.569 0.569 0.231 0.231 MEX 0.576 0.576 0.859 0.718 0.583 0.583 0.631 0.631 MKD 0.616 0.616 0.744 0.635 0.623 0.623 0.596 0.596 MLI 0.562 0.562 0.545 0.545 0.567 0.567 0.278 0.278 MNG 0.916 0.664 0.969 0.812 0.623 0.623 0.669 0.669 MOZ 0.543 0.543 0.723 0.723 0.548 0.548 0.172 0.172 MRT 0.585 0.585 0.803 0.677 0.590 0.590 0.185 0.185 MUS 0.699 0.699 0.941 0.792 0.705 0.705 0.718 0.704 MWI 0.975 0.908 0.960 0.960 0.555 0.555 0.229 0.229 MYS 0.553 0.553 0.693 0.693 0.562 0.562 0.491 0.491 NAM 0.441 0.441 0.746 0.746 0.448 0.448 0.457 0.457 NER 0.549 0.549 0.461 0.461 NGA 0.643 0.643 0.783 0.655 0.647 0.647 0.307 0.281 NIC 1.000 0.742 1.000 0.851 0.580 0.580 0.492 0.492 NLD 0.543 0.543 0.731 0.731 0.776 0.755 0.881 0.881 NOR 0.315 0.315 0.681 0.681 0.454 0.388 0.795 0.795 NPL 0.997 0.916 0.953 0.953 0.567 0.567 0.363 0.363 NZL 0.417 0.417 0.680 0.680 0.596 0.556 0.852 0.852 OMN 0.751 0.751 0.985 0.860 0.764 0.764 0.941 0.788 PAK 0.612 0.612 0.780 0.664 0.617 0.617 0.247 0.247 PAN 0.836 0.836 0.968 0.891 0.839 0.839 0.758 0.665 PER 0.675 0.675 0.963 0.795 0.680 0.680 0.735 0.699 PHL 0.955 0.691 0.992 0.831 0.648 0.648 0.671 0.615 POL 0.620 0.620 0.878 0.720 0.757 0.634 0.787 0.704 PRY 0.586 0.586 0.863 0.728 0.592 0.592 0.482 0.482 ROM 0.813 0.813 0.872 0.793 0.818 0.818 0.933 0.824 RUS 0.762 0.762 0.916 0.806 0.769 0.769 0.885 0.746 RWA 0.979 0.894 0.949 0.949 0.557 0.557 0.220 0.220 SAU 0.458 0.458 0.706 0.706 0.657 0.509 0.740 0.740 SDN 0.613 0.613 0.585 0.498 SEN 0.537 0.537 0.557 0.557 0.543 0.543 0.253 0.253 SLB 0.815 0.559 0.782 0.782 SLE 1.000 0.782 1.000 0.873 0.569 0.569 0.239 0.239 SLV 1.000 0.712 1.000 0.840 0.624 0.624 0.473 0.473 SVK 0.788 0.788 0.930 0.832 0.795 0.795 0.922 0.797 SVN 0.541 0.541 0.674 0.674 0.550 0.550 0.690 0.690 SWE 0.522 0.355 0.750 0.750 0.505 0.376 0.719 0.719 55 Table A.2 (continued) School enrollment, primary (% gross) Secondary Education Gross Enrollment Input Efficiency Output Efficiency Input Efficiency Output Efficiency Country FDH DEA FDH DEA FDH DEA FDH DEA SWZ 0.765 0.562 0.776 0.776 0.522 0.522 0.411 0.411 TCD 0.575 0.575 0.745 0.622 0.580 0.580 0.161 0.161 TGO 0.977 0.747 0.863 0.863 THA 0.648 0.648 0.869 0.731 0.655 0.655 0.647 0.598 TJK 0.560 0.560 0.675 0.675 0.565 0.565 0.611 0.611 TKM 0.711 0.711 0.818 0.694 TTO 0.566 0.566 0.863 0.728 0.577 0.577 0.627 0.627 TUN 0.763 0.519 0.750 0.750 0.521 0.521 0.649 0.649 TUR 0.641 0.641 0.900 0.752 0.649 0.649 0.659 0.604 TZA 0.562 0.562 0.622 0.622 0.568 0.568 0.214 0.214 UGA 0.918 0.642 0.937 0.793 0.574 0.574 0.190 0.190 UKR 0.510 0.510 0.715 0.715 0.517 0.517 0.682 0.682 URY 0.863 0.601 0.919 0.771 0.589 0.589 0.646 0.646 USA 0.620 0.620 0.881 0.723 0.634 0.634 0.757 0.681 VCT 0.552 0.552 0.713 0.713 VEN 0.591 0.591 0.840 0.712 0.599 0.599 0.600 0.600 VNM 0.558 0.558 0.728 0.728 0.564 0.564 0.515 0.515 VUT 0.923 0.673 0.819 0.819 ZAF 0.527 0.527 0.666 0.666 0.643 0.536 0.700 0.700 ZAR 0.557 0.557 0.683 0.683 ZWE 0.533 0.533 0.691 0.691 0.539 0.539 0.300 0.300 56 Table A.3 Efficiency Score for Selected Education Indicators (Two Inputs – Single Output) Input (Public Expenditure‐ Adult Literacy Rate)    Primary Net Enrollment  Secondary Net Enrollment    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  0.962  0.847  0.873  0.824  0.963  0.838  0.545  0.306  ALB  0.869  0.798  0.941  0.937  0.940  0.833  0.933  0.766  AZE  1.000  0.942  1.000  0.970  BDI  0.947  0.820  0.931  0.926  0.904  0.792  0.477  0.346  BEN  0.984  0.964  0.959  0.958  1.000  1.000  1.000  1.000  BFA  0.951  0.910  0.660  0.660  0.973  0.910  0.585  0.478  BGD  1.000  0.948  1.000  0.917  1.000  0.957  1.000  0.839  BGR  0.904  0.845  0.965  0.962  0.927  0.854  0.961  0.794  BLR  0.816  0.679  0.930  0.930  0.983  0.845  0.840  0.840  BOL  0.757  0.656  0.906  0.904  0.911  0.731  0.851  0.692  BRA  0.740  0.654  0.941  0.938  1.000  0.811  1.000  0.822  CAF  0.975  0.921  0.696  0.695  0.974  0.918  0.361  0.284  CHL  0.868  0.750  0.944  0.940  0.938  0.832  0.976  0.767  CHN  1.000  1.000  1.000  1.000  0.928  0.825  0.913  0.753  CIV  0.947  0.875  0.709  0.708  0.969  0.894  0.722  0.655  CMR  0.923  0.815  0.954  0.899  0.964  0.801  0.746  0.533  COG  0.804  0.719  0.917  0.909  0.874  0.710  0.685  0.513  COL  0.877  0.719  0.925  0.922  1.000  0.861  1.000  0.817  COM  0.957  0.861  0.826  0.823  1.000  0.908  1.000  0.833  CPV  0.856  0.766  0.973  0.971  1.000  0.833  1.000  0.790  CRI  0.727  0.623  0.980  0.976  1.000  0.850  1.000  0.868  CYP  0.551  0.504  0.982  0.982  0.923  0.717  0.756  0.756  DOM  1.000  0.848  1.000  0.878  0.999  0.862  0.898  0.731  ECU  0.873  0.742  0.952  0.949  0.995  0.859  1.000  0.820  EGY  1.000  0.904  1.000  0.980  1.000  0.900  1.000  0.848  ESP  1.000  0.867  1.000  0.997  1.000  1.000  1.000  1.000  EST  0.771  0.677  0.963  0.963  0.980  0.828  0.847  0.847  GEO  0.953  0.952  1.000  0.998  0.916  0.873  0.984  0.828  GHA  0.816  0.742  0.858  0.852  0.914  0.750  0.848  0.592  GIN  1.000  0.958  1.000  0.799  1.000  0.992  1.000  0.856  GMB  0.965  0.890  0.704  0.702  1.000  0.963  1.000  0.928  GNB  0.989  0.890  0.965  0.691  GTM  1.000  0.841  1.000  0.920  1.000  0.835  1.000  0.642  GUY  0.961  0.780  0.816  0.814  1.000  0.883  1.000  0.829  HND  0.857  0.732  0.961  0.960  0.857  0.725  0.796  0.626  HRV  0.851  0.728  0.880  0.880  0.997  0.850  0.976  0.799  57   Table A.3 (continued)    Input (Public Expenditure‐ Adult Literacy Rate)    Primary Net Enrollment  Secondary Net Enrollment    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  IDN  0.898  0.797  0.924  0.921  0.898  0.803  0.812  0.694  IND  0.939  0.836  0.975  0.919  1.000  0.854  1.000  0.742  IRN  1.000  1.000  1.000  1.000  1.000  0.917  1.000  0.876  ITA  0.891  0.854  0.977  0.977  1.000  0.891  1.000  0.850  KAZ  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  KEN  0.855  0.741  0.851  0.843  0.910  0.746  0.759  0.572  KGZ  0.812  0.673  0.883  0.883  0.860  0.747  0.691  0.691  KHM  0.997  0.871  0.982  0.962  LAO  1.000  0.926  1.000  0.945  1.000  0.903  1.000  0.673  LBN  1.000  1.000  1.000  1.000  1.000  1.000  1.000  0.739  LKA  1.000  0.993  1.000  0.995  1.000  0.983  1.000  0.974  LSO  0.798  0.709  0.807  0.801  0.835  0.713  0.647  0.501  LTU  0.845  0.733  0.971  0.971  0.981  0.832  0.811  0.811  LVA  0.737  0.649  0.974  0.974  0.980  0.789  0.827  0.827  MAR  0.951  0.824  0.973  0.967  1.000  0.829  1.000  0.724  MDA  0.742  0.629  0.878  0.878  0.861  0.715  0.688  0.688  MDG  0.905  0.795  0.662  0.385  MEX  0.819  0.709  0.958  0.954  0.898  0.751  0.851  0.694  MLI  0.968  0.933  0.621  0.621  0.995  0.972  0.895  0.768  MNG  0.858  0.796  0.966  0.966  0.928  0.819  0.947  0.749  MOZ  0.931  0.851  0.878  0.875  0.931  0.827  0.461  0.375  MUS  0.921  0.863  0.967  0.964  0.996  0.880  0.937  0.827  MWI  0.966  0.858  0.989  0.984  0.886  0.794  0.581  0.415  MYS  0.802  0.681  0.983  0.980  0.791  0.658  0.713  0.577  NAM  0.632  0.568  0.879  0.877  NER  1.000  1.000  1.000  1.000  1.000  1.000  1.000  0.639  NPL  0.987  0.880  0.988  0.982  1.000  0.866  1.000  0.750  PAK  1.000  0.898  1.000  0.734  1.000  0.898  1.000  0.500  PAN  1.000  1.000  1.000  1.000  1.000  0.906  1.000  0.723  PER  0.921  0.848  0.948  0.945  0.971  0.887  0.962  0.838  PHL  0.908  0.824  0.940  0.937  0.946  0.816  0.873  0.736  PRY  0.882  0.723  0.895  0.891  0.894  0.729  0.715  0.572  ROM  0.954  0.854  0.918  0.906  0.982  0.916  0.987  0.879  RUS  0.915  0.852  0.962  0.959  0.915  0.861  0.973  0.812  RWA  0.967  0.835  0.979  0.973  0.876  0.781  0.573  0.388  SAU  0.547  0.488  0.933  0.930  1.000  0.817  1.000  0.840  58 Table A.3 (continued)  Input (Public Expenditure‐ Adult Literacy Rate)    Primary Net Enrollment  Secondary Net Enrollment    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  SEN  0.902  0.820  0.716  0.714  0.924  0.830  0.718  0.591  SLE  1.000  1.000  1.000  1.000  1.000  0.997  1.000  0.863  SLV  0.983  0.816  0.944  0.943  0.983  0.811  0.854  0.703  SWZ  0.752  0.663  0.811  0.803  0.834  0.679  0.759  0.546  TCD  1.000  1.000  1.000  0.854  1.000  0.992  1.000  0.580  TGO  0.965  0.835  0.947  0.942  0.981  0.828  0.875  0.646  THA  0.894  0.782  0.938  0.935  0.967  0.871  0.986  0.818  TUN  1.000  0.721  1.000  0.993  1.000  0.820  1.000  0.843  TUR  0.896  0.754  0.951  0.948  0.969  0.816  0.929  0.745  TZA  0.894  0.777  0.847  0.841  0.895  0.767  0.409  0.322  UGA  0.914  0.814  0.990  0.933  0.914  0.788  0.342  0.239  UKR  0.749  0.660  0.961  0.961  0.912  0.755  0.755  0.755  URY  0.857  0.725  0.983  0.983  0.927  0.785  0.751  0.751  VNM  0.902  0.778  0.989  0.986  ZAF  0.975  0.767  0.935  0.754  ZAR  0.895  0.766  0.516  0.405  ZWE  0.830  0.710  0.879  0.877  0.832  0.697  0.523  0.413  59 Table A.4 Efficiency Score for Selected Health Indicators (Single Inputs – Output)   Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  0.504  0.504  0.76  0.746  0.439  0.439  0.704  0.704  ALB  0.527  0.527  0.954  0.94  0.775  0.638  0.997  0.997  ARG  0.551  0.551  0.951  0.94  0.495  0.495  0.944  0.941  ARM  0.521  0.521  0.927  0.913  0.454  0.454  0.951  0.951  ATG  0.844  0.594  0.96  0.952  0.826  0.68  1  0.997  AUS  1  1  1  1  0.415  0.415  0.929  0.929  AUT  0.768  0.653  0.98  0.98  0.31  0.31  0.918  0.918  AZE  0.619  0.619  0.875  0.874  0.565  0.565  0.899  0.897  BDI  0.472  0.472  0.7  0.7  0.404  0.404  0.962  0.962  BEN  0.481  0.481  0.777  0.759  0.413  0.413  0.789  0.789  BFA  0.476  0.476  0.711  0.711  0.409  0.409  0.915  0.915  BGD  0.491  0.491  0.882  0.863  0.424  0.424  0.968  0.968  BGR  0.529  0.529  0.924  0.911  0.463  0.463  0.941  0.941  BHR  0.897  0.897  0.984  0.975  0.881  0.881  1  0.999  BHS  0.586  0.586  0.921  0.916  0.523  0.523  0.983  0.98  BLR  0.529  0.529  0.893  0.88  0.472  0.472  0.987  0.987  BLZ  0.503  0.503  0.888  0.871  0.433  0.433  0.967  0.967  BOL  0.492  0.492  0.899  0.88  0.423  0.423  0.957  0.957  BRA  0.528  0.528  0.928  0.915  0.465  0.465  0.978  0.978  BRB  0.519  0.519  0.955  0.94  0.448  0.448  0.922  0.922  BWA  0.531  0.531  0.717  0.707  0.467  0.467  0.961  0.961  CAF  0.475  0.475  0.583  0.582  0.407  0.407  0.43  0.43  CAN  0.849  0.771  0.986  0.986  0.343  0.343  0.916  0.916  CHE  1  1  1  1  0.34  0.34  0.968  0.968  CHL  0.86  0.857  0.995  0.988  0.51  0.51  0.944  0.941  CHN  0.534  0.534  0.933  0.921  0.973  0.973  1  1  CIV  0.49  0.49  0.712  0.696  0.421  0.421  0.792  0.792  CMR  0.491  0.491  0.722  0.707  0.423  0.423  0.853  0.853  COG  0.501  0.501  0.752  0.737  0.431  0.431  0.828  0.828  COL  0.732  0.515  0.96  0.94  0.425  0.425  0.906  0.906  COM  0.478  0.478  0.826  0.807  0.411  0.411  0.853  0.853  CPV  0.495  0.495  0.906  0.888  0.429  0.429  0.957  0.957  CRI  0.973  0.777  0.98  0.979  0.394  0.394  0.906  0.906  CYP  1  1  1  1  1  0.823  1  0.999  CZE  0.706  0.578  0.95  0.95  0.42  0.414  0.993  0.993  DEU  0.776  0.612  0.974  0.974  0.311  0.311  0.961  0.961  DJI  0.473  0.473  0.785  0.785  0.406  0.406  0.85  0.85  60 Table A.4 (continued)    Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  DMA  0.512  0.512  0.915  0.899  0.442  0.442  0.986  0.986  DNK  0.532  0.496  0.967  0.967  0.278  0.278  0.931  0.931  DOM  0.546  0.546  0.933  0.922  0.477  0.477  0.863  0.863  DZA  0.773  0.527  0.957  0.942  0.444  0.444  0.96  0.96  ECU  0.784  0.558  0.961  0.947  0.454  0.454  0.877  0.877  EGY  0.537  0.537  0.89  0.879  0.472  0.472  0.962  0.962  ERI  0.48  0.48  0.778  0.76  0.412  0.412  0.945  0.945  ESP  1  0.943  1  0.998  0.386  0.386  0.977  0.977  EST  0.536  0.536  0.943  0.93  0.461  0.461  0.947  0.947  ETH  0.476  0.476  0.749  0.749  0.408  0.408  0.691  0.691  FIN  0.917  0.752  0.982  0.98  0.384  0.382  0.994  0.994  FJI  0.512  0.512  0.814  0.8  0.927  0.927  1  1  FRA  0.768  0.73  0.99  0.99  0.318  0.318  0.996  0.996  GAB  0.598  0.598  0.794  0.791  0.521  0.521  0.766  0.764  GBR  0.853  0.665  0.973  0.973  0.344  0.344  0.957  0.957  GEO  0.526  0.526  0.915  0.902  0.46  0.46  0.929  0.929  GHA  0.485  0.485  0.791  0.773  0.419  0.419  0.934  0.934  GIN  0.478  0.478  0.732  0.715  0.409  0.409  0.597  0.597  GMB  0.476  0.476  0.805  0.805  0.408  0.408  0.98  0.98  GNB  0.48  0.48  0.704  0.687  0.412  0.412  0.861  0.861  GRC  0.981  0.798  0.982  0.981  0.85  0.85  1  1  GRD  0.534  0.534  0.892  0.88  0.466  0.466  0.973  0.973  GTM  0.513  0.513  0.894  0.879  0.446  0.446  0.869  0.869  GUY  0.49  0.49  0.83  0.812  0.425  0.425  0.973  0.973  HND  0.483  0.483  0.897  0.877  0.415  0.415  0.98  0.98  HRV  0.46  0.46  0.932  0.931  0.399  0.399  0.967  0.967  HTI  0.481  0.481  0.765  0.747  0.413  0.413  0.638  0.638  HUN  0.515  0.515  0.933  0.917  0.953  0.953  1  1  IDN  0.548  0.548  0.878  0.868  0.484  0.484  0.816  0.814  IND  0.507  0.507  0.832  0.817  0.44  0.44  0.825  0.825  IRN  0.574  0.574  0.931  0.924  0.854  0.703  1  0.998  ISL  0.878  0.869  0.999  0.995  0.37  0.37  0.937  0.937  ITA  0.899  0.879  0.998  0.994  0.373  0.373  0.965  0.965  JAM  0.513  0.513  0.936  0.92  0.444  0.444  0.935  0.935  JOR  0.478  0.478  0.938  0.916  0.422  0.414  0.991  0.991  KAZ  0.632  0.632  0.86  0.856  0.576  0.576  0.993  0.991  KEN  0.485  0.485  0.784  0.766  0.416  0.416  0.918  0.918  KGZ  0.481  0.481  0.861  0.842  0.413  0.413  0.971  0.971  KHM  0.491  0.491  0.829  0.812  0.423  0.423  0.89  0.89  61 Table A.4 (continued)    Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  KNA  0.59  0.59  0.973  0.971  KOR  1  0.99  1  0.999  0.561  0.561  0.987  0.985  LAO  0.505  0.505  0.802  0.787  0.439  0.439  0.811  0.811  LBN  0.803  0.757  0.988  0.976  0.485  0.485  0.821  0.818  LCA  0.514  0.514  0.942  0.926  0.444  0.444  0.99  0.99  LKA  0.541  0.541  0.948  0.937  0.802  0.66  0.997  0.997  LSO  0.467  0.467  0.571  0.571  0.396  0.396  0.944  0.944  LTU  0.536  0.536  0.915  0.902  0.473  0.473  0.948  0.948  LUX  0.962  0.888  0.992  0.991  0.897  0.897  1  1  LVA  0.554  0.554  0.91  0.901  0.498  0.498  0.941  0.938  MAR  0.515  0.515  0.922  0.907  0.933  0.933  1  1  MDA  0.476  0.476  0.862  0.862  0.409  0.409  0.905  0.905  MDG  0.478  0.478  0.77  0.752  0.41  0.41  0.73  0.73  MEX  0.56  0.56  0.95  0.941  0.49  0.49  0.931  0.928  MKD  0.504  0.504  0.925  0.908  0.441  0.441  0.961  0.961  MLI  0.48  0.48  0.744  0.726  0.413  0.413  0.698  0.698  MNG  0.525  0.525  0.828  0.815  0.463  0.463  0.988  0.988  MOZ  0.475  0.475  0.688  0.688  0.407  0.407  0.775  0.775  MRT  0.495  0.495  0.853  0.836  0.427  0.427  0.75  0.75  MUS  0.592  0.592  0.92  0.915  0.528  0.528  0.993  0.991  MWI  0.47  0.47  0.663  0.663  0.402  0.402  0.934  0.934  MYS  0.655  0.655  0.94  0.938  0.596  0.596  0.983  0.981  NAM  0.494  0.494  0.733  0.718  0.423  0.423  0.867  0.867  NER  0.474  0.474  0.72  0.72  0.407  0.407  0.703  0.703  NGA  0.511  0.511  0.766  0.753  0.446  0.446  0.506  0.506  NIC  0.718  0.571  0.971  0.949  0.415  0.415  0.99  0.99  NLD  0.711  0.609  0.981  0.981  0.28  0.28  0.975  0.975  NOR  0.759  0.697  0.987  0.987  0.305  0.305  0.949  0.949  NPL  0.481  0.481  0.859  0.839  0.414  0.414  0.906  0.906  NZL  0.782  0.671  0.981  0.981  0.305  0.305  0.938  0.938  OMN  1  1  1  1  1  1  1  1  PAK  0.507  0.507  0.823  0.808  0.439  0.439  0.743  0.743  PAN  0.746  0.638  0.978  0.959  0.43  0.43  0.843  0.843  PER  0.785  0.709  0.983  0.969  0.462  0.462  0.921  0.921  PHL  0.515  0.515  0.865  0.85  0.448  0.448  0.804  0.804  PNG  0.478  0.478  0.741  0.724  0.411  0.411  0.737  0.737  POL  0.788  0.546  0.958  0.945  0.786  0.647  0.997  0.997  PRY  0.501  0.501  0.929  0.911  0.429  0.429  0.889  0.889  ROM  0.528  0.528  0.926  0.912  0.462  0.462  0.929  0.929  62 Table A.4 (continued)    Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  RUS  0.577  0.577  0.861  0.855  0.515  0.515  0.984  0.982  RWA  0.477  0.477  0.778  0.778  0.41  0.41  0.986  0.986  SAU  0.99  0.99  0.988  0.988  1  1  1  1  SDN  0.494  0.494  0.835  0.818  0.426  0.426  0.918  0.918  SEN  0.482  0.482  0.803  0.785  0.415  0.415  0.906  0.906  SLB  0.472  0.472  0.758  0.757  0.405  0.405  0.952  0.952  SLE  0.478  0.478  0.701  0.684  0.411  0.411  0.882  0.882  SLV  0.493  0.493  0.932  0.912  0.425  0.425  0.921  0.921  SVK  0.497  0.497  0.947  0.928  0.442  0.433  0.991  0.991  SVN  0.678  0.665  0.969  0.968  0.392  0.392  0.965  0.965  SWE  0.742  0.696  0.989  0.989  0.284  0.284  0.99  0.99  SWZ  0.473  0.473  0.613  0.613  0.403  0.403  0.937  0.937  TCD  0.484  0.484  0.712  0.696  0.416  0.416  0.372  0.372  TGO  0.478  0.478  0.748  0.73  0.41  0.41  0.85  0.85  THA  0.546  0.546  0.952  0.941  1  1  1  1  TJK  0.486  0.486  0.874  0.854  0.418  0.418  0.96  0.96  TKM  0.57  0.57  0.851  0.844  0.512  0.512  0.986  0.983  TON  0.486  0.486  0.876  0.856  0.422  0.422  0.815  0.815  TTO  0.697  0.697  0.946  0.912  0.631  0.631  0.93  0.928  TUN  0.759  0.542  0.961  0.944  0.441  0.441  0.99  0.99  TUR  0.799  0.686  0.978  0.966  0.475  0.475  0.978  0.978  TZA  0.482  0.482  0.752  0.735  0.415  0.415  0.929  0.929  UGA  0.479  0.479  0.725  0.708  0.412  0.412  0.798  0.798  UKR  0.496  0.496  0.875  0.857  0.43  0.43  0.535  0.535  URY  0.722  0.498  0.958  0.937  0.422  0.422  0.958  0.958  USA  0.562  0.494  0.952  0.952  0.316  0.316  0.958  0.958  UZB  0.495  0.495  0.859  0.841  0.89  0.74  0.999  0.999  VCT  0.506  0.506  0.893  0.877  0.436  0.436  0.986  0.986  VEN  0.605  0.605  0.94  0.937  0.542  0.542  0.822  0.82  VNM  0.494  0.494  0.917  0.898  0.425  0.425  0.912  0.912  VUT  0.481  0.481  0.789  0.771  0.413  0.413  0.652  0.652  WSM  0.481  0.481  0.895  0.874  0.411  0.411  0.629  0.629  YEM  0.497  0.497  0.827  0.81  0.43  0.43  0.726  0.726  ZAF  0.514  0.514  0.71  0.698  0.445  0.445  0.732  0.732  ZAR  0.475  0.475  0.717  0.717  0.407  0.407  0.745  0.745  ZMB  0.489  0.489  0.659  0.645  0.422  0.422  0.853  0.853  ZWE  0.481  0.481  0.656  0.641  0.412  0.412  0.899  0.899  63 Table A.5 Efficiency Score for Selected Health Indicators (Single Inputs – Output) Maternal Survival Rate  Infant Survival Rate      Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  Country  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  0.439  0.439  0.995  0.995  0.439  0.439  0.935  0.935  ALB  0.462  0.462  1.000  1.000  0.462  0.462  0.989  0.988  ARG  0.495  0.495  1.000  1.000  0.509  0.499  0.991  0.990  ARM  0.454  0.454  1.000  1.000  0.454  0.454  0.988  0.988  ATG  0.826  0.602  0.996  0.996  AUS  0.882  0.416  1.000  1.000  0.697  0.646  0.999  0.998  AUT  0.659  0.625  1.000  1.000  0.520  0.495  0.999  0.999  AZE  0.565  0.565  1.000  1.000  0.565  0.565  0.971  0.971  BDI  0.404  0.404  0.993  0.993  0.404  0.404  0.946  0.945  BEN  0.413  0.413  0.996  0.996  0.413  0.413  0.934  0.933  BFA  0.409  0.409  0.996  0.996  0.409  0.409  0.941  0.941  BGD  0.424  0.424  0.998  0.998  0.424  0.424  0.967  0.967  BGR  0.776  0.463  1.000  1.000  0.541  0.512  0.994  0.994  BHR  0.856  0.856  1.000  1.000  1.000  1.000  1.000  1.000  BHS  0.523  0.523  0.999  0.999  0.537  0.534  0.992  0.992  BLR  1.000  0.472  1.000  1.000  0.792  0.733  0.999  0.999  BLZ  0.433  0.433  1.000  1.000  0.433  0.433  0.988  0.987  BOL  0.423  0.423  0.998  0.998  0.423  0.423  0.968  0.968  BRA  0.465  0.465  1.000  1.000  0.465  0.465  0.986  0.986  BRB  0.448  0.448  1.000  1.000  0.461  0.449  0.990  0.990  BWA  0.467  0.467  0.999  0.999  0.467  0.467  0.964  0.964  CAF  0.407  0.407  0.991  0.991  0.407  0.407  0.905  0.904  CAN  0.577  0.344  1.000  1.000  0.576  0.491  0.997  0.997  CHE  0.723  0.341  1.000  1.000  0.571  0.524  0.998  0.998  CHL  0.510  0.510  1.000  1.000  0.595  0.585  0.995  0.995  CHN  0.467  0.467  1.000  1.000  0.480  0.471  0.991  0.991  CIV  0.421  0.421  0.994  0.994  0.421  0.421  0.928  0.928  CMR  0.423  0.423  0.994  0.994  0.423  0.423  0.939  0.939  COG  0.431  0.431  0.996  0.996  0.431  0.431  0.961  0.960  COL  0.425  0.425  0.999  0.999  0.425  0.425  0.987  0.987  COM  0.411  0.411  0.997  0.997  0.411  0.411  0.941  0.941  CPV  0.429  0.429  1.000  1.000  0.429  0.429  0.981  0.981  CRI  0.394  0.394  1.000  1.000  0.460  0.429  0.994  0.993  CYP  1.000  0.596  1.000  1.000  1.000  1.000  1.000  1.000  CZE  0.867  0.409  1.000  1.000  0.949  0.700  1.000  0.999  DEU  0.661  0.547  1.000  1.000  0.522  0.494  0.998  0.998  DJI  0.406  0.406  0.998  0.998  0.406  0.406  0.943  0.942  64 Table A.5 (continued)  Maternal Survival Rate  Infant Survival Rate  Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  Country  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  DMA  0.442  0.442  0.978  0.978  DNK  0.590  0.489  1.000  1.000  0.466  0.439  0.998  0.998  DOM  0.477  0.477  0.999  0.999  0.477  0.477  0.975  0.975  DZA  0.444  0.444  0.999  0.999  0.444  0.444  0.980  0.979  ECU  0.454  0.454  0.999  0.999  0.454  0.454  0.983  0.982  EGY  0.472  0.472  1.000  1.000  0.472  0.472  0.980  0.980  ERI  0.412  0.412  0.995  0.995  0.412  0.412  0.965  0.965  ESP  0.819  0.386  1.000  1.000  0.647  0.631  0.999  0.999  EST  0.774  0.462  1.000  1.000  0.774  0.745  0.999  0.999  ETH  0.408  0.408  0.996  0.997  0.408  0.408  0.953  0.953  FIN  0.799  0.375  1.000  1.000  0.869  0.801  1.000  1.000  FJI  0.445  0.445  1.000  1.000  0.445  0.445  0.983  0.982  FRA  0.520  0.466  1.000  1.000  0.519  0.491  0.998  0.998  GAB  0.521  0.521  0.997  0.997  0.521  0.521  0.963  0.963  GBR  0.578  0.345  1.000  1.000  0.577  0.518  0.998  0.998  GEO  0.460  0.460  1.000  1.000  0.460  0.460  0.990  0.989  GHA  0.418  0.418  0.997  0.997  0.419  0.419  0.955  0.955  GIN  0.409  0.409  0.993  0.993  0.409  0.409  0.936  0.936  GMB  0.408  0.408  0.993  0.993  0.408  0.408  0.956  0.956  GNB  0.412  0.412  0.995  0.995  0.412  0.412  0.935  0.935  GRC  0.871  0.408  1.000  1.000  0.684  0.657  0.999  0.999  GRD  0.466  0.466  1.000  1.000  0.466  0.466  0.989  0.989  GTM  0.446  0.446  0.999  0.999  0.446  0.446  0.975  0.975  GUY  0.425  0.425  0.998  0.998  0.425  0.425  0.973  0.973  HND  0.414  0.414  0.999  0.999  0.415  0.415  0.984  0.984  HRV  0.669  0.399  1.000  1.000  0.669  0.585  0.998  0.998  HTI  0.413  0.413  0.996  0.996  0.413  0.413  0.942  0.942  HUN  0.457  0.457  1.000  1.000  0.767  0.646  0.998  0.997  IDN  0.484  0.484  0.999  0.999  0.484  0.484  0.977  0.977  IND  0.440  0.440  0.998  0.998  0.440  0.440  0.961  0.960  IRN  0.509  0.509  1.000  1.000  0.509  0.509  0.987  0.987  ISL  0.791  0.371  1.000  1.000  1.000  1.000  1.000  1.000  ITA  0.791  0.373  1.000  1.000  0.625  0.600  0.999  0.999  JAM  0.444  0.444  0.999  0.999  0.444  0.444  0.988  0.987  JOR  0.411  0.411  0.999  0.999  0.411  0.411  0.985  0.985  KAZ  0.576  0.576  1.000  1.000  0.576  0.576  0.987  0.987  KEN  0.416  0.416  0.995  0.995  0.416  0.416  0.962  0.962  KGZ  0.413  0.413  0.999  0.999  0.413  0.413  0.979  0.979  KHM  0.423  0.423  0.998  0.998  0.423  0.423  0.969  0.969  65 Table A.5 (continued)  Maternal Survival Rate  Infant Survival Rate  Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  Country  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  KNA  0.689  0.628  0.994  0.994  KOR  0.941  0.561  1.000  1.000  0.941  0.896  0.999  0.999  LAO  0.439  0.439  0.998  0.998  0.439  0.439  0.947  0.947  LBN  0.485  0.485  1.000  1.000  0.566  0.539  0.994  0.994  LCA  0.444  0.444  1.000  1.000  0.444  0.444  0.990  0.989  LKA  0.478  0.478  1.000  1.000  0.558  0.507  0.993  0.993  LSO  0.396  0.396  0.995  0.995  0.396  0.396  0.927  0.927  LTU  0.794  0.473  1.000  1.000  0.793  0.697  0.998  0.998  LUX  0.725  0.433  1.000  1.000  1.000  0.955  1.000  1.000  LVA  0.498  0.498  1.000  1.000  0.835  0.659  0.997  0.997  MAR  0.448  0.448  0.999  0.999  0.448  0.448  0.976  0.975  MDA  0.408  0.408  1.000  1.000  0.409  0.409  0.988  0.988  MDG  0.410  0.410  0.996  0.997  0.410  0.410  0.962  0.962  MEX  0.490  0.490  1.000  1.000  0.490  0.490  0.989  0.988  MKD  0.740  0.441  1.000  1.000  0.515  0.457  0.993  0.993  MLI  0.413  0.413  0.994  0.994  0.413  0.413  0.927  0.927  MNG  0.463  0.463  1.000  1.000  0.463  0.463  0.983  0.983  MOZ  0.407  0.407  0.995  0.995  0.407  0.407  0.939  0.939  MRT  0.426  0.426  0.994  0.994  0.427  0.427  0.943  0.942  MUS  0.528  0.528  1.000  1.000  0.528  0.528  0.989  0.989  MWI  0.402  0.402  0.994  0.994  0.402  0.402  0.952  0.952  MYS  0.595  0.595  1.000  1.000  0.999  0.708  0.996  0.996  NAM  0.423  0.423  0.997  0.997  0.423  0.423  0.964  0.964  NER  0.407  0.407  0.994  0.995  0.407  0.407  0.944  0.944  NGA  0.446  0.446  0.992  0.992  0.446  0.446  0.926  0.926  NIC  0.415  0.415  0.999  0.999  0.415  0.415  0.983  0.983  NLD  0.470  0.457  1.000  1.000  0.469  0.438  0.998  0.998  NOR  0.648  0.576  1.000  1.000  0.709  0.591  0.999  0.999  NPL  0.413  0.413  0.997  0.997  0.414  0.414  0.968  0.968  NZL  0.513  0.344  1.000  1.000  0.512  0.430  0.997  0.997  OMN  0.972  0.972  1.000  1.000  1.000  1.000  1.000  1.000  PAK  0.438  0.438  0.998  0.998  0.439  0.439  0.932  0.932  PAN  0.430  0.430  0.999  0.999  0.430  0.430  0.986  0.986  PER  0.462  0.462  0.999  0.999  0.462  0.462  0.988  0.988  PHL  0.448  0.448  0.999  0.999  0.448  0.448  0.979  0.978  PNG  0.411  0.411  0.998  0.998  0.411  0.411  0.955  0.955  POL  1.000  0.469  1.000  1.000  0.786  0.671  0.998  0.998  PRY  0.429  0.429  0.999  0.999  0.429  0.429  0.983  0.983  ROM  0.462  0.462  1.000  1.000  0.540  0.494  0.994  0.993  66 Table A.5 (continued)  Maternal Survival Rate  Infant Survival Rate  Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  Country  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  RUS  0.515  0.515  1.000  1.000  0.601  0.580  0.995  0.995  RWA  0.410  0.410  0.997  0.997  0.410  0.410  0.964  0.964  SAU  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  SDN  0.426  0.426  0.997  0.997  0.426  0.426  0.953  0.953  SEN  0.415  0.415  0.997  0.997  0.415  0.415  0.963  0.962  SLB  0.404  0.404  0.999  0.999  0.405  0.405  0.979  0.979  SLE  0.410  0.410  0.986  0.986  0.411  0.411  0.904  0.904  SLV  0.425  0.425  0.999  0.999  0.425  0.425  0.987  0.987  SVK  0.911  0.430  1.000  1.000  0.721  0.578  0.997  0.997  SVN  0.658  0.392  1.000  1.000  0.910  0.818  1.000  1.000  SWE  0.604  0.572  1.000  1.000  0.661  0.535  0.999  0.999  SWZ  0.403  0.403  0.996  0.996  0.403  0.403  0.943  0.942  TCD  0.415  0.415  0.991  0.991  0.416  0.416  0.920  0.920  TGO  0.410  0.410  0.996  0.996  0.410  0.410  0.946  0.946  THA  0.480  0.480  1.000  1.000  0.493  0.483  0.991  0.990  TJK  0.417  0.417  1.000  1.000  0.418  0.418  0.960  0.960  TKM  0.512  0.512  1.000  1.000  0.512  0.512  0.954  0.953  TON  0.421  0.421  0.999  0.999  0.422  0.422  0.987  0.987  TTO  0.630  0.630  0.999  0.999  0.631  0.631  0.988  0.985  TUN  0.441  0.441  0.999  0.999  0.441  0.441  0.989  0.988  TUR  0.475  0.475  1.000  1.000  0.475  0.475  0.988  0.988  TZA  0.415  0.415  0.996  0.996  0.415  0.415  0.957  0.957  UGA  0.411  0.411  0.997  0.997  0.412  0.412  0.955  0.955  UKR  0.429  0.429  1.000  1.000  0.501  0.452  0.993  0.993  URY  0.422  0.422  1.000  1.000  0.493  0.458  0.994  0.993  USA  0.317  0.317  1.000  1.000  0.529  0.405  0.996  0.996  UZB  0.427  0.427  1.000  1.000  0.427  0.427  0.974  0.974  VCT  0.436  0.436  1.000  1.000  0.436  0.436  0.985  0.985  VEN  0.542  0.542  0.999  0.999  0.542  0.542  0.988  0.988  VNM  0.425  0.425  0.999  0.999  0.425  0.425  0.984  0.984  VUT  0.413  0.413  0.999  0.999  0.413  0.413  0.978  0.978  WSM  0.411  0.411  1.000  1.000  0.411  0.411  0.986  0.986  YEM  0.430  0.430  0.959  0.958  ZAF  0.445  0.445  0.999  0.999  0.445  0.445  0.965  0.965  ZAR  0.406  0.406  0.993  0.993  0.407  0.407  0.922  0.922  ZMB  0.422  0.422  0.998  0.998  0.422  0.422  0.952  0.951  ZWE  0.412  0.412  0.996  0.996  0.412  0.412  0.951  0.951  67 Table A.6 Efficiency Score for Selected Health Indicators (Two Inputs – Single Output)   Input (Health Expenditure‐ Adult Literacy Rate)    Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  0.924  0.780  0.824  0.814  0.931  0.765  0.704  0.704  ALB  0.927  0.832  0.993  0.963  0.923  0.754  0.997  0.997  ARG  0.914  0.783  0.956  0.953  0.791  0.642  0.941  0.941  ARM  0.899  0.744  0.970  0.938  0.909  0.693  0.954  0.951  AZE  0.899  0.833  0.916  0.905  0.909  0.843  0.900  0.899  BDI  0.922  0.788  0.840  0.802  0.893  0.830  0.982  0.965  BEN  1.000  0.926  1.000  0.974  1.000  0.921  1.000  0.892  BFA  0.991  0.915  0.988  0.922  1.000  1.000  1.000  1.000  BGD  1.000  0.957  1.000  0.983  1.000  1.000  1.000  1.000  BGR  0.851  0.693  0.929  0.926  0.749  0.613  0.941  0.941  BHR  0.948  0.862  0.984  0.969  0.959  0.871  1.000  0.998  BLR  0.798  0.639  0.898  0.894  0.791  0.647  0.987  0.987  BOL  0.873  0.703  0.909  0.907  0.796  0.665  0.957  0.957  BRA  0.922  0.747  0.939  0.938  0.815  0.679  0.978  0.978  CAF  0.970  0.882  0.768  0.726  0.987  0.865  0.488  0.464  CHL  1.000  1.000  1.000  1.000  0.800  0.649  0.941  0.941  CHN  0.915  0.758  0.976  0.943  0.969  0.969  1.000  1.000  CIV  1.000  0.885  1.000  0.842  1.000  0.874  1.000  0.840  CMR  0.936  0.768  0.782  0.765  0.947  0.743  0.853  0.853  COG  0.914  0.732  0.815  0.782  0.918  0.695  0.828  0.828  COL  0.859  0.823  0.971  0.963  0.685  0.574  0.906  0.906  COM  0.947  0.862  0.968  0.948  0.957  0.847  0.881  0.865  CPV  0.893  0.752  0.974  0.925  0.856  0.708  0.957  0.957  CRI  1.000  0.983  1.000  0.998  0.548  0.489  0.906  0.906  CYP  1.000  1.000  1.000  1.000  0.854  0.722  0.997  0.997  DOM  0.992  0.811  0.976  0.951  0.947  0.696  0.865  0.863  DZA  1.000  1.000  1.000  1.000  ECU  1.000  0.898  1.000  0.975  0.841  0.651  0.877  0.877  EGY  1.000  0.862  1.000  0.946  1.000  0.854  1.000  0.963  ERI  0.904  0.771  0.843  0.836  ESP  1.000  1.000  1.000  1.000  0.428  0.425  0.977  0.977  EST  0.768  0.697  0.947  0.938  0.601  0.528  0.947  0.947  ETH  0.987  0.868  0.987  0.923  GAB  1.000  0.829  1.000  0.830  1.000  0.787  1.000  0.764  GEO  0.900  0.732  0.958  0.928  0.910  0.694  0.932  0.930  GHA  0.882  0.733  0.857  0.835  0.898  0.763  0.934  0.934  68 Table A.6 (continued)    Input (Health Expenditure‐ Adult Literacy Rate)    Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  GIN  1.000  0.934  1.000  0.927  1.000  0.918  1.000  0.710  GMB  1.000  0.916  1.000  0.978  1.000  1.000  1.000  1.000  GNB  0.973  0.852  0.825  0.821  1.000  0.884  1.000  0.878  GRC  1.000  0.967  1.000  0.993  0.672  0.672  1.000  1.000  GTM  0.976  0.799  0.969  0.931  0.934  0.724  0.869  0.869  GUY  0.838  0.663  0.863  0.847  0.826  0.703  0.973  0.973  HND  0.866  0.730  0.964  0.915  0.824  0.706  0.980  0.980  HRV  0.745  0.739  0.951  0.943  0.483  0.479  0.967  0.967  HTI  0.964  0.840  0.896  0.879  IDN  0.958  0.769  0.919  0.898  0.965  0.755  0.817  0.815  IND  1.000  0.814  1.000  0.895  1.000  0.782  1.000  0.826  IRN  1.000  0.847  1.000  0.962  1.000  0.849  1.000  0.997  ITA  0.991  0.982  0.998  0.998  0.424  0.408  0.965  0.965  JOR  0.839  0.731  0.950  0.940  0.673  0.637  0.991  0.991  KAZ  0.899  0.755  0.895  0.875  0.909  0.784  0.994  0.992  KEN  0.904  0.731  0.850  0.821  0.890  0.722  0.918  0.918  KGZ  0.850  0.642  0.866  0.865  0.850  0.656  0.971  0.971  KHM  0.945  0.747  0.899  0.869  0.942  0.750  0.890  0.890  LAO  0.990  0.848  0.974  0.887  1.000  0.836  1.000  0.819  LBN  1.000  1.000  1.000  1.000  0.878  0.667  0.818  0.818  LKA  0.987  0.872  0.992  0.969  1.000  0.834  1.000  0.998  LSO  0.784  0.667  0.611  0.608  0.777  0.685  0.944  0.944  LTU  0.671  0.629  0.919  0.911  0.643  0.556  0.948  0.948  LVA  0.788  0.660  0.915  0.911  0.751  0.620  0.938  0.938  MAR  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  MDA  0.788  0.632  0.887  0.885  0.743  0.594  0.905  0.905  MDG  0.889  0.750  0.834  0.819  0.863  0.722  0.730  0.730  MEX  0.959  0.832  0.966  0.961  0.864  0.677  0.928  0.928  MLI  0.995  0.915  0.956  0.931  1.000  0.899  1.000  0.794  MNG  0.912  0.684  0.866  0.836  0.923  0.721  0.991  0.988  MOZ  0.933  0.813  0.826  0.803  0.931  0.788  0.800  0.785  MRT  1.000  0.991  1.000  0.996  MUS  0.979  0.804  0.962  0.941  0.990  0.803  0.994  0.991  MWI  0.843  0.745  0.736  0.734  0.867  0.787  0.953  0.935  MYS  0.963  0.849  0.978  0.961  0.974  0.838  0.984  0.982  NAM  0.779  0.647  0.763  0.753  0.736  0.595  0.867  0.867  69 Table A.6 (continued)  Health Expenditure‐ Adult Literacy Rate                Disability Adjusted Life Expectancy  Immunization, DTP    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  NER  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  NGA  1.000  0.893  1.000  0.881  NPL  0.945  0.848  0.974  0.940  0.927  0.815  0.936  0.910  OMN  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  PAK  1.000  0.867  1.000  0.925  1.000  0.852  1.000  0.754  PAN  0.961  0.888  0.990  0.978  0.616  0.503  0.843  0.843  PER  1.000  0.978  1.000  0.997  0.874  0.677  0.921  0.921  PHL  0.935  0.710  0.904  0.877  0.940  0.681  0.806  0.804  PRY  0.888  0.741  0.940  0.936  0.781  0.615  0.889  0.889  ROM  0.848  0.672  0.931  0.925  0.697  0.576  0.929  0.929  RUS  0.777  0.607  0.865  0.863  0.790  0.640  0.981  0.981  RWA  0.881  0.749  0.864  0.850  0.957  0.841  0.986  0.986  SAU  0.949  0.873  0.988  0.974  0.960  0.864  0.994  0.993  SDN  0.985  0.860  0.978  0.939  SEN  0.976  0.842  0.941  0.932  0.977  0.896  0.936  0.920  SLE  0.978  0.910  0.901  0.870  1.000  1.000  1.000  1.000  SLV  0.969  0.794  0.970  0.952  0.794  0.655  0.921  0.921  SWZ  0.707  0.605  0.653  0.642  0.692  0.615  0.937  0.937  TCD  1.000  0.976  1.000  0.927  1.000  0.959  1.000  0.474  TGO  0.885  0.783  0.848  0.819  0.925  0.772  0.878  0.853  THA  0.950  0.837  0.968  0.963  0.940  0.940  1.000  1.000  TON  0.824  0.630  0.880  0.878  0.783  0.602  0.815  0.815  TUN  1.000  0.977  1.000  0.995  0.807  0.759  0.990  0.990  TUR  0.974  0.910  0.990  0.983  0.715  0.628  0.978  0.978  TZA  0.885  0.728  0.816  0.789  0.893  0.750  0.929  0.929  UGA  0.887  0.739  0.786  0.765  0.870  0.709  0.798  0.798  UKR  0.800  0.620  0.879  0.877  0.747  0.585  0.535  0.535  URY  0.802  0.756  0.963  0.952  0.580  0.528  0.958  0.958  UZB  0.889  0.657  0.873  0.864  0.900  0.728  0.999  0.999  VEN  0.935  0.825  0.983  0.962  0.948  0.768  0.823  0.821  VNM  0.901  0.737  0.932  0.926  0.811  0.652  0.912  0.912  WSM  0.815  0.648  0.899  0.897  0.742  0.568  0.629  0.629  ZAF  0.788  0.629  0.718  0.715  0.760  0.589  0.732  0.732  ZAR  0.883  0.739  0.796  0.776  0.861  0.693  0.745  0.745  ZMB  0.902  0.745  0.714  0.695  0.887  0.678  0.853  0.853  ZWE  0.885  0.696  0.705  0.670  0.855  0.678  0.899  0.899  70 Table A.7 Efficiency Score for Selected Infrastructure Indicators (Single Inputs – Output)   Quality of the Overall Infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  0.100  0.100  0.321  0.321  0.100  0.100  0.233  0.233  ALB  0.143  0.143  0.522  0.522  0.143  0.143  0.555  0.555  ARE  0.259  0.254  0.936  0.936  0.259  0.243  0.956  0.956  ARG  0.230  0.230  0.507  0.484  0.230  0.230  0.502  0.491  ARM  0.258  0.258  0.614  0.594  0.258  0.258  0.700  0.687  AUS  0.334  0.193  0.772  0.772  0.364  0.216  0.905  0.905  AUT  1.000  0.614  1.000  0.948  1.000  0.835  1.000  0.987  AZE  0.087  0.087  0.662  0.662  0.087  0.087  0.650  0.650  BDI  0.153  0.153  0.369  0.369  0.153  0.153  0.329  0.329  BEL  0.877  0.699  0.922  0.911  0.877  0.870  0.999  0.992  BEN  0.168  0.168  0.427  0.427  0.168  0.168  0.371  0.371  BFA  0.150  0.150  0.392  0.392  0.150  0.150  0.357  0.357  BGD  0.177  0.177  0.398  0.398  0.177  0.177  0.290  0.290  BGR  0.165  0.165  0.471  0.471  0.165  0.165  0.573  0.573  BHR  0.383  0.258  0.824  0.824  0.139  0.139  0.848  0.848  BIH  0.151  0.151  0.361  0.361  0.151  0.151  0.783  0.783  BLZ  0.163  0.163  0.530  0.530  0.163  0.163  0.603  0.603  BOL  0.104  0.104  0.448  0.448  0.104  0.104  0.582  0.582  BRA  0.664  0.679  0.620  0.664  0.804  0.779  BRB  0.499  0.369  0.848  0.848  0.499  0.325  0.913  0.913  BTN  0.098  0.098  0.676  0.676  0.098  0.098  0.863  0.863  BWA  0.102  0.102  0.650  0.650  0.102  0.102  0.546  0.546  CAN  0.439  0.361  0.878  0.878  0.656  0.440  0.965  0.965  CHE  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  CHL  1.000  0.592  1.000  0.851  0.395  0.395  0.939  0.857  CIV  0.228  0.228  0.612  0.584  0.228  0.228  0.600  0.587  CMR  0.210  0.210  0.458  0.432  0.210  0.210  0.413  0.400  COL  0.133  0.133  0.497  0.497  0.133  0.133  0.747  0.747  CPV  0.109  0.109  0.542  0.542  0.109  0.109  0.324  0.324  CRI  0.242  0.242  0.533  0.512  0.242  0.242  0.847  0.831  CZE  0.411  0.195  0.737  0.737  0.448  0.361  0.936  0.936  DEU  1.000  1.000  1.000  1.000  1.000  1.000  1.000  1.000  DOM  0.217  0.217  0.554  0.525  0.217  0.217  0.295  0.287  DZA  0.084  0.084  0.516  0.516  0.084  0.084  0.647  0.647  ECU  0.098  0.098  0.531  0.531  0.098  0.098  0.556  0.556  EGY  0.377  0.377  0.716  0.599  0.377  0.377  0.741  0.672  EST  0.360  0.210  0.773  0.773  0.142  0.142  0.814  0.814  ETH  0.094  0.094  0.481  0.481  0.094  0.094  0.472  0.472  71 Table A.7 (continued)    Quality of the Overall Infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  FIN  0.802  0.584  0.963  0.963  0.802  0.753  0.995  0.995  FRA  0.795  0.545  0.958  0.958  0.791  0.64  0.984  0.984  GAB  0.121  0.121  0.459  0.459  0.121  0.121  0.354  0.354  GBR  0.669  0.42  0.856  0.823  1  0.742  1  0.982  GEO  0.136  0.136  0.619  0.619  0.136  0.136  0.723  0.723  GHA  0.162  0.162  0.536  0.536  0.162  0.162  0.438  0.438  GIN  0.57  0.442  0.388  0.57  0.228  0.218  GMB  0.156  0.156  0.638  0.638  0.156  0.156  0.59  0.59  GRC  0.41  0.41  0.939  0.741  0.41  0.41  0.858  0.788  GTM  0.602  0.9  0.801  0.602  0.862  0.828  GUY  0.104  0.104  0.527  0.527  0.104  0.104  0.421  0.421  HND  0.291  0.291  0.571  0.559  0.291  0.291  0.592  0.585  HRV  0.452  0.182  0.713  0.713  0.179  0.179  0.81  0.81  HTI  0.109  0.109  0.306  0.306  0.109  0.109  0.252  0.252  IDN  0.332  0.332  0.563  0.559  0.332  0.332  0.6  0.598  IND  0.136  0.136  0.535  0.535  0.136  0.136  0.481  0.481  IRL  0.235  0.235  0.723  0.692  0.648  0.455  0.949  0.929  IRN  0.28  0.28  0.642  0.627  0.28  0.28  0.765  0.754  ISL  0.559  0.545  0.934  0.934  0.973  0.959  0.999  0.999  ISR  1  1  1  1  1  1  ITA  0.26  0.26  0.639  0.619  0.26  0.26  0.861  0.845  JOR  0.434  0.216  0.745  0.745  0.172  0.172  0.836  0.836  JPN  0.367  0.32  0.896  0.896  0.367  0.332  0.951  0.951  KAZ  0.162  0.162  0.596  0.596  0.162  0.162  0.648  0.648  KEN  0.187  0.187  0.556  0.556  0.187  0.187  0.535  0.535  KHM  0.135  0.135  0.536  0.536  0.135  0.135  0.445  0.445  KOR  0.404  0.307  0.855  0.855  0.404  0.199  0.889  0.889  KWT  0.15  0.15  0.687  0.687  0.15  0.15  0.735  0.735  LKA  0.18  0.18  0.669  0.669  0.18  0.18  0.702  0.702  LSO  0.1  0.1  0.466  0.466  0.1  0.1  0.532  0.532  LTU  0.448  0.204  0.731  0.731  0.177  0.177  0.816  0.816  LUX  0.418  0.357  0.889  0.889  0.418  0.361  0.945  0.945  LVA  0.197  0.197  0.676  0.676  0.197  0.197  0.778  0.778  MAR  0.198  0.198  0.629  0.629  0.198  0.198  0.767  0.767  MDA  0.223  0.223  0.541  0.515  0.223  0.223  0.648  0.632  MDG  0.294  0.294  0.461  0.452  0.294  0.294  0.342  0.338  MEX  0.169  0.169  0.596  0.596  0.169  0.169  0.636  0.636  MLI  0.161  0.161  0.493  0.493  0.161  0.161  0.501  0.501  MMR  0.14  0.14  0.342  0.342  0.14  0.14  0.418  0.418  72 Table A.7 (continued)    Quality of the Overall Infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  MNE  0.145  0.145  0.489  0.489  0.145  0.145  0.572  0.572  MOZ  0.103  0.103  0.418  0.418  0.103  0.103  0.489  0.489  MUS  0.155  0.155  0.69  0.69  0.155  0.155  0.772  0.772  MWI  0.203  0.203  0.458  0.458  0.203  0.203  0.363  0.363  MYS  0.274  0.192  0.834  0.834  0.099  0.099  0.855  0.855  NAM  0.324  0.189  0.774  0.774  0.128  0.128  0.795  0.795  NGA  0.254  0.254  0.423  0.409  0.254  0.254  0.237  0.232  NIC  0.175  0.175  0.45  0.45  0.175  0.175  0.466  0.466  NLD  0.457  0.407  0.903  0.903  0.794  0.706  0.991  0.991  NPL  0.53  0.527  0.453  0.53  0.271  0.257  NZL  0.317  0.132  0.717  0.717  0.125  0.125  0.834  0.834  OMN  0.213  0.136  0.812  0.812  0.213  0.116  0.898  0.898  PAK  0.344  0.344  0.534  0.531  0.344  0.344  0.348  0.347  PAN  0.137  0.137  0.688  0.688  0.137  0.137  0.764  0.764  PER  0.165  0.165  0.477  0.477  0.165  0.165  0.708  0.708  PHL  0.403  0.403  0.692  0.543  0.403  0.403  0.664  0.608  POL  0.158  0.158  0.518  0.518  0.158  0.158  0.782  0.782  PRT  0.656  0.53  0.926  0.888  0.656  0.424  0.94  0.922  PRY  0.217  0.217  0.373  0.354  0.217  0.217  0.459  0.447  ROM  0.137  0.137  0.424  0.424  0.137  0.137  0.633  0.633  RUS  0.205  0.205  0.539  0.539  0.205  0.205  0.667  0.667  RWA  0.113  0.113  0.676  0.676  0.113  0.113  0.604  0.604  SAU  0.247  0.137  0.78  0.78  0.247  0.131  0.895  0.895  SEN  0.16  0.16  0.524  0.524  0.16  0.16  0.352  0.352  SGP  0.719  0.607  0.979  0.979  0.719  0.619  0.989  0.989  SLE  0.197  0.197  0.425  0.425  0.197  0.197  0.324  0.324  SLV  0.622  0.955  0.858  0.622  0.83  0.8  SRB  0.286  0.286  0.469  0.459  0.286  0.286  0.711  0.701  SUR  0.121  0.121  0.623  0.623  0.121  0.121  0.552  0.552  SVK  0.194  0.194  0.62  0.62  0.536  0.312  0.903  0.903  SWZ  0.163  0.163  0.626  0.626  0.163  0.163  0.58  0.58  SYC  0.14  0.14  0.699  0.699  0.14  0.14  0.73  0.73  TCD  0.127  0.127  0.337  0.337  0.127  0.127  0.22  0.22  THA  0.142  0.142  0.696  0.696  0.142  0.142  0.8  0.8  TJK  0.108  0.108  0.518  0.518  0.108  0.108  0.344  0.344  TTO  0.096  0.096  0.651  0.651  0.096  0.096  0.781  0.781  TUN  0.154  0.154  0.698  0.698  0.154  0.154  0.812  0.812  TUR  0.149  0.149  0.703  0.703  0.149  0.149  0.667  0.667  73  Table A.7 (continued)    Quality of the Overall Infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  TZA  0.186  0.186  0.441  0.441  0.186  0.186  0.345  0.345  UGA  0.152  0.152  0.482  0.482  0.152  0.152  0.364  0.364  UKR  0.504  0.817  0.69  0.504  0.763  0.718  URY  0.173  0.173  0.597  0.597  0.173  0.173  0.841  0.841  USA  0.449  0.361  0.872  0.872  0.449  0.327  0.925  0.925  VEN  0.093  0.093  0.411  0.411  0.093  0.093  0.364  0.364  VNM  0.12  0.12  0.465  0.465  0.12  0.12  0.53  0.53  YEM  0.398  0.398  0.575  0.449  0.398  0.398  0.248  0.226  ZAF  0.284  0.284  0.718  0.702  0.284  0.284  0.552  0.544  ZMB  0.346  0.346  0.516  0.514  0.346  0.346  0.516  0.515  ZWE  0.751  0.665  0.626  0.751  0.321  0.314  74 Table A.8 Efficiency Score for Selected Infrastructure Indicators (Two Inputs – Single Output) Input (Public ‐ Private Capital Expenditure)    Quality of overall infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  AGO  1  1  1  1  1  1  1  1  ALB  0.625  0.385  0.522  0.522  0.625  0.425  0.555  0.555  ARE  1  1  1  1  1  0.954  1  0.978  ARG  0.964  0.519  0.593  0.525  0.964  0.574  0.502  0.496  ARM  0.684  0.494  0.614  0.594  0.684  0.559  0.7  0.687  AUS  0.729  0.601  0.772  0.772  0.729  0.652  0.905  0.905  AUT  1  0.897  1  0.955  1  0.87  1  0.987  AZE  1  0.828  1  0.9  1  0.787  1  0.849  BDI  0.822  0.499  0.904  0.501  1  0.553  1  0.426  BEL  0.925  0.846  0.922  0.911  0.925  0.911  0.999  0.992  BEN  0.714  0.341  0.427  0.427  0.714  0.356  0.371  0.371  BFA  0.543  0.385  0.441  0.419  0.689  0.424  0.367  0.358  BGD  0.446  0.305  0.398  0.398  0.66  0.291  0.29  0.29  BGR  0.723  0.39  0.471  0.471  0.723  0.489  0.573  0.573  BHR  0.766  0.651  0.824  0.824  0.738  0.626  0.848  0.848  BIH  0.513  0.339  0.387  0.379  0.832  0.706  0.784  0.783  BLZ  0.714  0.5  0.568  0.553  0.895  0.58  0.603  0.603  BOL  0.894  0.484  0.715  0.552  0.894  0.643  0.999  0.675  BRA  0.995  0.799  0.679  0.62  0.995  0.808  0.804  0.779  BRB  1  1  1  1  1  1  1  1  BTN  0.528  0.379  0.676  0.676  0.528  0.453  0.863  0.863  BWA  0.608  0.458  0.65  0.65  0.503  0.404  0.546  0.546  CAN  0.833  0.765  0.912  0.884  0.801  0.781  0.967  0.966  CHE  1  1  1  1  1  1  1  1  CHL  1  0.798  1  0.851  0.903  0.776  0.939  0.857  CIV  1  0.829  1  0.831  1  0.864  1  0.822  CMR  0.864  0.417  0.495  0.446  0.864  0.452  0.413  0.401  COL  0.584  0.411  0.516  0.501  0.736  0.601  0.748  0.747  CPV  0.479  0.341  0.542  0.542  0.479  0.255  0.324  0.324  CRI  0.825  0.477  0.533  0.512  0.933  0.736  0.847  0.831  CZE  0.721  0.594  0.737  0.737  0.721  0.69  0.936  0.936  DEU  1  1  1  1  1  1  1  1  DOM  0.753  0.446  0.554  0.525  0.753  0.351  0.295  0.287  DZA  0.542  0.368  0.516  0.516  0.655  0.461  0.647  0.647  ECU  0.627  0.439  0.569  0.549  0.627  0.477  0.556  0.556  EGY  0.864  0.632  0.782  0.599  0.864  0.696  0.741  0.672  75 Table A.8 (continued)  Input (Public ‐ Private Capital Expenditure)    Quality of overall infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency  dmu  FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  EST  0.73  0.609  0.773  0.773  0.73  0.605  0.814  0.814  ETH  0.605  0.382  0.514  0.492  0.605  0.41  0.473  0.472  FIN  1  0.94  1  0.978  0.858  0.851  0.996  0.995  FRA  0.992  0.917  0.994  0.97  0.851  0.826  0.985  0.984  GAB  0.531  0.338  0.459  0.459  0.531  0.317  0.354  0.354  GBR  1  0.92  1  0.92  1  1  1  1  GEO  0.752  0.525  0.657  0.629  0.752  0.599  0.724  0.724  GHA  0.709  0.537  0.632  0.576  0.709  0.495  0.451  0.442  GIN  1  1  1  1  1  1  1  0.59  GMB  0.839  0.569  0.683  0.651  0.839  0.538  0.591  0.59  GRC  1  0.839  1  0.815  1  0.896  1  0.868  GTM  0.968  0.88  0.942  0.86  1  0.916  1  0.892  GUY  0.976  0.603  0.912  0.685  0.976  0.544  0.722  0.521  HND  0.711  0.498  0.571  0.559  0.711  0.523  0.592  0.585  HRV  0.795  0.634  0.753  0.717  0.795  0.677  0.811  0.81  HTI  0.24  0.23  0.306  0.306  0.516  0.258  0.252  0.252  IDN  0.652  0.483  0.563  0.559  0.652  0.5  0.6  0.598  IND  0.598  0.37  0.535  0.535  0.598  0.36  0.481  0.481  IRL  0.969  0.675  0.781  0.707  0.969  0.845  0.949  0.93  IRN  0.747  0.557  0.642  0.627  0.747  0.651  0.765  0.754  ISL  1  1  1  1  1  1  1  1  ISR  1  1  1  1  1  1  1  1  ITA  0.92  0.622  0.887  0.636  1  0.819  1  0.846  JOR  0.749  0.627  0.745  0.745  0.749  0.657  0.836  0.836  JPN  0.861  0.79  0.951  0.911  0.829  0.778  0.952  0.952  KAZ  0.765  0.501  0.596  0.596  0.765  0.542  0.648  0.648  KEN  0.799  0.499  0.589  0.559  0.799  0.505  0.536  0.535  KHM  0.619  0.472  0.574  0.552  0.619  0.439  0.445  0.445  KOR  0.722  0.647  0.855  0.855  0.684  0.621  0.889  0.889  KWT  0.891  0.711  0.81  0.76  0.891  0.739  0.805  0.763  LKA  0.724  0.526  0.669  0.669  0.715  0.542  0.702  0.702  LSO  0.619  0.383  0.499  0.48  0.619  0.462  0.533  0.533  LTU  0.856  0.685  0.783  0.752  0.856  0.72  0.817  0.816  LUX  1  0.914  1  0.951  0.973  0.9  0.974  0.951  LVA  0.781  0.575  0.676  0.676  0.81  0.634  0.778  0.778  76 Table A.8 (continued)  Input (Public ‐ Private Capital Expenditure)    Quality of overall infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  MAR  0.651  0.466  0.629  0.629  0.736  0.556  0.767  0.767  MDA  0.714  0.43  0.541  0.515  0.714  0.529  0.648  0.632  MDG  0.749  0.443  0.461  0.452  0.749  0.443  0.342  0.338  MEX  0.819  0.533  0.633  0.604  0.819  0.568  0.637  0.636  MLI  0.705  0.459  0.528  0.512  0.705  0.502  0.502  0.501  MMR  0.55  0.334  0.404  0.367  0.698  0.466  0.431  0.423  MNE  0.634  0.402  0.489  0.489  0.634  0.487  0.572  0.572  MOZ  0.789  0.414  0.493  0.477  0.789  0.521  0.535  0.523  MUS  0.796  0.592  0.716  0.693  0.796  0.632  0.773  0.772  MWI  0.888  0.523  0.794  0.529  0.888  0.497  0.622  0.385  MYS  0.781  0.759  0.893  0.869  0.781  0.715  0.856  0.855  NAM  0.707  0.599  0.774  0.774  0.707  0.586  0.795  0.795  NGA  0.639  0.417  0.423  0.411  0.863  0.409  0.237  0.232  NIC  0.719  0.354  0.45  0.45  0.719  0.408  0.466  0.466  NLD  0.93  0.872  0.967  0.942  0.93  0.915  0.992  0.992  NPL  0.762  0.649  0.527  0.453  0.762  0.649  0.271  0.257  NZL  0.778  0.642  0.768  0.748  0.778  0.705  0.835  0.835  OMN  0.802  0.746  0.868  0.852  0.802  0.784  0.899  0.898  PAK  0.993  0.618  0.754  0.584  0.993  0.535  0.45  0.37  PAN  0.62  0.474  0.688  0.688  0.62  0.502  0.764  0.764  PER  0.722  0.389  0.477  0.477  0.734  0.562  0.708  0.708  PHL  0.797  0.564  0.692  0.543  0.797  0.609  0.664  0.608  POL  0.694  0.471  0.555  0.533  0.861  0.682  0.783  0.782  PRT  1  0.907  1  0.923  0.98  0.874  0.94  0.923  PRY  0.554  0.387  0.436  0.39  0.953  0.541  0.459  0.454  ROM  0.601  0.321  0.424  0.424  0.713  0.485  0.633  0.633  RUS  0.843  0.509  0.571  0.55  0.843  0.625  0.688  0.668  RWA  0.743  0.571  0.724  0.696  0.743  0.519  0.605  0.605  SAU  0.82  0.715  0.834  0.823  0.82  0.799  0.896  0.896  SEN  0.701  0.411  0.524  0.524  0.701  0.336  0.352  0.352  SGP  0.943  0.881  0.979  0.979  0.768  0.752  0.989  0.989  SLE  0.863  0.454  0.527  0.463  0.863  0.433  0.334  0.33  SLV  1  0.966  1  0.96  1  0.93  1  0.91  SRB  0.845  0.47  0.469  0.459  0.845  0.681  0.711  0.701  SUR  0.522  0.365  0.623  0.623  0.522  0.336  0.552  0.552  SVK  0.801  0.554  0.62  0.62  0.801  0.744  0.903  0.903  SWZ  1  0.818  1  0.818  0.975  0.78  0.996  0.72  SYC  0.677  0.518  0.699  0.699  0.677  0.522  0.73  0.73  77 Table A.8 (continued)  Input (Public ‐ Private Capital Expenditure)    Quality of overall infrastructure  Quality of Electricity Supply    Input Efficiency  Output Efficiency  Input Efficiency  Output Efficiency    FDH  DEA  FDH  DEA  FDH  DEA  FDH  DEA  TCD  0.389  0.245  0.337  0.337  0.558  0.24  0.22  0.22  THA  0.738  0.552  0.696  0.696  0.738  0.602  0.8  0.8  TJK  0.917  0.578  0.828  0.65  0.917  0.467  0.59  0.407  TUN  0.788  0.593  0.724  0.699  0.788  0.651  0.812  0.812  TUR  0.726  0.558  0.703  0.703  0.726  0.522  0.667  0.667  TZA  0.642  0.332  0.441  0.441  0.642  0.313  0.345  0.345  UGA  0.655  0.353  0.482  0.482  0.655  0.321  0.364  0.364  UKR  0.917  0.749  0.855  0.712  0.917  0.786  0.885  0.733  URY  0.85  0.552  0.64  0.613  0.85  0.73  0.842  0.841  USA  0.945  0.838  0.933  0.914  0.909  0.818  0.926  0.925  VEN  1  0.637  1  0.736  1  0.662  1  0.676  VNM  0.524  0.341  0.465  0.465  0.524  0.412  0.53  0.53  YEM  1  0.805  1  0.809  1  0.681  1  0.449  ZAF  0.852  0.661  0.718  0.702  0.843  0.565  0.552  0.544  ZMB  0.681  0.483  0.516  0.514  0.681  0.483  0.516  0.515  ZWE  1  0.891  1  0.674  1  0.891  1  0.342  78 Table A.9. List of Countries Code  Region  Country  Code  Region  Country  Code  Region  Country  AGO  AFR  Angola  GRD  LAC  Grenada  PER  LAC  Peru  ALB  ECA  Albania  GTM  LAC  Guatemala  PHL  EAP  Philippines  ARE  MNA  United Arab Emirates  GUY  LAC  Guyana  PNG  EAP  Papua New Guinea  ARG  LAC  Argentina  HND  LAC  Honduras  POL  ECA  Poland  ARM  ECA  Armenia  HRV  ECA  Croatia  PRI  LAC  Puerto Rico  ATG  LAC  Antigua and Barbuda  HTI  LAC  Haiti  PRY  LAC  Paraguay  AZE  ECA  Azerbaijan  HUN  ECA  Hungary  QAT  MNA  Qatar  BDI  AFR  Burundi  IDN  EAP  Indonesia  ROM  ECA  Romania  BEN  AFR  Benin  IND  SAS  India  RUS  ECA  Russian Federation  BFA  AFR  Burkina Faso  IRN  MNA  Iran, Islamic Rep.  RWA  AFR  Rwanda  BGD  SAS  Bangladesh  JAM  LAC  Jamaica  SAU  MNA  Saudi Arabia  BGR  ECA  Bulgaria  JOR  MNA  Jordan  SDN  AFR  Sudan  BHR  MNA  Bahrain  KAZ  ECA  Kazakhstan  SEN  AFR  Senegal  BHS  LAC  Bahamas, The  KEN  AFR  Kenya  SGP  EAP  Singapore  Bosnia &  BIH  ECA  KGZ  ECA  Kyrgyz Republic  SLB  EAP  Solomon Islands  Herzegovina  BLR  ECA  Belarus  KHM  EAP  Cambodia  SLE  AFR  Sierra Leone  BLZ  LAC  Belize  KNA  LAC  St. Kitts and Nevis  SLV  LAC  El Salvador  BOL  LAC  Bolivia  KOR  EAP  Korea, Rep.  SRB  ECA  Serbia  BRA  LAC  Brazil  KWT  MNA  Kuwait  SUR  LAC  Suriname  BRB  LAC  Barbados  LAO  EAP  Lao PDR  SVK  ECA  Slovak Republic  BRN  EAP  Brunei Darussalam  LBN  MNA  Lebanon  SVN  ECA  Slovenia  BTN  SAS  Bhutan  LCA  LAC  St. Lucia  SWZ  AFR  Eswatini  BWA  AFR  Botswana  LKA  SAS  Sri Lanka  SYC  AFR  Seychelles  CAF  AFR  Central African Rep.  LSO  AFR  Lesotho  SYR  MNA  Syrian Arab Republic  CHL  LAC  Chile  LBR  AFR  Liberia  TCD  AFR  Chad  CHN  EAP  China  LBY  MNA  Lybia  TGO  AFR  Togo  CIV  AFR  Côte d'Ivoire  LTU  ECA  Lithuania  THA  EAP  Thailand  CMR  AFR  Cameroon  LVA  ECA  Latvia  TJK  ECA  Tajikistan  COG  AFR  Congo, Rep.  MAR  MNA  Morocco  TKM  ECA  Turkmenistan  COL  LAC  Colombia  MDA  ECA  Moldova  TLS  EAP  Timor‐Leste  COM  AFR  Comoros  MDG  AFR  Madagascar  TON  EAP  Tonga  CPV  AFR  Cabo Verde  MEX  LAC  Mexico  TTO  LAC  Trinidad and Tobago  CRI  LAC  Costa Rica  MKD  ECA  Macedonia, FYR  TUN  MNA  Tunisia  CZE  ECA  Czech Republic  MLI  AFR  Mali  TUR  ECA  Turkey  79 Table A.9. Continued  Code  Region  Country  Code  Region  Country  Code  Region  Country  DJI  MNA  Djibouti  MMR  EAP  Myanmar  TZA  AFR  Tanzania  DMA  LAC  Dominica  MNE  EAP  Montenegro, Rep. of  UGA  AFR  Uganda  Dominican  DOM  LAC  MNG  EAP  Mongolia  UKR  ECA  Ukraine  Republic  DZA  MNA  Algeria  MOZ  AFR  Mozambique  URY  LAC  Uruguay  ECU  LAC  Ecuador  MRT  AFR  Mauritania  UZB  ECA  Uzbekistan  St. Vincent &  EGY  MNA  Egypt, Arab Rep.  MUS  AFR  Mauritius  VCT  LAC  Grenadines  ERI  AFR  Eritrea  MWI  AFR  Malawi  VEN  LAC  Venezuela, RB  EST  ECA  Estonia  MYS  EAP  Malaysia  VNM  EAP  Vietnam  ETH  AFR  Ethiopia  NAM  AFR  Namibia  VUT  EAP  Vanuatu  FJI  EAP  Fiji  NER  AFR  Niger  WSM  EAP  Samoa  GAB  AFR  Gabon  NGA  AFR  Nigeria  YEM  MNA  Yemen, Rep.  GEO  ECA  Georgia  NIC  LAC  Nicaragua  ZAF  AFR  South Africa  GHA  AFR  Ghana  NPL  SAS  Nepal  ZAR  AFR  Congo, Dem. Rep.  GIN  AFR  Guinea  OMN  MNA  Oman  ZMB  AFR  Zambia  GMB  AFR  Gambia, The  PAK  SAS  Pakistan  ZWE  AFR  Zimbabwe  GNB  AFR  Guinea‐Bissau  PAN  LAC  Panama  Table A.9. Continued  Developed countries included in the efficiency estimation for learning scores  Code  Country  Code  Country  Code  Country  AUS  Australia  FIN  Finland  LUX  Luxembourg  AUT  Austria  FRA  France  MLT  Malta  BEL  Belgium  GBR  United Kingdom  NLD  Netherlands  CAN  Canada  GRC  Greece  NOR  Norway  CHE  Switzerland  IRL  Ireland  NZL  New Zealand  CYP  Cyprus  ISL  Iceland  PRT  Portugal  DEU  Germany  ISR  Israel  SWE  Sweden  DNK  Denmark  ITA  Italy  USA  United States  ESP  Spain  JPN  Japan  80 Table A.10. Definition and Source of Variables Definition of Variable   Source  Output variables for education    School enrollment, primary (% gross)   World Bank WDI  School enrollment, primary (% net)   World Bank WDI  School enrollment, secondary (% gross)   World Bank WDI  School enrollment, secondary (% net)   World Bank WDI  Tertiary level complete, ages 15+  Barro‐Lee Database  Literacy rate, youth total (% of people ages 15‐24)   World Bank WDI  Average years of school, ages 15+  Barro‐Lee Database  First level complete, ages 15+  Barro‐Lee Database  second level complete, ages 15+  Barro‐Lee Database  PISA Science scores  www.oecd.org/pisa/data  PISA Mathematics scores  www.oecd.org/pisa/data  PISA Reading scores  www.oecd.org/pisa/data  Quality of primary education, 1‐7 (best)  World Economic Forum  Primary education Net Enrollment, net %  World Economic Forum  Secondary Education Gross Enrollment, gross %  World Economic Forum  Tertiary education enrollment, gross %  World Economic Forum  Quality of the education system, 1‐7 (best)  World Economic Forum  Quality of math and science education, 1‐7 (best)  World Economic Forum  Input variables for education  Public education spending per capita in PPP terms, calculated  World Bank WDI  Literacy rate, adult total (% of people ages 15 and above)   World Bank WDI  Teachers per pupil, equal the reciprocal of pupils per teacher  World Bank WDI  Output variables for health  Life expectancy at birth, total (years)   World Bank WDI  Immunization, DPT (% of children ages 12‐23 months)   World Bank WDI  Immunization, measles (% of children ages 12‐23 months)   World Bank WDI  Institute for Health Metrics and  Disability Adjusted Life Expectancy  Evaluation (IHME)  Tuberculosis cases/100,000 pop.  World Economic Forum  Maternal Mortality Ratio (per 100 000 live births)  World Health Organization  Infant mortality rate (probability of dying between birth and age 1 per 1000 live  World Health Organization  births)  Input variables for health  Literacy rate, adult total (% of people ages 15 and above)   World Bank WDI  public spending on health per capita in PPP terms, calculated  World Bank WDI  public spending on health per capita in PPP terms, calculated  World Bank WDI  Output variables for Infrastructure  Quality of overall infrastructure, 1‐7 (best)  World Economic Forum  Quality of roads, 1‐7 (best)  World Economic Forum  Quality of railroad infrastructure, 1‐7 (best)  World Economic Forum  81 Table A.10 (Continued)    Definition of Variable  Source  Quality of port infrastructure, 1‐7 (best)  World Economic Forum  Quality of air transport infrastructure, 1‐7 (best)  World Economic Forum  A. Transport infrastructure  World Economic Forum  Quality of electricity supply, 1‐7 (best)  World Economic Forum  Input variables for Infrastructure  Private investment spending per capita in real terms, calculated  World Economic Outlook  Public investment spending per capita in real terms, calculated  World Economic Outlook  Variables used in the calculation  World Bank WDI  Pupil‐teacher ratio, primary  World Bank WDI  Public spending on education, total (% of GDP)  World Bank WDI  GDP per capita, PPP (constant 2011 international $)  World Bank WDI  Health expenditure, private (% of GDP)  World Bank WDI  Health expenditure, public (% of GDP)  World Bank WDI  GDP (constant 2010 US$)   World Economic Outlook  GDP (current US$)   World Economic Outlook  National Currency per PPP dollar (Units National Currency per PPP dollar)  World Economic Outlook  Population (Millions of Persons)  World Economic Outlook  Variables used in the Panel Tobit regression  Wages and salaries (% of total public expenditure)  World Bank WDI  Total government expenditure (% of GDP)  World Bank WDI  Share of expenditures publicly financed (public/total)  World Bank WDI  GDP per capita in constant 2011 US dollars  World Bank WDI  Urban population (% of total)  World Bank WDI  Dummy variable for HIV/AIDS  WHO mappings of diseases  Gini Coefficient  World Bank WDI  Aid (% of fiscal revenue) calculated as Official development assistance  World Bank WDI  and official aid (current US$) *official exchange rate * PPP conversion factor  / Revenue, excluding grants (current LCU)  Institutional Indicators including  a. Worldwide Governance Research Indicators  World Bank WDI  Control of Corruption: Estimate  World Bank WDI  Government Effectiveness: Estimate  World Bank WDI  Political Stability and Absence of Violence/Terrorism: Estimate  World Bank WDI  Regulatory Quality: Estimate  World Bank WDI  Rule of Law: Estimate  World Bank WDI  Voice and Accountability: Estimate  World Bank WDI  82 Figure A. 1. Efficiency Frontiers for Education Free Disposable Hull (FDH) Data Envelopment Analysis (DEA) Net Primary Enrollment vs Education Expenditure Net Primary Enrollment vs Education Expenditure 100 100 IRNESPCHN CAN ARG GEO NLDGBR SWE NOR IRNESPCHN CAN ARG GEO NLDGBR SWE NOR SLE VNMFRA TUN NZL FIN ISL DEU SLE VNMFRA TUN NZL FIN ISL ITA DEU KOREGY URY MYS MWI TJK SVN CRI CYP DNK ITA KOREGY URY MYS MWI TJK SVN NPL CRI CYP DNK AUS LTU POL FJI NPL NICCPVLVA AUS LTU POL FJI NICCPV RWALVA LKA MUS PAN RUS BGR ARMMNG EST RWA MAR UKR BLZ BGR ARM LKA MUS PAN RUS MNG ESTMAR UKR BLZ KHM TTO MEX HND KHM TTO MEX HND OMN PER TUR ECU BEN OMN PER TUR ECU BEN CHECHL ALB THA SLV LAO USA PHL BRA TGO BRB CHECHL ALB USA PHL THA SLV LAO BRA TGO BRB BLRUGA VCT SAU BLRUGA VCT SAU GTMCOL GTMCOL IDN Net Primary School Enrollment IDN HUN BDI HUN BDI Net Primary School Enrollment IND VEN IND VEN 90 90 BGD COG BOL BGD COG BOL AZE CMR AZE CMR PRY BWA PRY BWA KGZ MDA LBN ROM DOM KGZ MDA KAZLBN ROM DOM ATG HRV ZWE NAM KAZ ATG HRV ZWE MOZ NAM MKD MOZ VUT MKD VUT GHA GHA KEN KEN TZA TZA KNA AGO COM KNA AGO COM GUY GUY 80 80 SWZ ETH LSO SWZ ETH LSO PAK MRT PAK MRT GIN GIN 70 TCD SEN SLB 70 TCD SEN SLB CIV CIV GMB GMB CAF GNB CAF GNB BFA BFA NGA NGA MLI MLI 60 60 NER NER SDN DJI SDN DJI 50 50 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI Data Source: World Bank WDI (a.1) (a.2) Net Secondary Enrollment vs Education Expenditure 150 Net Secondary Enrollment vs Education Expenditure 150 AUS AUS Net Secondary School Enrollment Net Secondary School Enrollment ESP NLD ESP NLD FIN DNK FIN DNK NZL NZL SWE NOR SWE NOR FRA CRI ISL FRA CRI ISL EST BLRVCT GBR EST BLRVCT GBR ATG CAN LVA ATG CAN LVA BRB 100 BRB 100 KAZ ITA OMN DEU LTU HUN ARG SVN SAU KAZ ITA OMN DEUHUNLTU ARG SVN SAU CZE HRV POL COL AUT CZE HRV LKA CHE POLCOL ECU BRAAUT ROMLKA CHE THA KOR GEO ECU BRA ROM KOR GEO RUS BGRTHA CHL UKR CYP RUS BGR CHL PER USA URY UKR CYP PER USA URY KNA SVK MUS ALB MNG ARM TUR ZAF KNA SVK MUS ALB ARM TURMNG ZAF LCA CHN GUY CHN GUYLCA FJI CPV TUN FJI CPV KGZ JAMTUN MDA PHL KGZ JAM MDA IRNTKM PHL VEN TJK BOL MEX IRNTKM VEN TJK MEX BOL IDNMKD EGY IDN MKD EGY DOM SLV BLZ DOM SLV BLZ PAN NIC PAN NIC LBN PRYMYS HND LBN PRYMYS HND IND IND GTM MAR GTM MAR NPL NPL KEN SWZ KEN SWZ COM COM GMB GHA GMB GHA VUT VUT TGO COG BGDBENTGO COG BGDBEN 50 LAO CMR LSO 50 LAO CMR LSO SLB SLB NGA SLE ZWE NGA SLE ZWE DJI CIVSEN SDN DJI CIV ZARSEN SDN PAK MLIZAR GIN PAK MLI GIN MDGMWI RWA ETH MDGMWI RWA ETH TZA TZA BDI AGO BDI MOZ AGOMRT MOZ BFA MRT BFA UGA TCD UGA TCD CAF NER CAF NER 0 0 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI Data Source: World Bank WDI (b.1) (b.2) Tertiary Gross Enrollment vs Education Expenditure Tertiary Gross Enrollment vs Education Expenditure 100 100 KOR KOR FIN FIN USA USA SVN SVN 80 80 AUS UKR NZL AUS UKR NZL DNK DNK Tertiary Gross Enrollment Tertiary Gross Enrollment RUSESP ISL RUSESP ISL LTU VEN NOR LTU VEN NOR ARG SWE ARG SWE POL POL EST NLD EST NLD ITA LVA ITA LVA HUN CHL URY AUT HUN CHL URY AUT 60 60 CZE CAN GBR CZE CAN GBR ROM BRB ROM BRB BGR MNG FRA BGR MNG FRA SVK CHEHRV DEU SVK CHEHRV DEU LBN TUR LBN TUR THA KGZ THA KGZ KAZ PAN IRN ARM CYP KAZ PAN IRN ARM CYP 40 40 PER MKD ECU COL MDA SAU PER MKD ECU COL MDA SAU BOL BOL DOM MYS DOM MYS ALB PRY TUN CRI ALB PRY TUN CRI GEO EGY GEO EGY OMN MUS PHL MEX BRA OMN MUS PHL MEX BRA CHN IDN JAM CHN IDN JAM SLV SLV 20 20 TJKHND TJKHND AZE LAO NIC GTM IND CPV VNM AZE LAO NIC GTM IND CPV VNM LKA ZAF LKA ZAF MAR MAR GUY TTO BLZ GUY TTO BLZ NGA KHM BGD CMR GINGHA NPL NGA KHM BGD CMR GINGHA NPL GAB BEN CIVSEN NAM BWA GAB BEN CIVSEN NAM BWA PAK MLIRWA LSO PAK MLIRWA LSO AGO UGA MRT ZWE ETH KEN SWZ AGO UGA MRT ZWE ETH KEN SWZ MDG BFA GMB SLE TZAMOZ BDI MDG BFA GMB SLE TZAMOZ BDI TCD MWI TCD MWI 0 0 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/World Economic Forum Data Source: World Bank WDI/World Economic Forum (c.1) (c.2) 83 Quality of primary education vs Education Expenditure Quality of primary education vs Education Expenditure 7 7 FIN FIN CHE 6 CHE 6 Quality of primary education Quality of primary education BRB BRB NZL NZL LBN CAN NLD ISL LBN CAN NLD ISL AUSEST CYP AUSEST CYP MYS SVN SWE MYS SVN SWE 5 5 KOR AUT GBR FRA DNK KOR DEU AUT GBR FRA DNK DEU NOR NOR LTU USA CRI LKA LTU USA CRI LKA CZE ITA HRV TTO LVA CZE ITA HRV TTO LVA CHN UKR CHN UKR GUY GMB POL GUY GMB TUN POL ALB SVK MUSIDNALB TUN SVK MUSIDN RUS HUN SAU RUS OMN HUN SAU OMN IRN 4 IRN BWA 4 KAZ ROM BGR MKD CPV RWA ZWESWZ BWA KAZ ROM BGR MKD CPV RWA ZWESWZ KEN MDA KEN MDA ESP CMR BLZ ESP CMR BLZ THA ARM PHLLAOIND ARM THA PHLLAOIND URY MNG URY VNMSEN MNG VNMSEN LSO JAM GHA CIV LSO JAM GHA GEOTUR COLTJK CIV BEN GEOTUR COLTJK BEN ETH ETH NPL ECU NPL ECU ARG KGZ NAM ARG KGZ NAM AZEGAB KHM PAK MAR UGA AZEGAB KHM PAK MAR UGA BFA 3 BFA PAN 3 PAN CHL VEN SLE BOL CHL VEN SLE BOL NGA BGD MDG MWI NGASLVBGD MDG MEX MWI SLV MEX TZA TZA NIC HND NIC HND TCD MLI BRA HTI ZAF TCD MLI BRA HTI ZAF GTM GIN GTM GIN BDI PER EGY MRT MOZ BDI PER EGY MRT MOZ DOM PRY DOM PRY 2 2 AGO AGO 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/World Economic Forum Data Source: World Bank WDI/World Economic Forum (d.1) (d.2) Quality of the education system vs Education Expenditure Quality of the education system vs Education Expenditure CHE CHE 6 6 FIN FIN CAN ISL CAN ISL Quality of the education system Quality of the education system NZL NZL NLD BRB NLD BRB MYS CYP SWE DNK AUS MYS CYP SWE DNK LBN DEU AUS NOR LBN DEU NOR 5 5 USA CRI USA CRI GBR AUT GBR AUT LKA GMBFRA LKA GMBFRA KEN KEN TTO TTO IDN INDEST ZWE SAU IDN INDEST ZWE SAU PHL TUN PHL MUS ALB TUN CZE MUS ALB SVN RWA CZE SVN RWA 4 4 CHN GUY CHN GUY OMN KORLTU LAO CPV BWA OMN KORLTU LAO CPV UKR BWA MKD GHA LVA UKR MKD GHA LVA POL SEN POL SEN THA CMRMWILSO ETH THA CMRMWILSO ETH ITA RUS BEN UGA VNM ITA RUS BEN UGA VNM COL KHM TJK JAM COL KHM TJK JAM KAZ ROM ESP HUN PAK CIV SWZ KAZ ROM ESP HUN NGA CHL TURPAK CIV SWZ BGR NGA CHL TUR ARM BGR HRV ARM NPL HRV URY BGD NPL URY BGDTZA ECU ECU TZA ARG MDA AZE IRN ARG MDA AZE IRN BOL BOL PAN GEO MAR MDG PAN GEO MAR MDG SLV MEXMOZ KGZ 3 SLV MEXMOZ KGZ 3 NAM NAM SVK SLE SVK SLE MLI BRA TCD MLI BRA TCD HND HND GAB GAB BFA BLZ GTMVEN BFA BLZ GTMVEN NIC NIC PER MNG GIN PER MNG GIN DOM DOM MRT MRT BDI EGY BDIZAF EGY ZAF PRY HTI PRY HTI AGO AGO 2 2 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/World Economic Forum Data Source: World Bank WDI/World Economic Forum (e.1) (e.2) Pisa Mathematics Literacy score vs Education Expenditure Pisa Mathematics Literacy score vs Education Expenditure 550 550 KOR CHN KOR CHE FIN CHN CHE FIN CAN NLD CAN NLD Pisa Mathematics Literacy score Pisa Mathematics Literacy score DEU DEU 500 SVN NZL DNK NZL 500 AUS AUS SVN DNK POL AUT ISL NOR POL NOR CZE FRA VNM GBR SWE AUT FRA VNM ISL SWE ITA SVK ESP CZE GBR RUS HUN LVA ITA SVK ESP HUN USALTU RUS USA LVA LTU HRV HRV 450 450 ROM BGR TUR CYP AZE ROM BGR TUR CYP AZE CHL URY ARG MDA THA TTO CHL URY ARG MDA MEX THA TTO MEX 400 KAZ GEO 400 CRI KAZ GEO LBN ALB CRI LBN ALB COL BRA IDN COL PER IDN BRA MKD TUN PER MKD TUN PAN 350 PAN 350 DOM KGZ DOM KGZ 300 300 1000 1500 2000 2500 Orthogonalized Public Expditure on Education 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Data Source: World Bank WDI/NCES Data Source: World Bank WDI/NCES (f.1) (f.1) 84 Pisa Reading Literacy vs Education Expenditure Pisa Reading Literacy vs Education Expenditure 550 550 KORCAN FIN KORCAN FIN NZL NZL AUS NOR AUS NOR 500 500 DEU POL NLD DEU POL NLD CHE USA FRA GBR SWE DNK CHE USA FRA GBR SWE DNK CHN SVN CHN SVN ISL ISL ITACZE ESP HUN VNM LVA ITACZE ESP HUN VNM LVA Pisa Reading Literacy Pisa Reading Literacy HRV HRV RUS AUT RUS AUT LTU LTU SVK SVK 450 450 CHL CHL TUR TUR CYP CYP ARG ARG BGR URY ROM BGR URY ROM CRI MEX CRI MEX TTO TTO COL COL MDA THA MDA THA BRA BRA 400 400 GEO IDN GEO IDN ALB ALB KAZ KAZ PER TUN PER TUN PAN PAN AZE AZE DOM 350 DOM 350 MKD MKD LBN LBN KGZ KGZ 300 300 1000 1500 2000 2500 1000 1500 2000 2500 Orthogonalized Public Expditure on Education Orthogonalized Public Expditure on Education Data Source: World Bank WDI/NCES Data Source: World Bank WDI/NCES (g.1) (g.1) Average Year of School vs Education Expenditure Average Year of School vs Education Expenditure 15 15 USA USA SVK CZE CHE SVK CZE CHE KOR CAN DEU KOR CAN DEU GBR AUS GBR AUS EST EST ISL HUN SVN ISL SWE NOR HUN SVN SWE NOR KAZ RUS HRVLTU POL NLD CYP DNK KAZ RUS HRVLTU POL NLD CYP DNK LVA LVA Average Year of School FRA Average Year of School BGR FRA MYS UKR BLZ NZL BGR MYS UKR BLZ NZL ROM ESP TTO KGZ ROM ESP TTO KGZ ARMFJI MDA ITA ARMFJI MDA FIN ITA TJK FIN TJK 10 10 LKA CHL ALB LKA CHL ALB AUT ZAF BWA AUT ZAF JAM BWA JAM PAN MNG ARG PAN MNG ARG BRB MUS COL BRB MUS COL GABIRN PER MEX SAU GABIRN PER MEX SAU PHL VEN PHL VEN GUYURY BOL CRI THA GUYURY BOL CRI DOM THA CHN DOMCHN IDN ECU PRY TUN IDN ECU PRY TUN SLV ZWE BRA SLV TUR ZWE BRA TUR VNM EGY VNMGHA EGY GHA NIC IND NIC IND CMR HND CMR HND NAM BGD TGO KEN COG NAM BGD TGO KEN COG LSO TZA UGA LSO TZA UGA MAR SWZ PAK HTI MAR SWZ PAK HTI LAO GTM LAO GTM KHM CIV 5 KHM CIV 5 MRT BEN MWI NPL MRT BEN MWI NPL RWA RWA SLE GMB SLE GMB CAF ZAR BDI CAF ZAR BDI SDN SDN SEN SEN MOZ MOZ MLI MLI NER NER 0 0 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expenditure on Education Orthogonalized Public Expenditure on Education Data Source: World Bank WDI, Barro-Lee database Data Source: World Bank WDI, Barro-Lee database (h.1) (h.2) Second Level Complete vs Education Expenditure Second Level Complete vs Education Expenditure 80 80 CZE CZE KGZ TJKKGZ KAZ SVK TJK KAZ SVK 60 60 POL Second Level Complete POL Second Level Complete ARM SVN ARM SVN ZAF MDA ROM ZAF MDA ROM HUN HUN HRVLTU TTO HRVLTU TTO DEU BGR SWE DEU BGR SWE CHE EST GBR CHE EST GBR MUS ALBMNG MUS ALB MNG AUT AUT CHL FJI MYS CHL FJI MYS PER AUS PER AUS LVA FRA 40 LVA FRA 40 NLD DNK ITA NLD DNK ITA NOR NOR UKR CYP KOR JAMUKR CYP KOR VEN JAM VEN GUY GUY USA USA CAN ARG BLZ CAN ARG BLZ SAU SAU IRN ECU BWA IRN ECU BWA PAN FIN PAN BGD BRA EGYCOL FIN BGD BRA EGYCOL IND GAB DOMCHN IND GAB DOM IDNSLV IDNSLV RUSESPCHN PHL PRY GHA TUR ISL PHL PRY GHA RUSESP TUR HTI ISL HTI MEX VNM BRB CRI THAPAK MEXVNM BRB CRI 20 THAPAK 20 LKA URY LKA URY BEN BOL NICBENHNDBOL NIC HND NAM CMR NAM CMR TUN TUN LSO NZL NPL KENLSO NZL GTM NPL KEN ZAR GTM ZAR MAR MAR TGO TGO CAF CAF CIV MRT CIV MRT MWI MWI SLE RWA LAO SLE RWACOG LAO KHM UGA COG ZWE BDI SWZ KHM UGA ZWE BDI SWZ MLI GMBMOZ MLI MOZ GMB NER SDN TZA NER SEN SDN TZA SEN 0 0 500 1000 1500 2000 2500 500 1000 1500 2000 2500 Orthogonalized Public Expenditure on Education Orthogonalized Public Expenditure on Education Data Source: World Bank WDI, Barro-Lee database Data Source: World Bank WDI, Barro-Lee database (i.1) (i.2) 85 Youth Literacy Rate vs Education Expenditure Youth Literacy Rate vs Education Expenditure 100 100 LBN AZE ROM ITA RUSGEO ARM CHN PER HRV IDNESP ALB CHL LTU BLR URYEST KGZ ARG MYSUKR BOLLVA MDA CRI SAU CYP LBN AZE ROM ITA RUSGEO ARM CHN PER HRV IDNESP ALB CHL LTU BLR TURPRY URYEST KGZ ARG MYSUKR BOL ZAFLVA MDA CRI SAU CYP PAN LKA IRN PHL MUS OMN TURPRY MNG COLECUMEX VEN CPV ZAF BRA PAN LKA IRN PHL MUS OMNMNG COLECU VENMEX CPV BRA DOM BGR THA SLV VNM TUN DOM BGR THA SLV VNM TUN GUY HND NAM GUY HND NAM GTM SWZ GTM SWZ EGY ZWE EGY ZWE GAB GAB KHM IND UGA KEN GHA KHM IND UGA KEN GHA NPL MAR LSO NPL MAR LSO ZAR ZAR TZA TGO COG TZA TGO COG 80 80 CMR RWABDI CMR RWA BDI AGO AGO Youth Literacy Rate Youth Literacy Rate MWI MWI LAO PAK COM LAO PAK COM MDG MDG MOZ MOZ BGD BGD SEN SEN 60 GMB 60 GMB GNB GNB SLE SLE BEN BEN CIV BFA CIV BFA MLI MLI 40 40 TCDGIN TCDGIN CAF CAF NER NER 20 20 0 200 400 600 800 1000 0 200 400 600 800 1000 Orthogonalized Public Expenditure on Education Orthogonalized Public Expenditure on Education Data Source: World Bank WDI Data Source: World Bank WDI (j.1) (j.2) 86 Figure A.2. Efficiency Frontiers for Health Free Disposable Hull (FDH) Data Envelopment Analysis (DEA) Life Expectancy vs Health Expenditure Life Expectancy vs Health Expenditure 90 90 CHE AUS ESP ITA ISL CHE FRA SWE LUXAUS ESP ITA ISL CAN FRA NOR SWE KOR LUX CAN NOR NZL AUT NLD KOR GRC GBR NZL AUT NLD GRC FIN GBR SVNFIN 80 DEU DEU 80 CYP SVN DNK CYP DNK LBN CRI CHL LBN CRI USA CHL CZE USA CZE ALB PAN HRV ALB PAN HRV POL POL URY OMNBHR MEX EST SVK URY OMNBHR Life Expectancy MEX EST SVK ARG ATGCHN Life Expectancy ARG ATGCHNECU JAM VNM BHS ECU JAM BRB DZA TUN HUN MKD VNM MYS BHS IRN THA TUR BRB DZA TUN HUN MKD LCA MYS IRN THA TUR LKA BRAMAR ROM BGRLCA NIC SAU LKA BRAMAR ROM BGR PER NIC SAU VEN MUS LVA PER LTUARM COL WSM JOR VEN MUS LVA LTUARM COL WSM JOR DOM GRD DOM GRD GEO VCT HND GEO VCT PRYHND TON SLV PRY TON SLV BLR GTM CPV AZE BLR GTM CPV VUT AZE EGY UKR VUT BGD BGD 70 UZB MDA TJK TTO KAZ RUS EGY FJI UKR TTO KAZ 70 UZB BLZKGZMDA TJK RUS FJIBLZ KGZ SLB SLB IDN PHL NPL IDNMNGPHL NPL MNG BOL TKM IND BOL KHM TKM IND KHM GUY GUY PAK SEN PAK SEN LAO KEN LAO PNG RWA PNG RWA KEN GAB YEM MDG GAB YEM MDG SDNERI ETH SDNERI ETH BWA COG COM TZA MRTHTI BWA COG COM TZA MRTHTI GHA GHA DJI 60 DJI NAMGMB 60 NAMGMB BEN AGO BEN MWI AGO MWI ZAF ZMB TGO BFA UGA ZMB ZAF UGA TGO BFA NER ZAR GIN NER ZAR GIN CMR MLI ZWE MOZ GNB BDI CMR MLI ZWE MOZ GNB BDI SWZ SWZ NGACIV LSO NGACIV LSO TCD 50 TCD 50 SLE CAF SLE CAF 1000 2000 3000 4000 5000 1000 2000 3000 4000 5000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI Data Source: World Bank WDI (a.1) (a.2) Immunization DPT vs Health Expenditure Immunization DPT vs Health Expenditure 100 100 OMN BHR CYP IRN THA LKA ATG HUN CHN POL ALB FJI LUX MAR UZB SVK GRC CZE JOR FIN FRA OMN BHR CYP IRN THA ATG LKACHN POLHUN ALBMAR FJI LUX UZB SVK GRC CZE JOR FIN FRA SAU KAZ MUS MNG BLR LCA TUN NIC SWE SAU KAZ MUS MNG BLR TUN LCA NIC SWE MYSKOR BHS KNA TKM TUR RUS BRA DMA GRD VCT GUY RWA GMB ESP HND KGZ NLD MYSKOR BHS KNA TKM TUR RUS BRA DMA GRD VCT GUY RWA HND GMB ESP KGZ NLD EGY BLZ MKD BWA DZA BGD HRV BDISVN ITA CHE DEU EGY BWA DZA BLZ MKD BGD TJK HRV BDI SVN ITA CHE DEU ARM CPV TJK URY BOL SLB GBR USA ARM CPV URY BOL SLB GBR USA NOR CHL ARG LTU EST BGR ERI LSO NOR ARG CHL LTU EST BGR ERI LSO LVA JAM GHA SWZ MWI ISL NZL DNK LVA JAM GHA SWZ MWI ISL NZL DNK TTO ROM MEX PERGEO BRB SDN AUS SLVTZA TTO MEX ROMGEO PERBRB SDN AUS SLVTZA KEN VNM BFA CAN AUT KEN VNM BFACRI CAN AUT COLSEN NPL MDA ZWE CRI COLNPL SEN MDA ZWE AZE KHM PRY AZE KHM PRY ECU SLE ECU SLE DOM GTM NAMGNB DOM GTM NAMGNB CMR ZMBCOM ZMB CMRCOM DJI PAN TGO PAN TGODJI INDCOG VEN INDCOG VEN LBN LBN 80 80 IDN LAOTON IDN LAOTON Immunization DPT Immunization DPT PHL UGA PHL UGA CIV BEN CIV BEN MOZ MOZ GAB GAB PAKMRT ZAR PAKMRT ZAR ZAF PNG ZAF PNG YEM MDG YEM MDG AGO MLI NER AGO MLI NER ETH ETH VUT VUT HTI HTI WSM WSM 60 60 GIN GIN UKR UKR NGA NGA CAF CAF 40 40 TCD TCD 1000 2000 3000 4000 1000 2000 3000 4000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI Data Source: World Bank WDI (b.1) (b.2) Immunization Measles vs Health Expenditure Immunization Measles vs Health Expenditure 100 100 OMN BHR KAZ TKM MUS IRN BLR LKACHN HUN MAR NIC OMN BHR KAZ MUSTKM IRN LKA THA ATG CHN BLR HUN MAR NICCZE GRC SAU KOR RUS THA ATG BRA ALB GUY VCT UZB GRC CZE KGZ SAU KOR RUS ALB BRA POL VCTUZB GUY LUX SVK KGZ KNA POL MNG LCA DMA LUX SVK BLZ VNM DEU SWE KNA MNG BWA TUR LCA DMA ARM BLZ TUN VNM JOR ESP FIN DEU SWE BWA TUR ARMTUN JOR ESP FIN MEXGRD TZA RWA NLD MEXGRD TZA URY BOLRWA NLD BGR MKD DZA URY BOL ERI TJKGMB SVN HRV MYS BGR EGY MKD DZA PAN ERI TJK HRV GMB SVN MYS ARG LVA EGY LTU ESTFJI PAN CPVAUS AZE BHS ARG LVA LTU EST FJI CPVAUS BDI NOR AZE BHS CHL BDI CHE NOR CHL ECU SLV SWZ CHE ECU SLV SWZ GEO ROM ISL GEO ROM PERBRB BGD ISL USA NZL TTO PERBRB BGD COL MDA GBR CAN USA NZL 90 TTO JAM COL GHAMDA GBR CAN JAM PRYGHA MWI 90 PRY MWI SLB SLB ZWE LSO ITA FRA ZWE LSO ITA FRA BFA HND BFA HND NPL CRI AUT DNK CYP NPL CRI ZMB AUT DNK CYP VEN VEN DOM SDN Immunization Measles DOM ZMB SDN GTM Immunization Measles GTM PHL TON PHL TON IND KHM KENMOZ IND KHM KENMOZ COM SEN COM 80 SEN PNG SLE DJI SLE PNG 80 DJI LBN CMR NAMUGA GNB LBN CMR NAMUGA GNB IDN COG IDN COG LAO MLI ZAR LAO MLI ZAR TGO TGO MRT NER 70 MRT NER BEN 70 BEN ZAF ZAF WSM WSM GAB CIV UKR YEM GAB CIV UKR YEM ETH ETH AGO AGO MDG PAK HTI MDG 60 PAK HTI 60 GIN GIN VUT TCD VUT TCD NGA 50 NGA 50 CAF CAF 1000 2000 3000 4000 1000 2000 3000 4000 Orthogonalized Public Expditure on Health Orthogonalized Public Expditure on Health Data Source: World Bank WDI Data Source: World Bank WDI (c.1) (c.2) 87 Table B.1. Spearman rank-correlation on input efficiency rankings (DEA)   Infant  Maternal  Second  Average  Gross  Net  Gross  Net  Free  PISA  DALE  Survival  Survival  Level  year of  Secondary  Secondary  Primary  Primary  Tuberculosis  Science    Rate  Rate  Complete  School  Enrollment  Enrollment  Enrollment  Enrollment        PISA  1.000  0.72***  0.80***  0.74***  0.76***  0.63***  0.78***  0.19  0.37***  0.34**  ‐0.18  Science  (55)  (51)  (51)  (44)  (52)  (55)  (52)  (55)  (55)  (55)  (55)  Second Level  1.000  0.97***  0.96***  0.95***  0.80***  0.85***  0.23**  0.28***  0.32***  0.23**  Complete  (113)  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  Average year of    1.000  0.95***  0.95***  0.75***  0.87***  0.28***  0.30***  0.36***  0.17*  School  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Gross Secondary      1.000  0.96***  0.75***  0.88***  0.33***  0.27***  0.39***  0.35***  Enrollment  (103)  (103)  (101)  (100)  (102)  (93)  (103)  (101)  Net Secondary        1.000  0.73***  0.87***  0.40***  0.34***  0.45***  0.32***  Enrollment  (128)  (126)  (119)  (127)  (111)  (128)  (126)  Gross Primary          1.000  0.74***  0.08  0.09  0.14  0.13  Enrollment  (133)  (126)  (132)  (116)  (133)  (131)  Net Primary            1.000  0.50***  0.41***  0.50***  037***  Enrollment  (126)  (125)  (111)  (133)  124)  DALE              1.000  0.78***  0.88***  0.56***    (151)  (126)  (151)  (148)  Free Tuberculosis                1.000  0.79***  0.55***    (126)  (126)  (148)  Infant Survival Rate                  1.000  0.62***    (152)  (148)  Maternal Survival                    1.000  Rate  (148)    Note 1. Figures are correlation coefficients from Spearman test, number of observations are in parentheses. 2. ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 88 Table B.2. Spearman rank-correlation on output efficiency rankings (DEA)   Free  Infant  Maternal  Second  Average  Gross  Net  Gross  Net  PISA  DALE  Tuberculosi Survival  Survival  Level  year of  Secondary  Secondary  Primary  Primary  Science    s  Rate  Rate  Complete  School  Enrollment  Enrollment  Enrollment  Enrollment        PISA  1.000  0.21  0.61***  0.76***  0.68***  ‐0.08  0.53***  0.39***  0.43***  0.64**  0.66***  Science  (55)  (51)  (51)  (44)  (52)  (55)  (52)  (55)  (55)  (55)  (55)  Second Level  1.000  0.84***  0.71***  0.69***  ‐0.008  0.41***  0.53***  0.50***  0.66***  0.72***  Complete  (113)  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  Average year of    1.000  0.86***  0.86***  ‐0.04  0.55***  0.69***  0.66***  0.85***  0.86***  School  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Gross Secondary      1.000  0.91***  ‐0.06  0.65***  0.73***  0.70***  0.87***  0.89***  Enrollment  (103)  (103)  (101)  (100)  (102)  (93)  (103)  (101)  Net Secondary        1.000  0.0.16*  0.66***  0.81***  0.75***  0.88***  0.88***  Enrollment  (128)  (126)  (119)  (127)  (111)  (128)  (126)  Gross Primary          1.000  0.38***  0.10  ‐0.007  0.08  0.05  Enrollment  (133)  (126)  (132)  (116)  (133)  (131)  Net Primary            1.000  0.65***  0.56***  0.67***  0.66***  Enrollment  (126)  (125)  (111)  (126)  124)  DALE              1.000  0.88***  0.90***  0.84***    (151)  (126)  (151)  (148)  Free Tuberculosis                1.000  0.84***  0.82***    (126)  (126)  (148)  Infant Survival                  1.000  0.93***  Rate  (152)  (148)    Maternal Survival                    1.000  Rate  (148)  Note 3. Figures are correlation coefficients from Spearman test, number of observations are in parentheses. 4. ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 89 Table B.3. Spearman rank-correlation on input efficiency rankings (FDH)   PISA  Second  Average  Gross  Net  Gross  Net  DALE  Free  Infant  Maternal  Science  Level  year of  Secondary  Secondary  Primary  Primary  Tuberculosis  Survival  Survival  Complete  School  Enrollment  Enrollment  Enrollment  Enrollment    Rate  Rate      PISA  1.000  0.69***  0.80***  0.69***  0.71***  0.41***  0.76***  0.18  0.24*  0.30**  0.26*  Science  (55)  (51)  (51)  (44)  (52)  (55)  (52)  (55)  (55)  (55)  (55)  Second Level  1.000  0.91***  0.87***  0.80***  0.50***  0.83***  0.19**  0.16*  0.26***  0.28***  Complete  (113)  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Average year of    1.000  0.88***  0.86***  0.44***  0.85***  0.30***  0.24**  0.40***  0.39***  School  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Gross Secondary      1.000  0.89***  0.43***  0.87***  0.37***  0.19*  0.40***  0.41***  Enrollment  (103)  (103)  (101)  (100)  (102)  (93)  (103)  (101)  Net Secondary        1.000  0.35***  0.82***  0.47***  0.34***  0.51***  0.47***  Enrollment  (128)  (126)  (119)  (127)  (111)  (128)  (126)  Gross Primary          1.000  0.50***  ‐0.05  ‐0.13  ‐0.11  ‐0.13  Enrollment  (133)  (126)  (132)  (116)  (133)  (131)  Net Primary            1.000  0.49***  0.32***  0.49***  0.46***  Enrollment  (126)  (125)  (111)  (133)  124)  DALE              1.000  0.75***  0.86***  0.81***    (151)  (126)  (151)  (148)  Free Tuberculosis                1.000  0.80***  0.73***    (126)  (126)  (148)  Infant Survival Rate                  1.000  0.93***    (152)  (148)  Maternal Survival                    1.000  Rate  (148)  Note 5. Figures are correlation coefficients from Spearman test, number of observations are in parentheses. 6. ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 90 Table B.4. Spearman rank-correlation on output efficiency rankings (FDH)   PISA  Second  Average  Gross  Net  Gross  Net  DALE  Free  Infant  Maternal  Science  Level  year of  Secondary  Secondary  Primary  Primary    Tuberculos Survival  Survival  Complete  School  Enrollment  Enrollment  Enrollment  Enrollment  is  Rate  Rate        PISA  1.000  0.20  0.63***  0.70***  0.64***  ‐0.04  0.47***  0.34**  0.43***  0.63**  0.64***  Science  (55)  (51)  (51)  (44)  (52)  (55)  (52)  (55)  (55)  (55)  (55)  Second Level  1.000  0.84***  0.72***  0.70***  0.10  0.41***  0.52***  0.50***  0.67***  0.72***  Complete  (113)  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Average year of    1.000  0.87***  0.86***  0.06  0.57***  0.68***  0.65***  0.85***  0.86***  School  (113)  (87)  (106)  (110)  (105)  (113)  (106)  (113)  (113)  Gross Secondary      1.000  0.90***  0.03  0.66***  0.72***  0.70***  0.87***  0.89***  Enrollment  (103)  (103)  (101)  (100)  (102)  (93)  (103)  (101)  Net Secondary        1.000  0.20***  0.68***  0.79***  0.74***  0.86***  0.88***  Enrollment  (128)  (126)  (119)  (127)  (111)  (128)  (126)  Gross Primary          1.000  0.37***  0.19**  0.00  0.11  0.11  Enrollment  (133)  (126)  (132)  (116)  (133)  (131)  Net Primary            1.000  0.65***  0.56***  0.67***  0.67***  Enrollment  (126)  (125)  (111)  (133)  124)  DALE              1.000  0.87***  0.88***  0.83***  (151)  (126)  (151)  (148)  Free Tuberculosis                1.000  0.84***  0.82***    (126)  (126)  (148)  Infant Survival                  1.000  0.93***  Rate  (152)  (148)    Maternal Survival                    1.000  Rate  (148)  Note Figures are correlation coefficients from Spearman test, number of observations are in parentheses. ***0.01 significance level, ** 0.05 significance level, *0.10 significance level, and insignificant otherwise 91