Journal Issue
Share
Article

Identifying Structural Reform Gaps in Emerging Europe, the Caucasus, and Central Asia1

Author(s):
Norbert Funke, Asel Isakova, and Maksym Ivanyna
Published Date:
March 2017
Share
  • ShareShare
Show Summary Details

I. Introduction

Structural reforms are often seen as central to increasing potential growth, which has declined since the global financial crisis began, and to reduce public debt. In many advanced economies, interest rates have hit the zero lower bound and fiscal space is limited. Structural reforms are advocated to stimulate growth and reduce public debt. In emerging markets, they are expected to promote faster economic convergence and to overcome the middle-income trap. At the 2014 G-20 meeting, structural reforms were emphasized as a means of promoting recovery, with governments agreeing to adopt national growth strategies. Since then the emphasis on structural reforms has been renewed.

There is empirical cross-country evidence that structural reforms are beneficial for growth,2 with a more recent analysis suggesting that short- and medium-term impacts depend on economic conditions (Duval, Furceri, et al. 2016). A number of empirical microeconomic studies have demonstrated the positive effect of particular national reforms (e.g., Besley and Burgess 2002; Fabiano and Viviano 2011; Banerjee, Duflo, and Qian 2012). However, macroeconomic and microeconomic studies of structural reforms do not often make it easy to compare reform needs in different areas. Cross-country regressions usually use an aggregate index of reforms that lacks detail, and microeconomic studies are often too granular, with their results dependent on country- and reform-specific institutions. Meanwhile, because government resources are limited, there is a need to prioritize reforms.

On average, countries with higher per capita income tend to score better on structural indicators, as evidenced by the close correlation between level of development and structural reform indicators. Moreover, reform priorities may change over time as development advances, moving from basic institutional and macroeconomic reforms to reforms to improve efficiency, such as those related to human capital, product, labor and financial markets, and reforms that focus on innovation policies (see also World Economic Forum 2015).

In identifying essential reforms, authorities may want to take into account explicitly the level of development. As countries aspire to achieve a higher level of development, authorities might ask how a country compares to a (generic) country with similar characteristics but which has already achieved a certain higher level of income?3 Such comparisons can provide additional insights compared to looking at absolute scores or rankings of various reform indicators. Countries that may look quite similar in terms of structural indicators—for example, countries with a similar overall ranking in the 2015–16 Global Competitiveness Index, such as Slovenia (59), Macedonia (60), Hungary (63), Georgia (66), Slovak Republic (67) and Montenegro (70)—may in fact be viewed differently once the level of development is taken into account.

Prior to the global financial crisis, between 2000–2007 per capita GDP in the ECA region increased by about 60 percent on average but is projected to grow by only about 20 percent between 2015–21. It appears unlikely that countries will on average achieve pre-crisis growth rates in the coming years. At the same time, projected growth rates for the coming years are unsatisfactorily low. The best performing country in the region, Georgia, is projected to grow by some 40 percent during 2015-2021, broadly in line with the average of the regional average during the years prior to the global financial crisis and the projected regional average growth rate for the coming years. We take some middle ground and use a 40 percent increase in income as one of two benchmarks.

In this paper, we compare in a cross-sectional analysis structural indicators for 25 countries in emerging Europe, the Caucasus, and Central Asia (ECA)4 with those of a generic country that is 40 percent wealthier as well as a country with the average EU income. Using data from the World Economic Forum (WEF) 2015-16 Global Competitiveness Report (GCR) as example, Section 2 describes the empirical methodology. Section 3 presents the findings, transforms the gaps into a heat map, and discusses the results. Section 4 identifies the largest gaps in the region on a more disaggregated level. Section 5 performs some robustness checks. Finally, Section 6 discusses some limitations of the approach and possible extensions.

In the benchmarking exercise that assumes that countries in the region aspire to increase their incomes by 40 percent in the coming years, about one-third to more than half of the countries were found to have large reform needs in institutions, financial market development, infrastructure, goods and labor market efficiency, and areas related to innovation. With the generally more ambitious goal of reaching the average income in the EU (with the exception of Slovenia), the list of reforms expands considerably. The findings are reasonably robust to changes in various model specifications.

The analysis involves several caveats, including those related to (1) the quality and consistency of data, (2) the nature of the link between structural indicators and per capita income, and (3) and reform complementarities. It is inherently difficult to collect reliable structural reform data; country-specific biases cannot be ruled out. To illustrate our approach, we use data from the GCR, which is based on survey data, unlike other indicators, such as the World Bank (WB) Doing Business Indicators, which are generally based on the application of rules and regulations. While the assessment of certain reform areas covered in the GCR and WB reports leads on average to broadly similar results, in a given country the results based on different data sources may differ. The paper abstracts from explicitly linking structural reforms to growth. The time series available are short and include the period after the global financial crisis, during which macro policies differed widely. That is why the focus of this analysis is on comparing structural indicators to a “generic” country with higher income per capita.

II. Structural Reform Gaps: Estimation

A number of studies have used benchmarking to identify reform needs. In its “Going for Growth” analysis, the OECD (2013, 2015) benchmarks countries against the OECD median. 5 The World Bank (2015)Doing Business report computes scores for distance to the frontier. The Transition Report of the European Bank for Reconstruction and Development compares countries to the absolute maximum score on a number of reform indicators. For CESEE countries IMF (2016) uses various benchmarks, including advanced Europe and benchmarks for drivers of growth (labor, investment, and productivity). However, none of these methodologies take the level of income explicitly into account in benchmarking. Benchmarking indicators to level of development or other structural characteristics is common in other areas of economics, such as for example in the tax effort literature6 or the IMF External Balance Assessment methodology (Phillips et al. 2013).

To measure the degree of structural development, we use the GCR data from the 2015-16 report. The GCR reports more than 126 indicators for up to 148 countries; these are grouped in 12 broad areas, which the report calls pillars: 1–institutions, 2–infrastructure, 3–macroeconomic environment, 4–health and primary education, 5–higher education and training, 6–goods market efficiency, 7–labor market efficiency, 8–financial market development, 9–technological readiness, 10–market size, 11–business sophistication, and 12–innovation. Each country is scored on each pillar from 1 (worst) to 7 (best).

Empirically, there is a close positive correlation between per capita income and structural indicators (Figure 1): countries with higher per capita income tend to have better structural indicators. There is a similar correlation between disaggregated structural reform indicators and real GDP per capita (Figure 2).

Figure 1.GCR Score and GDP per Capita in 2015

Source: WEF 2015, JVI.

Figure 2.Global Competitiveness and Real GDP per Capita, 2015

Source: WEF 2015.

Note: For each pillar, scores are specified on the vertical axis and each country is scored from 1 (worst) to 7 (best). Pillar 10 (market size) is excluded from the analysis.

In what follows, we estimate the formal link between reform indicators and per capita income, as well as other structural characteristics. We then use the results to compare structural indicators of a given country to those with a generic country that has a 40 percent higher per capita income. We begin by running the following regression for each indicator:

where Iik is indicator k in country i and Xi is the set of controls; α, β estimated parameters, εi the error term. To proxy for level of development, we take the logarithm of annual income per capita in 2015 (measured by GDP per capita) in 2005 prices.

As the initial specification, we choose the regression that includes only a constant and per capita income. As with GDP per capita, all structural reform indicators are logged in order to give less weight to potential outliers. The summary statistics are provided in Appendix Table A1.1.

The difference (or gap) k between the structural indicator in country i and that of a generic country with 40 percent higher income is defined as the residual in regression evaluated for a country with a 40 percent higher income (1):

Each gap is weighted by the inverse of its standard deviation to unify the units of measurement; this allows for comparisons both between countries and between indicators within the same country. For example, if a k-gap in country i is Z, that means that in that country indicator k is Z standard deviations from the trend of a country which is 40 percent richer. If the distribution of the k-gap is close to normal, a gap of –1.65 means that relative to a generic country with 40 percent higher per capita income, country i performs worse than 95 percent of the sample. A positive Z implies a positive gap for country i—it performs better than an average hypothetical country with a 40 percent higher income per capita and other structural characteristics.

Table 1 shows the regression results based on the initial specification for 11 out of 12 GCR pillars.7 As expected, in all areas covered by the GCR index, income per capita has a positive and statistically significant coefficient. On average, across all reform areas richer countries tend to be more advanced on structural reforms. However, the fit varies across different reform indicators. It appears closest in the areas of infrastructure and technological readiness. The link appears weakest in the area of labor markets, in part reflecting that labor market rigidities persist in several advanced economies. To check for the sensitivity of the results, for each pillar, Appendix 2 presents eight different specifications of equation (1), including dummies for ECA and resource-rich countries8, estimations with other regional and country-specific dummies and GDP per capita squared, to account for possible nonlinearity. Resource-rich countries trail others with the same income in almost all reform areas, which suggests that resource rents may at least partially crowd out structural reforms. On average ECA emerging countries tend to perform relatively better in a few areas than countries elsewhere with a similar income level, for example in infrastructure quality, macroeconomic environment, education and health care, and technological readiness. However, on average these countries tend to lag in the areas of institutions, business sophistication, and innovation.9 Overall, the results are qualitatively similar across different specifications. Countries with a higher per capita income tend to score better on structural reform indicators.

Table 1.Regression Results: Initial Specification Dependent Variable – Logarithm of GCR Pillar Score
VariablesPillar 1Pillar 2Pillar 3Pillar 4Pillar 5Pillar 6Pillar 7Pillar 8Pillar 9Pillar 11Pillar 12
lngdp_pc0.07***

(0.01)
0.15***

(0.01)
0.05***

(0.01)
0.07***

(0.01)
0.12***

(0.01)
0.05***

(0.00)
0.03***

(0.01)
0.06***

(0.01)
0.15***

(0.00)
0.07***

(0.00)
0.09***

(0.01)
Constant1.47***

(0.02)
1.31***

(0.02)
1.64***

(0.02)
1.73***

(0.02)
1.42***

(0.01)
1.59***

(0.01)
1.60***

(0.01)
1.48***

(0.01)
1.32***

(0.01)
1.48***

(0.01)
1.33***

(0.02)
Observations132132132132132132132132132132132
R-squared0.450.800.210.560.770.480.150.380.860.630.54
Adjusted R20.450.800.200.550.770.480.150.380.860.620.54
AIC−179.4−199.7−121.1−232.4−239.8−318.3−236−202.6−264.1−277.6−185.9
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ calculations.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ calculations.

III. Benchmarking: Results

In what follows, we analyze structural reform indicators using two benchmarks: countries with per capita income that is higher by 40 percent, and a hypothetical country with income per capita equivalent to the EU average. Figure 3 shows the resulting differences (gaps) for structural indicators for 25 ECA countries. The blue bars represent the range of gaps relative to the case with 40 per cent higher per capita income calculated from eight different specifications; the green diamonds show the size of the gaps based on our initial specification again relative to a country with 40 per cent higher per capita income; and the red diamonds mark the average weighted gap across all specifications. The average gap is a weighted average of gaps from all eight specifications, using the absolute values of the Akaike information criterion as weights.10 Finally, blue circles define the size of reform gaps relative to a hypothetical country with EU average per capita income. The gap from our initial specification (green diamonds) is generally close to the average gap (red diamonds) for most pillars and countries, except several cases, but reform gaps relative to the EU average are in most cases larger.

Figure 3.Reform Gaps: Comparing Structural Reform Indicators to Various Benchmarks

The vertical axis depicts the standard deviation from the trend of the generic benchmark. A negative number implies that the country lags (see text). 1–institutions, 2–infrastructure, 3–macroeconomic environment, 4–health and primary education, 5–higher education and training, 6–goods market efficiency, 7–labor market efficiency, 8–financial market development, 9–technological readiness, 11–business sophistication, and 12–innovation.

Tables 23 transform visually the gaps between the structural indicators of a country and its two generic comparators into reform heat maps: differences compared to a generic country with a 40 percent higher per capita income, and gaps relative to the EU average.11 We define a gap to be very large if it is smaller than –1.65; large if it is between –0.5 and –1.65; medium if it is between –0.5 and 0.5; and low if it is above 0.5 standard deviations. While the thresholds involve some judgment, the interpretation is intuitive if the distribution of gaps approximates a normal distribution.12 For example, a gap of –0.5 means that the country is performing worse than about 70 percent of the sample (assuming a specific income level). A gap of –1.65 implies that the country is below the 5th percentile. A gap of zero means the country is performing as well as about half of the countries in the sample. While a positive gap implies that a country is on an income-adjusted basis in the better half of the sample, it does not mean that there is no need for reform.

Table 2.Reform Needs Based On Comparing Structural Reform Indicators in Emerging ECA to a Generic Country with 40 Percent Higher per Capita Income
InstitutionsInfra-structureMacro-economic environmentHealth and primary educationHigher education and trainingGoods market efficiencyLabor market efficiencyFinancial market developmentTechnological readinessBusiness sophisticationInnovation
AlbaniaLARGELARGELARGELOWLOWMEDIUMMEDIUMLARGELARGELARGELARGE
ArmeniaMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMLARGELARGE
AzerbaijanMEDIUMMEDIUMLOWLARGELARGEMEDIUMLOWLARGEMEDIUMlargeMEDIUM
Bosnia and HerzegovinaLARGEVERY LARGEMEDIUMLOWLARGEVERY LARGEVERY LARGELARGELARGEVERY LARGELARGE
BulgariaLARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLOWLARGELARGE
CroatiaLARGEMEDIUMLARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMLARGELARGE
Czech RepublicLARGELARGELOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGE
EstoniaLOWMEDIUMLOWMEDIUMLOWLOWLOWLOWMEDIUMLARGEMEDIUM
GeorgiaLOWLOWMEDIUMLOWMEDIUMMEDIUMLOWMEDIUMMEDIUMLARGELARGE
HungaryLARGEMEDIUMMEDIUMMEDIUMLARGELARGEMEDIUMLARGELARGEVERY LARGELARGE
KazakhstanMEDIUMMEDIUMLOWLARGEMEDIUMMEDIUMLOWLARGELARGELARGELARGE
Kyrgyz RepublicMEDIUMMEDIUMMEDIUMLOWLOWLOWMEDIUMMEDIUMLOWMEDIUMLARGE
LatviaMEDIUMLARGELOWMEDIUMMEDIUMMEDIUMLOWMEDIUMLOWLARGELARGE
LithuaniaLARGEMEDIUMLOWMEDIUMLOWMEDIUMMEDIUMMEDIUMLOWMEDIUMMEDIUM
MacedoniaMEDIUMMEDIUMMEDIUMMEDIUMLOWLOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUM
MoldovaLARGELOWMEDIUMlowLOWMEDIUMMEDIUMLARGELOWLARGELARGE
MontenegroMEDIUMMEDIUMMEDIUMLOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGELARGE
PolandLARGELARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMMEDIUMLARGELARGE
RomaniaLARGELARGELOWMEDIUMMEDIUMLARGEMEDIUMMEDIUMMEDIUMLARGELARGE
Russian FederationLARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMLARGELARGELARGELARGE
SerbiaLARGEMEDIUMLARGEMEDIUMMEDIUMVERY LARGELARGELARGELOWVERY LARGELARGE
Slovak RepublicVERY LARGELARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMLARGELARGELARGE
SloveniaLARGELARGELARGEMEDIUMMEDIUMLARGELARGEVERY LARGEMEDIUMLARGELARGE
TajikistanLOWMEDIUMMEDIUMLOWLOWMEDIUMLOWMEDIUMmediumLOWLOW
TurkeyLARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMVERY LARGEMEDIUMLARGELARGELARGE
UkraineLARGELOWVERY LARGELOWLOWLARGEMEDIUMLARGEMEDIUMMEDIUMMEDIUM
Table 3.Reform Needs Based on Comparing Structural Reform Indicators s in Emerging ECA Relative to EU Average Income
InstitutionsInfra-structureMacro-economic environmentHealth and primary educationHigher education and trainingGoods market efficiencyLabor market efficiencyFinancial market developmentTechnological readinessBusiness sophisticationInnovation
AlbaniaLARGELARGELARGEMEDIUMMEDIUMLARGELARGEVERY LARGELARGELARGEVERY LARGE
ArmeniaLARGELARGEMEDIUMLARGELARGELARGEMEDIUMLARGELARGELARGELARGE
AzerbaijanLARGELARGELOWLARGELARGELARGEMEDIUMLARGELARGELARGELARGE
Bosnia and HerzegovinaVERY LARGEVERY LARGELARGEMEDIUMLARGEVERY LARGEVERY LARGELARGELARGEVERY LARGEVERY LARGE
BulgariaLARGELARGEMEDIUMMEDIUMLARGELARGEMEDIUMLARGEMEDIUMLARGELARGE
CroatiaLARGEMEDIUMLARGEMEDIUMLARGELARGELARGELARGEMEDIUMLARGELARGE
Czech RepublicLARGEMEDIUMLOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUM
EstoniaMEDIUMMEDIUMLOWMEDIUMMEDIUMMEDIUMLOWMEDIUMMEDIUMLARGEMEDIUM
GeorgiaMEDIUMLARGEMEDIUMMEDIUMLARGELARGEMEDIUMLARGELARGEVERY LARGEVERY LARGE
HungaryLARGELARGEMEDIUMLARGELARGELARGELARGELARGELARGELARGELARGE
KazakhstanLARGELARGEMEDIUMLARGELARGELARGELOWLARGELARGELARGELARGE
Kyrgyz RepublicVERY LARGEVERY LARGELARGELARGELARGELARGELARGELARGEVERY LARGEVERY LARGEVERY LARGE
LatviaLARGELARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGELARGE
LithuaniaLARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMMEDIUMMEDIUM
MacedoniaLARGELARGEMEDIUMLARGEMEDIUMMEDIUMLARGEMEDIUMLARGELARGELARGE
MoldovaVERY LARGELARGEMEDIUMLARGELARGELARGELARGELARGELARGEVERY LARGEVERY LARGE
MontenegroLARGELARGELARGEMEDIUMLARGELARGEMEDIUMMEDIUMLARGELARGELARGE
PolandLARGELARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMMEDIUMLARGELARGE
RomaniaLARGELARGEMEDIUMLARGELARGELARGELARGEMEDIUMMEDIUMLARGELARGE
Russian FederationLARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMLARGELARGELARGELARGE
SerbiaVERY LARGELARGEVERY LARGEMEDIUMLARGEVERY LARGELARGEVERY LARGELARGEVERY LARGELARGE
Slovak RepublicLARGELARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMMEDIUMLARGELARGE
SloveniaLARGEMEDIUMLARGEMEDIUMMEDIUMMEDIUMLARGEVERY LARGEMEDIUMLARGEMEDIUM
TajikistanLARGEVERY LARGELARGELARGELARGELARGEMEDIUMLARGEVERY LARGELARGELARGE
TurkeyLARGELARGEMEDIUMLARGELARGEMEDIUMVERY LARGELARGELARGELARGELARGE
UkraineVERY LARGELARGEVERY LARGEMEDIUMMEDIUMLARGEMEDIUMVERY LARGELARGELARGELARGE
Source: Authors’ calculationsSee text for thresholds.
Source: Authors’ calculationsSee text for thresholds.

While the rule on differences between structural indicators could be formulated directly in terms of percentiles, showing distances from the mean has two advantages: (1) It is likely to provide a more objective pattern of the gaps if the distribution is skewed,1 and (2) the percentiles are harder to interpret when the gaps are relative to an absolute benchmark, such as the EU average.

To gauge the robustness of each gap, we look at the spread between specifications. A gap is interpreted as robust if its maximum or minimum distance from all eight specifications is no more than one threshold apart from the gap based on the main specification. For example, if the main gap is 0.2, the minimum is –0.4, and the maximum is 0.7, then the gap is considered robust, because only one threshold is crossed (0.5). If the main gap is 0.2, the minimum is –0.4, but the maximum is 1.7, then the gap is not robust, because two thresholds are crossed (0.5 and 1.65). Non-robust gaps are marked on the heat maps by small letters, robust gaps are capitalized (e.g., LARGE—robust, large—not robust). Almost all of our main specification gaps are robust.2

Tables 23 show that despite some relatively common reform challenges, ECA countries are quite heterogeneous. Assuming countries in the region want to raise per capita income by 40 percent over the next 7–10 years, the largest needs are in the areas of institutions, financial market development, infrastructure, product and labor market efficiency, and areas related to innovation. While reform needs are generally large, a few selected countries have a small number of reform areas in “red”, including some low-income countries. In addition to reform progress, this also reflects that the methodology explicitly takes into account the level of income, thus countries with a lower per capita income are not expected to score as well on structural indicators as countries with a higher level of income. However, other factors may play a role as well, such as measurement errors in the perception-based structural indicators and omitted variables.

Compared to countries where income is equivalent to that of an average EU country, most ECA countries have ample space for improvement (Table 3). The choice of the benchmark obviously affects the size of the difference in structural indicators. The higher the income associated with the benchmark, the larger the gap tends to be.

IV. Reform Gaps Disaggregated

Differences between structural indicators of a country and a hypothetical country with higher per capita income can also be calculated at a more disaggregated level. GCR reports sub-pillar data for over 100 indicators. To illustrate the most common challenges, Table 4 shows for each pillar the sub pillars (areas) with the largest differences (gaps). Based on our initial specification, the table lists areas that in at least three countries were among the top 10 largest gaps.3

Table 4.Largest Reform Gaps at a More Disaggregated Level
PillarSubpillarsPillarSubpillars
Institutions (53)Efficiency in challenging regulations, settling disputes Diversion of public funds Judicial independence Property rights protection (incl. intellectual, minority shareholders) Favoritism in gov't decisions, bribes Transparency of policymaking Strength of reporting standardsInfrastructure (25)Quality of air transport

Quality of roads

Quality of ports, railroads
Macroeconomic environment (4)Health and primary education (11)Business impact of HIV/AIDS, tuberculosis, malaria
Higher education and training (15)Extent of staff training

Quality of management schools
Goods market efficiency (31)Time, procedures required to start business, Intensity of local competition, anti-monopoly policy

Buyer sophistication

Prevalance of foreign ownership, impact of rules on FDI
Labor market efficiency (32)Country capacity to attract talent

Capacity to retain talent

Redundancy costs, flexibility of wage determination

Reliance on professional management
Financial market development (21)Soundness of banks

Financing through local equity mkts

Availability, affordability of financial services

Regulation of security exchanges
Technological readiness (15)Firm-level technology absorption

Internet bandwidth, subscriptions

Availability of latest technologies
Business sophistication (39)Local supplier quality, quantity

Extent of marketing

Production process sophistication

State of cluster development Willingness to delegate authority

Value chain breadth Nature of competitive advantage
R&D Innovation (26)Availability of scientists and engineers

PCT patent applications Quality of scientific research institutions, collaboration
Subpillars, which are among the 10 top largest gaps at least in three countries.Numbers in brackets – total number of subpillars among the 10 top largest gaps in a given pillar.Source: author’s calculations.
Subpillars, which are among the 10 top largest gaps at least in three countries.Numbers in brackets – total number of subpillars among the 10 top largest gaps in a given pillar.Source: author’s calculations.

Results at a more granular level broadly reflect those of the aggregate level. In 10 out of 11 main GCR pillars, there are three or more large reform gaps at a more granular level in emerging ECA countries, which underlines the breadth and heterogeneity of reform needs in the region. Challenges in a few areas, however, do seem more widespread. The list of reforms at a more granular level is longer in the areas of institutions, business sophistication, goods market efficiency, labor market efficiency, financial market development. The granular analysis identifies at least three key areas in infrastructure, technological readiness and R&D innovation.

V. Robustness Checks

To test further for the robustness of the main results, we re-estimated the relationship for the whole time period separately for each income group and for a different set of reform indicators: the World Bank Doing Business indicators. In addition, we use a different methodology.

V.1 Reform Gaps over Time and across Income Groups

When the eight specifications described are re-estimated using the average for the 2006–14 period, the findings are consistent with previous results: income per capita is positively correlated with structural indicators for each pillar, and again, being a resource-rich country was associated with weaker performance across most pillars, except for the macroeconomic environment. Results for the ECA region are somewhat different for two pillars, infrastructure and technological readiness. For the whole time period, the ECA dummy is no longer positive and significant, pointing to some improvement in these two areas in recent years.

When re-estimating our specifications separately for four different income groups, based on the World Bank classifications of high-income, upper-middle-income, lower-middle-income, and low-income countries, in most cases the results are consistent with our main findings. Even within the more homogenous income groups and with much smaller samples, income per capita remains positively correlated with structural reform indicators. However, we find a statistically significant relationship between income per capita and the labor market pillar only for high-income countries and the sample as a whole.

V.2 Reform Gaps with Doing Business Indicators

So far, the analysis has used GCR indicators as examples. These indicators are based on a survey of business executives on a number of topics. They fall into the category of perception-based indicators, which aim at capturing the views of relevant groups, for example in terms of the quality of various institutions and policies in a given country. The advantage of perception-based indicators is that they capture the views of those who benefit from enhanced legislation and better rules and they not only focus on the existence of a law but also on the quality of its implementation. Disadvantages may be related to the sampling design, a possible sample selection bias, and room for interpretation in the formulation of the questions. Perception-based measures are also often scaled in a somewhat arbitrary way or units that are difficult to interpret, and such a scale can sometimes be unclear to respondents. It may be difficult to link the results of the assessment to particular policy interventions.

Indicators based on primary data, which are considered fact based, focus on the existence of specific laws, regulations, or rules “on the books”. Fact-based indicators tend to involve more clarity in documenting whether a country has a regulation in place in a certain area or certain types of regulatory institutions. Furthermore, these indicators often reflect existing legislation, which makes these indicators “actionable” as policy makers can change laws. However, indicators based on primary data also have a number of drawbacks that one needs to take into account. Laws may not be observed or effective and there could be laws that potentially conflict with each other. The connection to outcomes may be challenging to establish. In sum, both indicators have their advantages and shortcomings, and therefore, a detailed analysis of a particular country should include a comparison of different types of indicators.

While testing a large number of alternative indicators is beyond the scope of this paper, as an example we apply the same methodology to estimate reform gaps using a few selected World Bank Doing Business (DB) indicators4, an example of an indicator based on primary data, and compare the DB and the GCR reform gaps. While the scope of these indicators and the methodology differ, there is also substantial overlap in reform areas.

Income per capita remains positively correlated with the DB indicators. Figure 4 suggests that at the aggregate level gaps based on GCR and DB indicators (overall GCR score vs. DB score) are positively correlated. This relationship holds not only at the aggregate level but also for sub-indicators that look at relatively similar areas—for example, for the DB getting credit indicator and the GCR financial markets development indicator, or the DB paying taxes and the GCR total tax rate and incentives of taxation to invest indicators, even when controlling for overall GCR and DB scores. While the two sets of indicators are positively correlated, the R2 ranges from 0.3 to 0.5, which is relatively low, so in some cases they may give a different signal.5 In 2015 in emerging ECA, countries that show up more favorably in the GCR than the DB indicators in these areas are Tajikistan, Azerbaijan, Kazakhstan, Ukraine, the Russian Federation, and Turkey. Assessed less favorably are Georgia, Estonia, Macedonia, and Armenia.

Figure 4.GCR and Doing Business Score Gaps Compared

Table 5.GCR and Doing Business Indicator Gaps
DB overall

b/se
DB overall

b/se
DB overall

b/se
DB getting credit

b/se
DB paying taxes

b/se
DB paying taxes

b/se
GCR overall0.49***

(0.08)
0.29***

(0.11)
−0.24**

(0.10)
GCR goods markets efficiency0.49***

(0.08)
0.28**

(0.11)
0.54***

(0.08)
0.23***

(0.07)
GCR financial markets development0.46***

(0.09)
DB overall0.49***

(0.08)
GCR effect of taxation on incentives to invest0.22**

(0.08)
GCR total tax rate0.50***

(0.07
Constant0.15**

(0.08)
0.15**

(0.08)
0.15**

(0.07)
0.09

(0.07)
0.04

(0.08)
0.05

(0.06)
Observations130130130130130130
R20.250.240.280.390.270.57

V.3 Reform Gaps Using Stochastic Frontier Analysis

As an alternative to OLS we also estimate the reform needs using stochastic frontier analysis. Stochastic frontier analysis estimates the productivity frontier – which in our case refers to the country with the highest level of a structural reform indicator given its income – and the distance of other countries to this frontier. Here, the parametric approach assumes a linear relationship between the output (level of structural reform) and the input (log GDP per capita). One difference between the stochastic frontier analysis and OLS is that stochastic frontier analysis estimates an additional error term (called the inefficiency term), which is always non-negative – the distance of a country to the frontier. In OLS the reform gap is the difference between actual and the trend, whereas in stochastic frontier analysis it is the difference between actual and the most productive economy (frontier), and so it is always positive by definition. The stochastic frontier analysis relies on more assumptions than OLS, in particular about the distribution of the inefficiency term, and its estimation procedure is less straightforward.

We employ the parametric version as formulated in Aigner et. al. 1977, where the inefficiency term is distributed half-normally. The regression specification is the same as is for the OLS – the only dependent variable we use is log GDP per capita. The dependent variable the aggregate GCR score.

Figure 5 depicts the reform gaps based on the OLS versus those based on the stochastic frontier analysis. The fit between the two is close (R2 of regression one on another is 0.93), while the ranking of countries is generally preserved. Overall, we expect both methods to yield similar results in terms of country ranking and relative magnitude of the gaps.

Figure 5:Reform Gaps: OLS vs. Stochastic Frontier

Note: Underlying indicator – aggregate GCR score, 2015. Main specification is used both for OLS and stochastic frontier estimations (the only dependent variable is log GDP per capita). For the stochastic frontier estimation, the distribution of the inefficiency term is half-normal.

V.4 Reform Gaps: Additional Alternative Specifications

We run two additional alternative specifications to estimate the structural reform gaps and see if there is a large difference with our main specification. First, instead of actual GDP per capita as in our main specification, we use potential GDP per capita to estimate the reform gaps, reflecting the medium-term nature of the analysis. Potential GDP per capita is estimated for each country using a Hodrick-Prescott filter, and taking into account a three-year ahead forecast. Without taking into account business cycle fluctuations, countries with large negative output gaps may appear to perform deceptively well. However, as reported in Table A.3.1, in 2015 there is little difference between the two measures – the corresponding heat maps are almost identical. Another alternative specification relates to the choice of the benchmark. In the main specification, the benchmark is a 40% increase in income over the next 7-10 years for all countries. This bar may be high for richer countries of the region, while for the poorest countries in the region 40 percent growth over ten years may not be ambitious enough, given the average speed of convergence between rich and poor countries over the last decade. Therefore, in our alternative specification we use country-specific GDP growth targets as implied by a simple growth regression. To obtain the projections, we use results of the regression of average GDP per capita growth during 2005-2015 and log GDP per capita in 2004, estimated for the sample of Emerging ECA countries. A negative coefficient on log GDP per capita suggests that there was convergence between richer and poorer countries in the region during the last ten years. This implies higher GDP growth projections for poorer countries (as high as 64% for Tajikistan, and 63% for Kyrgyz Republic), and lower GDP growth projections for richer countries (as low as 19% for Czech Republic and 17% for Slovenia). Consequently, as depicted in Table A.3.2, the reform needs are higher for poorer countries as compared to the main specification, and they are lower for the richer countries in the region. The within-country ranking of reform gaps does not change though.

VI. Conclusions

On average countries with higher per capita income tend to score better on structural indicators. Empirical evidence suggests that at least in the medium term structural reforms are conducive to increasing potential growth. However, in some cases reverse causality or mutually reinforcing developments cannot be ruled out. Better-off countries may be able to more easily afford to be more advanced in the structural area.

While this paper does not analyze directly the link between structural reforms and growth, it offers a bird’s-eye view of structural reform needs as a first step to a more detailed analysis. As countries strive to increase per capita income, the answers to certain auxiliary questions may provide guidance to policymakers on reform priorities: Assuming a country aspires to increase per capita income by say 40 percent in the next 7-10 years, how does a country compare to a generic country that has already achieved the higher level of income? How does a given country compare to one with the income of an average EU country?

While reform needs are country-specific, the results here suggest that in coming years reform needs in the region are largest in the areas of institutions, financial market development, infrastructure, goods and labor market efficiency, and areas related to innovation. The approach also helps detect more granular reform elements in each of these areas. For example, at a more granular level, there are important reform needs in the labor market related to wage flexibility and the ability to attract and retain talent.

The analysis does not directly link structural reforms to growth. Still, the underlying hypothesis is that if the difference between an indicator in country i and the same indicator in a generic country with 40 percent higher income is particularly large, then it is likely that closing the difference will be desirable and would bring a certain growth dividend. But it could still be true that closing a smaller difference in one area may be more beneficial than closing a larger gap elsewhere.

As with any analysis, the results depend on (i) how reliable the underlying data are and (ii) how well the underlying regressions characterize the true process. Measurement errors in the data would distort the estimated gaps and may lead to misleading conclusions. Moreover, the estimated gaps could also reflect missing explanatory variables in the regression and the unobserved factors may bias results. Despite these caveats, we see the approach as a first step in a comprehensive analysis of the structural reforms needs—an overview, which needs to be supplemented by further analysis, country knowledge, and judgment.

Appendix 1: Summary Statistics
Table A1.1Structural Reform Indicators: Summary Statistics 1/
countmeansdp10p90
Pillar 1: Institutions1324.070.863.195.46
Pillar 2: Infrastructure1324.041.222.435.71
Pillar 3: Macroeconomic environment1324.790.963.606.15
Pillar 4: Health and primary education1325.520.894.286.44
Pillar 5: Higher education and training1324.261.012.785.59
Pillar 6: Goods market efficiency1324.380.533.785.13
Pillar 7: Labor market efficiency1324.240.563.615.00
Pillar 8: Financial market development1323.950.703.114.99
Pillar 9: Technological readiness1324.091.212.635.91
Pillar 10: Market size1323.911.162.575.45
Pillar 11: Business sophistication1324.070.713.315.28
Pillar 12: Innovation1323.540.852.674.98
GDP per capita, thousands 2010 USD13215.8120.410.8345.41
=1 if from Emerging ECA0.200.400.001.00
=1 if from SSA0.230.420.001.00
=1 if OECD member and HIC as of 20100.230.420.001.00
=1 if resource-rich (IMF 2012)0.300.460.001.00

132 countries for which all data is available.

Source: GCR 2015, authors’ calculations.

132 countries for which all data is available.

Source: GCR 2015, authors’ calculations.
Table: A1.2Global Competiveness Index: Score (1-7 (best))
CountryInstitutionsInfrastructureMacroeconomic environmentHealth and primary educationHigher education and trainingGoods market efficiencyLabor market efficiencyFinancial market developmentTechnological readinessBusiness sophisticationInnovation
Albania3.683.553.965.974.744.343.973.243.403.652.76
Armenia3.783.724.715.354.264.464.303.533.673.653.02
Azerbaijan3.944.156.355.223.904.314.573.334.263.863.33
Bosnia and Herzegovina3.183.084.326.033.773.693.363.343.603.312.79
Bulgaria3.393.994.945.974.484.354.233.984.873.643.11
Croatia3.634.584.195.854.624.053.833.594.653.743.13
Czech Republic4.094.695.976.315.104.634.444.625.434.493.79
Estonia5.034.876.156.345.504.935.004.635.324.264.03
Georgia4.384.194.955.794.004.484.563.873.813.482.71
Hungary3.524.514.945.714.564.294.153.934.603.703.44
Kazakhstan4.164.255.725.374.534.484.903.564.193.793.27
Kyrgyz Republic3.292.844.625.304.094.234.063.443.273.412.67
Latvia4.184.475.566.185.054.644.724.395.294.063.33
Lithuania4.124.685.566.195.354.644.353.995.634.323.73
Macedonia4.143.775.095.614.794.654.074.094.153.873.38
Moldova3.203.694.865.394.094.064.073.284.393.292.56
Montenegro3.893.984.626.214.584.304.184.264.333.623.28
Poland4.074.305.116.155.054.514.114.264.784.093.32
Romania3.663.615.445.494.554.284.134.054.633.713.24
Russian Federation3.464.815.295.944.964.164.403.534.223.793.29
Serbia3.243.873.605.874.273.743.723.234.473.142.90
Slovak Republic3.434.285.216.014.624.433.904.414.644.073.29
Slovenia3.934.794.456.445.414.504.002.855.144.153.83
Tajikistan4.102.934.645.614.124.124.423.382.813.803.32
Turkey3.844.434.755.694.584.533.463.934.084.073.35
Ukraine3.074.073.126.065.034.024.333.183.453.703.41
Source: GCR 2015-16.
Source: GCR 2015-16.
Appendix 2: Regression Results: Different Specifications*
Table A2.1Regression Results for Pillar 1, Institutions
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.07***

(0.01)
0.07***

(0.01)
0.07***

(0.01)
0.08***

(0.01)
emerg_ECA−0.08***

(0.02)
−0.07***

(0.02)
−0.10**

(0.05)
−0.02

(0.02)
−0.02

(0.02)
−0.00

(0.04)
oecd_hic0.04

(0.04)
−0.02

(0.04)
−0.08**

(0.03)
−0.08**

(0.04)
RR_dummy−0.04

(0.02)
−0.03

(0.02)
−0.03

(0.02)
−0.03

(0.02)
−0.04*

(0.02)
−0.04*

(0.02)
regionl−0.01

(0.07)
0.06

(0.08)
region2−0.01

(0.05)
0.02

(0.05)
region3−0.02

(0.06)
0.05

(0.06)
region40.01

(0.06)
0.07

(0.06)
region5−0.16***

(0.06)
−0.04

(0.06)
region60.01

(0.06)
0.08

(0.05)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.47***

(0.02)
1.50***

(0.02)
1.50***

(0.02)
1.52***

(0.06)
1.50***

(0.01)
1.51***

(0.02)
1.51***

(0.02)
1.47***

(0.06)
Observations132132132131132132132131
R-squared0.450.490.500.610.570.580.600.66
Adjusted R20.450.480.480.580.570.570.580.62
AIC−179.4−185.5−184.8−205.1−210.4−208.9−211.3−218
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.2Regression Results for Pillar 2, Infrastructure
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.15***

(0.01)
0.15***

(0.01)
0.15***

(0.01)
0.14***

(0.01)
emerg_ECA0.01

(0.02)
0.01

(0.02)
−0.03

(0.03)
0.08***

(0.02)
0.08***

(0.02)
0.03

(0.04)
oecd_hic−0.01

(0.03)
−0.03

(0.03)
−0.09**

(0.04)
−0.07*

(0.04)
RR_dummy−0.07***

(0.02)
−0.07***

(0.03)
−0.06**

(0.03)
−0.05*

(0.03)
−0.07**

(0.03)
−0.06**

(0.03)
regionl−0.07

(0.06)
−0.05

(0.08)
region2−0.02

(0.02)
0.04*

(0.02)
region3−0.05

(0.04)
0.03

(0.05)
region4−0.10**

(0.04)
−0.11**

(0.06)
region5−0.10**

(0.04)
0.01

(0.05)
region60.05

(0.03)
0.10**

(0.04)
gdpcap_thous0.02***

(0.00)
0.02***

(0.00)
0.02***

(0.00)
0.02***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.31***

(0.02)
1.33***

(0.02)
1.33***

(0.02)
1.41***

(0.04)
1.37***

(0.02)
1.37***

(0.03)
1.37***

(0.03)
1.42***

(0.05)
Observations132132132131132132132131
R-squared0.800.810.810.840.690.710.720.77
Adjusted R20.800.810.810.830.680.710.710.75
AIC−199.7−205.8−204−208.5−138.4−146.7−148.7−161.6
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.3Regression Results for Pillar 3, Macroeconomic Environment
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.05***

(0.01)
0.06***

(0.01)
0.06***

(0.01)
0.07***

(0.01)
emerg_ECA0.04

(0.03)
0.04

(0.03)
0.09

(0.06)
0.08***

(0.03)
0.08***

(0.03)
0.19***

(0.07)
oecd_hic−0.03

(0.04)
−0.01

(0.05)
−0.11**

(0.05)
−0.06

(0.05)
RR_dummy0.07**

(0.03)
0.06**

(0.03)
0.07**

(0.03)
0.07**

(0.03)
0.05*

(0.03)
0.06*

(0.03)
regionl0.13

(0.11)
0.20*

(0.12)
region20.03

(0.07)
0.05

(0.07)
region30.06

(0.09)
0.13

(0.10)
region40.11

(0.09)
0.16

(0.10)
region50.06

(0.09)
0.17*

(0.10)
region60.17*

(0.09)
0.24**

(0.10)
gdpcap_thous0.00***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00

(0.00)
−0.00

(0.00)
−0.00***

(0.00)
−0.00**

(0.00)
Constant1.64***

(0.02)
1.61***

(0.03)
1.61***

(0.03)
1.50***

(0.09)
1.68***

(0.02)
1.63***

(0.02)
1.63***

(0.02)
1.45***

(0.11)
Observations132132132131132132132131
R-squared0.210.250.250.300.200.260.290.35
Adjusted R20.200.230.230.240.190.240.260.29
AIC−121.1−123.5−122.1−116.8−117.5−123.6−126.6−124.8
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.4Regression Results for Pillar 4, Health and Primary Education
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.07***

(0.01)
0.07***

(0.01)
0.07***

(0.01)
0.05***

(0.01)
emerg_ECA0.05***

(0.01)
0.05***

(0.01)
−0.00

(0.01)
0.08***

(0.02)
0.09***

(0.02)
0.01

(0.01)
oecd_hic−0.02

(0.02)
−0.02

(0.02)
−0.06**

(0.02)
−0.03*

(0.02)
RR_dummy−0.07***

(0.02)
−0.07***

(0.02)
−0.06***

(0.02)
−0.06***

(0.02)
−0.07***

(0.02)
−0.06***

(0.02)
regionl−0.01

(0.05)
−0.00

(0.06)
region20.02

(0.03)
0.04

(0.03)
region30.02

(0.03)
0.05

(0.03)
region4−0.15***

(0.05)
−0.15***

(0.04)
region5−0.04

(0.03)
−0.00

(0.04)
region60.04

(0.03)
0.06**

(0.03)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.73***

(0.02)
1.74***

(0.02)
1.74***

(0.02)
1.83***

(0.04)
1.76***

(0.02)
1.77***

(0.02)
1.77***

(0.02)
1.83***

(0.03)
Observations132132132131132132132131
R-squared0.560.630.630.750.440.540.550.73
Adjusted R20.550.620.620.730.430.520.530.70
AIC−232.4−250.8−249.6−286.5−199.3−220.9−221.6−273.8
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1*Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1*Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.5Regression Results for Pillar 5, Higher Education and Training
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.12***

(0.01)
0.12***

(0.00)
0.11***

(0.01)
0.09***

(0.01)
emerg_ECA0.08***

(0.02)
0.08***

(0.02)
0.03

(0.02)
0.14***

(0.02)
0.14***

(0.02)
0.07**

(0.03)
oecd_hic0.01

(0.02)
−0.00

(0.02)
−0.04

(0.03)
−0.03

(0.02)
RR_dummy−0.04**

(0.02)
−0.04*

(0.02)
−0.03

(0.02)
−0.03

(0.02)
−0.04

(0.03)
−0.02

(0.02)
regionl−0.09*

(0.05)
−0.08

(0.06)
region2−0.02

(0.02)
0.02

(0.02)
region3−0.09**

(0.03)
−0.04

(0.04)
region4−0.16***

(0.04)
−0.17***

(0.05)
region5−0.08**

(0.03)
−0.01

(0.04)
region60.01

(0.03)
0.05*

(0.03)
gdpcap_thous0.02***

(0.00)
0.02***

(0.00)
0.02***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.42***

(0.01)
1.42***

(0.01)
1.42***

(0.01)
1.54***

(0.04)
1.47***

(0.02)
1.46***

(0.02)
1.46***

(0.02)
1.54***

(0.04)
Observations132132132131132132132131
R-squared0.770.810.810.850.630.710.710.80
Adjusted R20.770.800.800.830.620.700.700.78
AIC−239.8−259.3−257.6−272.7−174.6−203.1−202.2−235.2
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central, Eastern Europe, Central Asia and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central, Eastern Europe, Central Asia and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.6Regression Results for Pillar 6, Goods Market Efficiency
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.05***

(0.00)
0.04***

(0.00)
0.04***

(0.01)
0.05***

(0.01)
emerg_ECA−0.01

(0.01)
−0.01

(0.01)
−0.03

(0.02)
0.01

(0.01)
0.02

(0.01)
0.02

(0.02)
oecd_hic0.01

(0.02)
−0.02

(0.02)
−0.05**

(0.02)
−0.05**

(0.02)
RR_dummy−0.03*

(0.02)
−0.03

(0.02)
−0.02

(0.02)
−0.02

(0.02)
−0.03*

(0.02)
−0.03*

(0.02)
regionl−0.04*

(0.02)
−0.01

(0.03)
region2−0.03**

(0.01)
−0.02

(0.01)
region3−0.05**

(0.03)
−0.01

(0.03)
region4−0.04

(0.03)
−0.02

(0.03)
region5−0.11***

(0.03)
−0.04

(0.03)
region60.01

(0.02)
0.05**

(0.02)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.59***

(0.01)
1.60***

(0.01)
1.60***

(0.01)
1.65***

(0.02)
1.61***

(0.01)
1.61***

(0.01)
1.61***

(0.01)
1.63***

(0.03)
Observations132132132131132132132131
R-squared0.480.500.500.590.520.540.550.61
Adjusted R20.480.480.480.550.510.520.540.57
AIC−318.3−318.8−316.9−327−327.3−327.6−330.4−332.5
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.7Regression Results for Pillar 7, Labor Market Efficiency
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.03***

(0.01)
0.03***

(0.01)
0.02*

(0.01)
0.03***

(0.01)
emerg_ECA−0.01

(0.02)
−0.00

(0.02)
−0.04

(0.03)
0.02

(0.02)
0.02

(0.02)
0.03

(0.03)
oecd_hic0.06*

(0.03)
−0.01

(0.03)
−0.02

(0.03)
−0.06*

(0.03)
RR_dummy−0.01

(0.02)
0.00

(0.02)
0.00

(0.02)
−0.00

(0.02)
−0.00

(0.02)
−0.00

(0.02)
regionl−0.19***

(0.05)
−0.12**

(0.05)
region2−0.11***

(0.02)
−0.10***

(0.02)
region3−0.21***

(0.04)
−0.16***

(0.03)
region4−0.11***

(0.03)
−0.05

(0.04)
region5−0.23***

(0.03)
−0.15***

(0.04)
region6−0.09***

(0.03)
−0.04

(0.03)
gdpcap_thous0.00***

(0.00)
0.00***

(0.00)
0.00***

(0.00)
0.01***

(0.00)
gdp_sq−0.00

(0.00)
−0.00

(0.00)
−0.00

(0.00)
−0.00*

(0.00)
Constant1.60***

(0.01)
1.60***

(0.02)
1.60***

(0.02)
1.75***

(0.03)
1.60***

(0.01)
1.60***

(0.02)
1.60***

(0.02)
1.68***

(0.04)
Observations132132132131132132132131
R-squared0.150.160.180.370.310.310.310.48
Adjusted R20.150.140.160.320.300.290.290.43
AIC−236−232.4−235−254.3−260.6−257.2−255.8−276.7
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.8Regression Results for Pillar 8, Financial Market Development
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.06***

(0.01)
0.06***

(0.01)
0.06***

(0.01)
0.08***

(0.01)
emerg_ECA−0.05***

(0.02)
−0.06***

(0.02)
0.03

(0.06)
−0.02

(0.02)
−0.02

(0.02)
0.11**

(0.06)
oecd_hic−0.04

(0.04)
−0.02

(0.05)
−0.10**

(0.04)
−0.05

(0.05)
RR_dummy−0.03*

(0.02)
−0.04**

(0.02)
−0.04*

(0.02)
−0.03

(0.02)
−0.05**

(0.02)
−0.05**

(0.02)
regionl−0.05

(0.06)
−0.01

(0.06)
region2−0.20***

(0.03)
−0.18***

(0.02)
region3−0.15***

(0.05)
−0.09*

(0.05)
region4−0.07

(0.06)
−0.05

(0.06)
region5−0.11**

(0.05)
−0.01

(0.05)
region6−0.05

(0.05)
0.01

(0.05)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00**

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.48***

(0.01)
1.50***

(0.02)
1.50***

(0.02)
1.58***

(0.05)
1.51***

(0.01)
1.53***

(0.02)
1.52***

(0.02)
1.55***

(0.05)
Observations132132132131132132132131
R-squared0.380.420.420.510.360.380.410.50
Adjusted R20.380.400.410.470.350.360.380.46
AIC−202.6−205.7−205.5−211.3−196.4−194.9−199.5−208.6
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.9Regression Results for Pillar 9, Technological Readiness
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.15***

(0.00)
0.14***

(0.00)
0.13***

(0.01)
0.13***

(0.01)
emerg_ECA0.05***

(0.02)
0.06***

(0.02)
0.01

(0.03)
0.13***

(0.02)
0.13***

(0.02)
0.09**

(0.03)
oecd_hic0.06***

(0.02)
0.03

(0.03)
−0.04

(0.03)
−0.02

(0.03)
RR_dummy−0.07***

(0.02)
−0.05***

(0.02)
−0.05***

(0.02)
−0.05**

(0.02)
−0.06***

(0.02)
−0.05***

(0.02)
regionl−0.11***

(0.03)
−0.07

(0.04)
region20.01

(0.01)
0.06***

(0.01)
region3−0.04

(0.03)
0.04

(0.04)
region4−0.06*

(0.04)
−0.05

(0.04)
region5−0.07**

(0.03)
0.05

(0.04)
region60.03

(0.02)
0.10***

(0.03)
gdpcap_thous0.02***

(0.00)
0.02***

(0.00)
0.02***

(0.00)
0.02***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.32***

(0.01)
1.34***

(0.01)
1.34***

(0.01)
1.40***

(0.03)
1.38***

(0.02)
1.38***

(0.02)
1.38***

(0.02)
1.38***

(0.04)
Observations132132132131132132132131
R-squared0.860.890.900.910.780.840.840.87
Adjusted R20.860.890.890.900.770.840.840.86
AIC−264.1−288.6−293.9−298.8−197.3−236.7−235.9−248.7
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.10Regression Results for Pillar 11, Business Sophistication
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.07***

(0.00)
0.07***

(0.00)
0.06***

(0.01)
0.06***

(0.01)
emerg_ECA−0.08***

(0.01)
−0.07***

(0.01)
−0.10***

(0.03)
−0.03***

(0.01)
−0.03***

(0.01)
−0.04

(0.03)
oecd_hic0.05**

(0.02)
0.02

(0.02)
−0.01

(0.02)
−0.01

(0.02)
RR_dummy−0.05***

(0.02)
−0.04**

(0.02)
−0.03*

(0.02)
−0.04**

(0.02)
−0.04**

(0.02)
−0.03**

(0.02)
regionl−0.04

(0.05)
−0.00

(0.06)
region2−0.02

(0.04)
−0.00

(0.04)
region3−0.06

(0.05)
−0.01

(0.05)
region4−0.06

(0.05)
−0.03

(0.05)
region5−0.10**

(0.05)
−0.02

(0.05)
region6−0.01

(0.05)
0.04

(0.05)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.48***

(0.01)
1.51***

(0.01)
1.51***

(0.01)
1.57***

(0.05)
1.50***

(0.01)
1.52***

(0.01)
1.52***

(0.01)
1.54***

(0.05)
Observations132132132131132132132131
R-squared0.630.690.710.730.700.730.730.74
Adjusted R20.620.690.700.710.700.720.720.72
AIC−277.6−299.6−303.9−301.8−306.1−312.3−310.6−303.9
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Table A2.11Regression Results for Pillar 12, Innovation
VARIABLES(1)(2)(3)(4)(5)(6)(7)(8)
Ingdp_pc0.09***

(0.01)
0.08***

(0.01)
0.06***

(0.01)
0.06***

(0.01)
emerg_ECA−0.09***

(0.02)
−0.07***

(0.02)
−0.11**

(0.04)
−0.03**

(0.02)
−0.03**

(0.02)
−0.02

(0.04)
oecd_hic0.15***

(0.03)
0.10***

(0.04)
0.06*

(0.03)
0.05

(0.04)
RR_dummy−0.06***

(0.02)
−0.03

(0.02)
−0.03

(0.02)
−0.05**

(0.02)
−0.04**

(0.02)
−0.03*

(0.02)
regionl−0.07

(0.08)
−0.01

(0.08)
region2−0.05

(0.06)
−0.03

(0.06)
region3−0.10

(0.08)
−0.04

(0.08)
region4−0.07

(0.08)
−0.02

(0.08)
region5−0.16**

(0.07)
−0.06

(0.07)
region6−0.00

(0.07)
0.06

(0.07)
gdpcap_thous0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
0.01***

(0.00)
gdp_sq−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
−0.00***

(0.00)
Constant1.33***

(0.02)
1.38***

(0.02)
1.38***

(0.02)
1.46***

(0.07)
1.36***

(0.01)
1.38***

(0.02)
1.39***

(0.02)
1.40***

(0.08)
Observations132132132131132132132131
R-squared0.540.610.670.720.700.720.730.75
Adjusted R20.540.600.660.700.700.710.720.73
AIC−185.9−200.7−223−229.4−239.1−243.6−244.6−244
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Note: emerg_ECA is a dummy for countries in Central and Eastern Europe, Central Asia, and the Caucasus; oecd_hic is a dummy for members of the OECD; RR_dummy is a dummy for resource-rich countries; and regional dummies are dummies for six geographical regions in the following order: South Asia, Europe and CIS, MENA, Sub-Saharan Africa, Latin America, and Asia and the Pacific, relative to the seventh region, North America.
Appendix 3. Differences in Structural Reform Indicators in Emerging ECA: Alternative Specifications
Table A3.1.Heat Map: Differences in Structural Reform Indicators s in Emerging ECA Relative to a Generic Country with 40% Higher Income – Potential Output instead of Actual Output
InstitutionsInfrastructureMacro-economic environmentHealth and primary educationHigher education and trainingGoods market efficiencyLabor market efficiencyFinancial market developmentTechnological readinessBusiness sophisticationInnovation
AlbaniaLARGELARGELARGELOWLOWMEDIUMMEDIUMLARGELARGELARGELARGE
ArmeniaMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMLARGELARGE
AzerbaijanMEDIUMMEDIUMLOWLARGELARGEMEDIUMLOWLARGEMEDIUMlargeMEDIUM
Bosnia and HerzegovinaLARGEVERY LARGEMEDIUMLOWLARGEVERY LARGEVERY LARGELARGELARGEVERY LARGELARGE
BulgariaLARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLOWLARGELARGE
CroatiaLARGEMEDIUMLARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMVERY LARGELARGE
Czech RepublicLARGELARGELOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGE
EstoniaLOWMEDIUMLOWMEDIUMLOWLOWLOWLOWMEDIUMLARGEMEDIUM
GeorgiaLOWLOWMEDIUMLOWMEDIUMMEDIUMLOWMEDIUMMEDIUMLARGELARGE
HungaryLARGEMEDIUMMEDIUMMEDIUMLARGELARGEMEDIUMLARGELARGEVERY LARGELARGE
KazakhstanMEDIUMMEDIUMLOWLARGEMEDIUMMEDIUMLOWLARGELARGELARGELARGE
Kyrgyz RepublicMEDIUMMEDIUMMEDIUMLOWLOWLOWMEDIUMMEDIUMLOWMEDIUMLARGE
LatviaMEDIUMLARGELOWMEDIUMMEDIUMMEDIUMLOWMEDIUMLOWLARGELARGE
LithuaniaLARGEMEDIUMLOWMEDIUMLOWMEDIUMMEDIUMMEDIUMLOWMEDIUMMEDIUM
MacedoniaMEDIUMMEDIUMMEDIUMMEDIUMLOWLOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUM
MoldovaLARGELOWMEDIUMlowLOWMEDIUMMEDIUMLARGELOWLARGELARGE
MontenegroMEDIUMMEDIUMMEDIUMLOWMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMLARGELARGE
PolandLARGELARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMMEDIUMLARGELARGE
RomaniaLARGELARGELOWMEDIUMMEDIUMLARGEMEDIUMMEDIUMMEDIUMLARGELARGE
Russian FederationLARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMLARGELARGELARGELARGE
SerbiaLARGEMEDIUMLARGEMEDIUMMEDIUMVERY LARGELARGELARGELOWVERY LARGELARGE
Slovak RepublicVERY LARGELARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMLARGELARGELARGE
SloveniaLARGELARGELARGEMEDIUMMEDIUMLARGELARGEVERY LARGEMEDIUMLARGELARGE
TajikistanLOWMEDIUMMEDIUMLOWLOWMEDIUMLOWMEDIUMMEDIUMLOWLOW
TurkeyLARGEMEDIUMMEDIUMMEDIUMMEDIUMMEDIUMVERY LARGEMEDIUMLARGELARGELARGE
UkraineLARGELOWVERY LARGELOWLOWLARGEMEDIUMLARGEMEDIUMMEDIUMMEDIUM
Table A3.2.Heat Map: Differences in Structural Reform Indicators in Emerging ECA Relative to a Generic Country with Higher Income – Projected 10-year GDP Per Capita Growth
InstitutionsInfrastructureMacro-economic environmentHealth and primary educationHigher education and trainingGoods market efficiencyLabor market efficiencyFinancial market developmentTechnological readinessBusiness sophisticationInnovation
AlbaniaLARGELARGELARGELOWLOWMEDIUMLARGELARGELARGElargeVERY LARGE
ArmeniaMEDIUMMEDIUMMEDIUMmediumlowLOWMEDIUMLARGEMEDIUMLARGELARGE
AzerbaijanMEDIUMMEDIUMLOWLARGELARGEMEDIUMLOWLARGEMEDIUMlargelarge
Bosnia and HerzegovinaVERY LARGEVERY LARGELARGELOWLARGEVERY LARGEVERY LARGELARGELARGEVERY LARGEVERY LARGE
BulgariaVERY LARGElargeMEDIUMLOWMEDIUMMEDIUMMEDIUMMEDIUMLOWVERY LARGEvery large
CroatiaVERY LARGEMEDIUMLARGEmediummediumVERY LARGELARGELARGEMEDIUMVERY LARGEVERY LARGE
Czech RepublicLARGELARGELOWMEDIUMMEDIUMMEDIUMMEDIUMLOWlowMEDIUMLARGE
EstoniaLOWMEDIUMLOWlowLOWLOWLOWLOWLOWLARGEMEDIUM
GeorgiaLOWLOWMEDIUMLOWMEDIUMLOWLOWMEDIUMMEDIUMVERY LARGEVERY LARGE
HungaryVERY LARGELARGEMEDIUMLARGELARGELARGELARGElargeMEDIUMVERY LARGELARGE
KazakhstanMEDIUMLARGELOWLARGEMEDIUMMEDIUMLOWLARGELARGEvery largeLARGE
Kyrgyz RepublicLARGELARGEMEDIUMLOWLOWLOWMEDIUMMEDIUMlowlargeLARGE
LatviaMEDIUMLARGELOWlowlowMEDIUMLOWMEDIUMLOWLARGEVERY LARGE
LithuaniaLARGEMEDIUMLOWlowLOWMEDIUMMEDIUMLARGELOWMEDIUMlarge
MacedoniaMEDIUMlargeMEDIUMMEDIUMLOWLOWMEDIUMMEDIUMMEDIUMlargeMEDIUM
MoldovaLARGELOWLOWlowlowMEDIUMMEDIUMLARGELOWVERY LARGEVERY LARGE
MontenegroLARGElargeMEDIUMLOWMEDIUMLARGEMEDIUMLOWMEDIUMVERY LARGELARGE
PolandLARGELARGEMEDIUMMEDIUMMEDIUMMEDIUMLARGEMEDIUMMEDIUMLARGEvery large
RomaniaLARGEVERY LARGELOWLARGEMEDIUMLARGEMEDIUMMEDIUMMEDIUMVERY LARGELARGE
Russian FederationVERY LARGELOWMEDIUMMEDIUMLOWLARGEMEDIUMLARGELARGEvery largeLARGE
SerbiaVERY LARGElargeVERY LARGELOWMEDIUMVERY LARGELARGEVERY LARGELOWVERY LARGEVERY LARGE
Slovak RepublicVERY LARGEVERY LARGEMEDIUMMEDIUMLARGELARGELARGEMEDIUMLARGELARGEVERY LARGE
SloveniaLARGELARGELARGElowlowLARGELARGEVERY LARGEmediumLARGELARGE
TajikistanLOWmediumLOWLOWLOWMEDIUMLOWMEDIUMLARGELOWLOW
TurkeyLARGEMEDIUMMEDIUMLARGEMEDIUMMEDIUMVERY LARGEMEDIUMLARGELARGELARGE
UkraineVERY LARGELOWVERY LARGELOWLOWLARGEMEDIUMLARGELARGELARGEMEDIUM
Note: Country X is compared with a generic country whose income per capita is the same as that of country X, projected for 2024. The projected GDP per capita growth in country X is implied from a regression of GDP per capita growth in 2005-2015 on GDP per capita in 2004, estimated on a sample of Emerging ECA countries.
Note: Country X is compared with a generic country whose income per capita is the same as that of country X, projected for 2024. The projected GDP per capita growth in country X is implied from a regression of GDP per capita growth in 2005-2015 on GDP per capita in 2004, estimated on a sample of Emerging ECA countries.
Appendix 4. Standard Deviations of Gaps across Specifications
PillarAlbaniaArmeniaAzerbaijanBosnia and HerzeBulgariaCroatiaCzech RepublicEstoniaGeorgiaHungaryKazakhstanKyrgyz RepublicLatvia
Institutions0.290.160.310.130.160.180.250.420.220.310.310.410.18
Infrastructure0.160.240.150.400.310.250.210.230.210.300.190.590.27
Macroenvironment0.300.080.260.100.120.190.160.190.080.210.300.320.15
Health and primary education0.060.440.190.330.350.370.290.290.370.330.160.430.34
Higher education and training0.220.440.420.570.460.460.390.360.490.490.310.740.39
Goods market efficiency0.160.110.180.170.170.190.210.290.120.280.180.320.16
Product market efficiency0.080.080.130.160.100.190.190.320.120.250.130.100.12
Finanacial sector development0.240.110.250.100.140.160.250.280.130.310.240.460.13
Technological readiness0.360.500.190.570.340.440.410.430.480.610.300.840.34
Business sophistication0.440.220.480.180.250.230.270.290.200.290.460.560.27
Innovation0.340.190.440.160.250.260.260.330.140.370.430.440.29
PillarLithuaniaFYR MacedoniaMoldovaMontenegroPolandRomaniaRussiaSerbiaSlovak RepublicSloveniaTajikistanTurkeyUkraine
Institutions0.180.210.190.190.180.170.250.140.220.200.350.170.14
Infrastructure0.240.280.370.310.290.410.150.290.270.200.660.270.26
Macroenvironment0.160.090.120.120.170.130.320.130.180.160.250.160.13
Health and primary education0.340.400.470.320.340.430.110.360.290.300.580.400.34
Higher education and training0.370.370.520.440.390.460.230.470.460.380.820.470.41
Goods market efficiency0.160.140.170.170.170.170.170.180.220.150.350.170.13
Product market efficiency0.140.080.050.100.160.110.100.110.160.120.110.220.08
Finanacial sector development0.150.140.190.140.140.140.240.110.270.260.400.140.12
Technological readiness0.330.430.600.460.420.420.290.380.570.460.930.580.53
Business sophistication0.300.270.230.240.270.250.460.170.260.220.460.290.21
Innovation0.360.290.200.280.290.280.430.190.310.240.310.290.24
Appendix 5. Histograms of Gaps

Figure A5.1Histograms of Gaps: Main Pillars

Source: Own computations. Note: gaps are relative to a peer with the same income per capita.

References

    AcemogluD.J.Simon and J.Robinson. 2004. “Institutions as the Fundamental Cause of Long-Run Growth.” NBER Working Paper 10481. Cambridge, MA: National Bureau of Economic Research.

    • Search Google Scholar
    • Export Citation

    AignerD.C.Lovell and P.Schmidt. 1977. “Formulation and Estimation of Stochastic Frontier Production Function Models.” Journal of Econometrics6: 2137.

    • Crossref
    • Search Google Scholar
    • Export Citation

    AndersonDerekBergljotBarkbuLusineLusinyan and DirkMuir. 2014. “Assessing the Gains from Structural Reforms for Jobs and Growth” in Jobs and Growth: Supporting the European Recovery edited by MartinSchindlerHelgeBergerBas B.Bakker and AntonioSpilimbergo. Washington DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    BabeckyJ. and T.Havranek. 2013. “Structural Reforms and Growth in Transition. A Meta-analysis.” A European Bank for Reconstruction and Development publication. Oxford, UK: Blackwell Publishing.

    • Search Google Scholar
    • Export Citation

    BanerjeeAbhijitEstherDuflo and NancyQian. 2012. “On the Road: Access to Transportation and Economic Growth in China.” NBER Working Paper No. 17897. Cambridge, MA: National Bureau of Economic Research.

    • Search Google Scholar
    • Export Citation

    BarkbuB.etal. 2012. “Fostering Growth in Europe.” IMF Staff Discussion Note European Department. Washington, D.C.: International Monetary Fund.

    • Crossref
    • Search Google Scholar
    • Export Citation

    BesleyTimothy and RobinBurgess. 2004. “Can Labor Regulation Hinder Economic Performance? Evidence from India.” Quarterly Journal of Economics119(1): 91134.

    • Crossref
    • Search Google Scholar
    • Export Citation

    BordonAnne RoseChristianEbeke and KazukoShirono. 2016. “When Do Structural Reforms Work? On the Role of the Business Cycle and Macroeconomic Policies.” IMF Working Paper No. 16/62. Washington, D.C.: International Monetary Fund.

    • Crossref
    • Search Google Scholar
    • Export Citation

    CantonE.etal. 2014. “The Role of Structural Reform and Growth.” ECFIN Economic Brief34: 16.

    Dabla-NorrisE.et al. 2013. “Anchoring Growth: The Importance of Productivity-Enhancing Reforms in Emerging Market and Developing Economies.” IMF Staff Discussion Note. Strategy Policy and Review Department. Washington, DC: International Monetary Fund.

    • Crossref
    • Search Google Scholar
    • Export Citation

    Dabla-NorrisEraGiangHo and AnnetteKyobe. 2016. “Structural Reforms and Productivity Growth in Emerging Market and Developing Economies.” IMF Working Paper No. 16/15. Washington, D.C.: International Monetary Fund.

    • Crossref
    • Search Google Scholar
    • Export Citation

    DuvalRomain and DavideFurcerietal. 2016. “Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies,” Chapter 3IMF World Economic Outlook: pp. 10142. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    ChepteaCristina and DeliaVelculescu. 2014. “A Disaggregated Approach to Prioritizing Structural Reforms for Growth and Employment.” In Jobs and Growth: Supporting the European Recovery edited by MartinSchindlerHelgeBergerBas B.Bakker and AntonioSpilimbergo. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    European Bank for Reconstruction and Development. 2013. “Structural Reform.” Transition Report 2013: Stuck in Transition? pp. 10617. London, UK: EBRD.

    • Search Google Scholar
    • Export Citation

    FabianoSchivardi and ElianaViviano. 2011. “Entry Barriers in Retail Trade.” Economic Journal121(551): 14570.

    GomesSandra; PascalJacquinotMatthiasMohrMassimilianoPisani (2011) Structural Reforms and Macroeconomic Performance in the Euro Area Countries. A Model-Based Assessment European Central Bank Working Paper Series No. 1323. April.

    • Search Google Scholar
    • Export Citation

    HaldenwangChristian von and MaksymIvanyna. 2012. “A Comparative View on the Tax Performance of Developing Countries: Regional Patterns, Non-tax Revenue and Governance.” Economics: The Open-Access Open-Assessment E-Journal6: 201232.

    • Search Google Scholar
    • Export Citation

    IMF. 2015. “Structural Reforms and Macroeconomic Performance: Initial Considerations for the Fund.” IMF Staff Report. Washington, D.C.: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    IMF2016. “Regional Economic Issues Central Eastern and Southeastern Europe. How to Get Back on the Fast Track” International Monetary Fund May 2016.

    • Search Google Scholar
    • Export Citation

    KaufmannD.KraayA. (2007) “Governance Indicators: Where Are We, Where We Be Going?World Bank Policy Research Working Paper No. 4370 November 2007.

    • Search Google Scholar
    • Export Citation

    KrugmanPaul. 2014. “Structural Deformity.” New York Times: November 20 Opinion Pages.

    McAdamPeter and LivioStracca (2015) “Structural Reforms for Inclusive Growth? Empirical Evidence from the OECD” paper presented at the ECB-CBRT conference on “Balanced and Sustainable Growth – Operationalising the G20 Framework” 28 August

    • Search Google Scholar
    • Export Citation

    Organisation for Economic Co-operation and Development. 2013. “Taking Stock of Reform Actions and Identifying Priorities.” Economic Policy Reforms 2013: Going for Growth pp 1555.

    • Search Google Scholar
    • Export Citation

    Organisation for Economic Co-operation and Development. 2015. Going for Growth: Breaking the Vicious Circle. Paris: OECD.

    OstryJonathanAlessandroPrati and AntonioSpilimbergo. 2009. “Structural Reform and Economic Performance in Advanced and Developing Countries.” IMF Occasional Paper No.268. Washington, DC: International Monetary Fund.

    • Search Google Scholar
    • Export Citation

    PhillipsStevenetal. 2013. “The External Balance Assessment (EBA) MethodologyIMF Working Paper 13/272. Washington, D.C.: International Monetary Fund.

    • Crossref
    • Search Google Scholar
    • Export Citation

    VamvakidisAthanasios. 2009. “Convergence in Emerging Europe: Sustainability and Vulnerabilities.” Eastern European Economics47(3): 527

    • Search Google Scholar
    • Export Citation

    VargaJanos and Jan. in’tVeld. 2014. “The Potential Growth Impact of Structural Reforms in the EU. A Benchmarking Exercise.” European Economy Economic Papers 541. Brussels: European Commission.

    • Search Google Scholar
    • Export Citation

    World Bank. 2015. Doing Business 2015: Going Beyond Efficiency. Washington, DC: World Bank.

    World Economic Forum. 2015. Global Competiveness Report 2015–16. Geneva, Switzerland: WEF.

1Author affiliations: Norbert Funke (Assistant Director, ICD) was director of the Joint Vienna Institute (JVI), when the paper was written; Asel Isakova and Maksym Ivanyna are JVI economists. The authors thank the other JVI economists and Andy Berg, Helge Berger, Romain Duval, Edward Gemayel, Mariya Hake, Albert Jaeger, Martin Petri, SeokHyun, Yoon, Johannes Wiegand, and other colleagues in the IMF European, Middle East and Central Asia, and Research departments and country representatives for their very valuable suggestions. Many thanks also to Anne Grant for editorial suggestions.
2See Acemoglu, Simon, and Robinson 2004; Barkbu et al. 2012; Bordon, Ebeke, and Shirono. 2016; Dabla-Norris, Ho, and Kyobe 2013, 2016; Gomes et al. (2011); IMF 2015, 2016; McAdam and Stracca (2015); Ostry, Prati, and Spilimbergo 2009; Vamvakidis 2009. However, some studies recommend caution: Krugman (2014) considers a blanket call for structural reforms to be “intellectually lazy and destructive”; and some reforms may hurt, especially in the short run (see, e.g., Babecky and Havranek 2013).
3This question is often implied in reports like the IMF Article IV reports, when a country’s reform indicators are compared to such benchmarks as regional neighbors, countries with similar income, or other structural characteristics.
4The paper looks at the region in a holistic manner, notwithstanding that countries are at different development stages. For example, the Czech Republic belongs to the “other advanced economies” grouping while Slovakia and Slovenia are both “advanced euro area economies” in the Fund’s World Economic Outlook classification.
5The advantage of the OECD “Going for Growth” project over this study is that it links particular policies to outcomes, while we mostly deal with outcomes directly.
7Because part of market size, the size of the domestic market, is not really a reform area and the result is not statistically significant, this pillar is excluded from the analysis.
8For example, resource-rich countries generally have lower public debt and higher saving rates, which would boost their score on macroeconomic environment, only controlling for their income. Emerging ECA countries generally tend to have relatively higher levels of higher education – a legacy of the past policies. An additional reason to include the dummies is to more precisely estimate the relationship between structural reform stance and income per capita
9Debt as a percent of GDP is an important subcomponent of pillar 3 that measures the macroeconomic environment.
10Specifically, the weight of a gap from specification i is calculated as: ωi=|AIGi|Σi|AIG|.
11Appendix table A1.2 shows the (absolute) scores of the global competitiveness index (1-7(best)) for each pillar.
12On average, for 2015 the reform gaps in 135 countries approximate a normal distribution reasonably well: 5.6 percent of them are very large (smaller than –1.65), and 28 percent are smaller than –0.5. See also Appendix 5.
1While on average for all pillars the distribution of gaps approximates a normal distribution, for some pillars the distribution is skewed to the left: More countries with large negative gaps than large positive ones. See Appendix 4.
2To provide an even fuller picture of the variation of gaps between the eight specifications, Appendix 4 reports the standard deviations of gaps for all specifications. The average deviation from the mean in emerging ECA is 0.28.
3We dropped some subpillars that in our judgment are likely to be unimportant for economic growth (e.g., fixed telephone lines). In some cases, we merged similar subpillars.
4See also World Bank, “Doing-Business – Answers to Frequently Asked Questions” for a description and comparison of Doing Business Indicators.
5Note that in principle the two sets of indicators measure different things: GCR encompasses significantly more reform areas than DB does. That may be driving part of the difference.

Other Resources Citing This Publication