Credibility of Rules and Economic Growth

Credibility of Rules and Economic Growth Evidence from a Worldwide Survey of the Private Sector 1 Aymo Brunetti, Gregory Kisunko, Beatrice Weder University of Saarland and University of Basel World Bank University of Basel Revised version, December 1997 Abstract The purpose of this paper is twofold: the first is to present a new data set on firms perception of institutional uncertainty and the second is to show evidence on the link between such uncertainty and economic performance across countries. The data set is based on a private sector survey which was conducted in 73 countries. From the survey data we derive indicators of the perceived predictability of rules, the reliability of the judiciary, property rights security, political instability and corruption and use them to construct a summary indicator of the credibility of rules. Cross-country regression analysis suggests that this indicator is closely associated with both growth and investment and that this relation is reasonably robust to variations in the sample and to the inclusion of standard control variables. 1 The survey was conducted in the context of the World Development Report 1997. We thank Ajay Chhibber for his support. Financial support for the surveys in the OECD countries was provided by the World Bank Research Support Budget and by the WWZ Förderverein, University of Basel. For valuable comments we thank William Easterly, Markus Kobler, Guy Pfeffermann, Moshe Syrquin and the participants at research seminars held at the World Bank, the University of Bern, the Max-Planck-Institute in Jena, the Swiss National Bank, the Deutsche Bank and the University of the United Nations in Tokio. Three referees provided extremely helpful suggestions and comments. We would also like to thank Michael Geller and Barbara Lévy for able editorial and research assistance. 1

Introduction The general idea that an instable political framework lowers growth is hardly controversial. One would expect that a business environment characterized by incredible rules such as unclear property rights, constant policy surprises and policy reversals, uncertain contract enforcement and high corruption would translate into lower investment and growth. In such an uncertain environment entrepreneurs are reluctant to commit resources especially in projects that are characterized by large sunk cost. 2 This reaction of the private sector not only reduces aggregate investment but also distorts the allocation of resources and reduces economic growth. How the relevant uncertainties can be adequately captured empirically is less clear. Early papers in the recent wave of empirical growth analysis included measures of political instability, proxied for instance by the number of coups and revolutions. 3 Such measures certainly have the advantage of being universally observable and therefore objective but they are also very crude measures of the relevant uncertainties that affect the private entrepreneurs. Subjective measures have been used to proxy for property rights insecurity and corruption by relying on country risk indicators from expert opinions. 4 These second kind of indicators are likely to reflect more closely the concerns of entrepreneurs than the overall measures of political instability. However they are based on the perceptions of country experts and not on those of the local businessmen themselves. In this paper we propose a new measurement approach based on firm level surveys and we construct an indicator of the credibility of rules to be used in growth regressions. The data was sampled in a private sector survey conducted in 73 countries and covering more than 3,800 enterprises. The survey was designed to capture local entrepreneurs views of the predictability of changes in laws and policies, of the reliability of law enforcement, of the impact of discretionary and corrupt 2 3 4 See e.g. Dixit and Pindyck (1994) and Aizenman and Marion (1993). See in particular the influential paper by Barro (1991). Brunetti (1997) provides an updated survey. See Mauro (1995) and Knack and Keefer (1995) 2

bureaucracies and of the danger of policy reversals due to changes in governments. We test this indicator and its various components in standard cross-country growth and investment regressions and find that low credibility of rules is associated with lower rates of investment and growth. The paper is organized as follows. Section 1 discusses in which respect the existing measures of political uncertainty might be incomplete and why we designed a different measurement. Section 2 presents the survey approach, gives information on the surveyed firms and discusses possible problems with selection bias and measurement error. Section 3 explains the construction and gives some regional information on the overall indicator of credibility and its various subindicators. Section 4 presents the empirical approach and section 5 details on the results of growth and investment regressions for the credibility indicator and its components for the 52 countries with reliable data. 1. Why a new approach for measuring policy uncertainty? Figure 1 shows how the existing literature on policies and growth can be grouped and in which respect our approach can be distinguished from the other attempts to measure the degree of policy uncertainty. At the most general level we can distinguish two channels through which policies may influence economic growth: the first focuses on the efficiency of policies and the second on the reliability of policies. The first branch of the literature explains differences in growth with differences in macro-and microeconomic policies. In a large number of studies fiscal policy variables (e.g. Easterly and Rebelo 1993), monetary policy variables (e.g. Fischer 1993), or trade policy variables (e.g. Edwards 1992) have been found to be related to differences in cross-country growth performance. 5 5 For a survey see Barro and Sala-i-Martin (1995) and for a comparative analysis Levine and Renelt (1992). 3

Figure 1: A classification of the literature on policies and growth Policies and Growth Efficiency of policies (e.g. Fischer 1993, Edwards 1992 or Easterly/Rebelo 1993) Reliability of policies Objective measures of instability (e.g. Barro 1991, Alesina et al. 1996 Aizenman/Marion 1993 or Hausmann/Gavin 1996) Subjective measures of uncertainty Country experts (Mauro 1995, Knack/Keefer 1995) Local entrepreneurs The second branch of the literature emphasizes the reliability of policies, i.e. their stability and uncertainties surrounding their implementation. Within this branch most studies use objective measures of political instability to proxy for uncertainties. The variables most often used are average numbers of violent political events 6 (e.g. riots or political assassinations), the number of or the estimated probability of government change 7 (e.g. orderly government changes or revolutions and coups) and/or the volatility of macroeconomics variables 8 (e.g. standard deviations of inflation or tax incomes). Clearly these objective variables are incomplete proxies for the variety of institutional uncertainties that entrepreneurs are confronted with in their daily business operations. For instance they 6 7 8 E.g. Alesina, Oezler, Roubini and Swagel (1996) or Barro (1991). E.g. Londregan and Poole (1990) or Cukierman, Edwards and Tabellini (1992) E.g. Aizenman and Marion (1993) Easterly and Rebelo (1993) or, more recently, Hausmann and Gavin (1996). 4

disregard more micro-aspects which entrepreneurs consider important such as uncertainties in tax legislation, large and unpredictable changes in labor regulations, uncertain and arbitrary decisions of courts or unclear proceedings in the allocation of all sorts of licenses etc. 9 Two examples can help making the point. Take for instance Thailand 10. Indicators of political instability which are based on counting the number of coups would characterize Thailand as a country with high political uncertainty. But the interviews we conducted with businessmen suggest that the coups did not affect the credibility of the institutional framework and that entrepreneurs did not fear wide-ranging policy swings or reversals. The opposite case is Peru in the 1980s. 11 Despite the apparent stability of the government, legislation through executive and emergency decrees was so extensive that the private sector faced a much more uncertain environment than measures of the number of government changes could capture. These examples highlight the two problems of all objective indicators of political instability as proxies for policy reliability: The first problem is that they concentrate on events that the private sector may not perceive as important and the second is that they fail to capture many uncertainties that the private sector perceives as crucial. In essence the problem of objective variables is that they measure instability and not uncertainty. Instability can be objectively observed whereas uncertainty is subjective to the individual investor. Because investment decisions are based on the subjective evaluations of businessmen, a variable that captures these perceptions would seem more promising when trying to explain investment and growth. The subjective measures of political uncertainty that have been used in the literature 12 are based on opinions of external experts. Companies that specialize in assessing country risks commercially 9 10 11 12 See Borner, Brunetti and Weder (1995) for reports on interviews on these issues conducted with private businessmen in 10 LDCs. See Brunetti and Weder (1995). See Keefer (1990). See Mauro (1995) and Knack and Keefer (1995). 5

provide such indicators. The drawback of these indicators is that they are aimed at foreign firms and the problems for foreign investors and local entrepreneurs may differ quite substantially. For instance, to a large degree these indicators reflect risks of nationalization and impediments to repatriation of revenues that do not arise in similar intensity for domestic entrepreneurs. Also, the degree to which the investors are kept abreast of regulatory changes may differ significantly for multinational and domestic firms. Finally, multinationals may receive a very different treatment from politicians and bureaucrats than the large majority of small local firms. Given that in most countries, the development of the private sector mainly depends on local investors, an indicator based on their perceptions would seem a promising way for a more encompassing measurement of political uncertainty and its effects on investment and growth. 13 In this paper we aim at filling this gap by constructing a measure of the credibility of rules based on a private sector survey among domestic entrepreneurs. In this respect our survey extends the research based on country-level interviews with domestic firms done at the World Bank (see e.g. Stone, Levy and Paredes 1992) to the cross-country level. 2. The private sector survey This section gives a short overview of the survey. We discuss the structure of the questionnaire, explain how the survey was implemented, give an overview of the characteristics of responding firms and discuss possible problems of the data. 13 The importance of local investors is underlined by the findings of Feldstein and Horioka (1980) and the literature it triggered, which shows that a large portion of countries investment generally comes from domestic savings. 6

2.1. Questionnaire, survey implementation and sample characteristics The purpose of the questionnaire was to capture all relevant forms of policy uncertainties related to the development and enforcement of laws, regulations and policies. In preparatory interviews and tests of this questionnaire, firms that were confronted with unpredictable state action usually came up with very different examples of such uncertainties. 14 These answers ranged from surprising executive decrees to unpredictable court decisions, from uncertainty on the severity of tax audits to unpredictable custom procedures, and from policy reversals whenever a new minister is appointed to uncertainty whether a bribe would lead to blackmailing by government officials. The questionnaire tries to cover the most important forms of such uncertainties. The main part of the questionnaire consists of 25 mainly multiple choice questions. A subset of these questions will be used in the empirical sections of this paper. In addition, we asked respondents to judge the situation ten years ago; this allows to construct ten-year-averages of the indicators used in the econometric work. The process of implementing the survey began in August 1996 and ended in June 1997. At the survey s conclusion 73 countries had participated. In 60 of these countries the questionnaires were distributed through World Bank missions and/or local consulting companies. In 9 European countries the survey was undertaken end of 1996 as a separate exercise under our direction at the University of Basel. 15 Those surveyed at the University of Basel used exactly the same methodology. As the 14 15 The survey instrument was developed over the last four years. It started with a large number of interviews of private entrepreneurs in different Latin American countries that resulted in a short multiple choice questionnaire and small scale survey (results were reported in Borner, Brunetti and Weder 1995). Based on the results of this pretest the survey instrument was refined and expanded. In preparation for the WDR 1997 survey the expanded questionnaire was discussed with a number of country experts at the World Bank and at IFC. After these discussions the questionnaire was revised and finalized and resulted in the survey presented in this paper. The countries were: Austria, France, Germany, Ireland, Italy, Portugal, Spain, Switzerland and United Kingdom. 7

coverage of Southeast Asian economies proved to be rather poor after completion of the WDR, we later conducted surveys with the same method in 4 additional countries. 16 Companies were selected based on stratification criteria including firm size, geographic location within their country and foreign participation. The survey was conducted by direct mailing where possible. In some countries where mail delivery systems were unreliable, hand delivery was used. The average rate of return on the mailed survey in LDC countries was 30%. 17 Due to various constraints, not all the country surveys were based on a random sample of companies. Nevertheless the stratification criteria should ensure a reasonable coverage. Table 1 shows regional averages and some descriptive statistics of response patterns. 18 Over all countries, the average number of questionnaires is 53 and the median 50. The average number of responses was lowest in the developed countries (DC). Minimum number of questionnaires was 13 and the maximum 124. Regional information about the standard deviations and coefficients of variation is provided in appendix 6. Table 1: Returned questionnaires per region Number of surveyed countries Number of surveyed firms Average Median Minimum Maximum All countries 73 3,883 53 50 13 124 High income OECD 11 254 23 20 14 56 South and South- 7 337 48 43 29 88 East Asia Middle East and 3 109 36 42 15 52 North Africa Central and Eastern 11 771 70 70 46 114 16 17 18 The countries were: Hong-Kong, Republic of Korea, Singapore and Thailand. The response rate in high income OECD-countries was considerably lower (18% on average). For regional details on rates of response see Brunetti, Kisunko and Weder (1997). For reasons of confidentiality no individual country ratings can be revealed at this point. Therefore, we present these tables in the form of regional averages. 8

Europe Latin America and 9 474 53 47 17 87 Caribbean Sub-Saharan Africa 22 1,288 59 48 13 124 Commonwealth of Independent States 10 650 65 62 31 91 Table 2 gives regional information about distribution of size, geographic location and ownership of responding firms. About 40 percent of firms were small (less than 50 employees), about 30 percent were large firms (more than 200 employees) and the remaining 30 percent a third were of medium size. The sample, therefore, is reasonably diversified according to this criterion. The regional decomposition shows considerable variation in the percentage of firm size. This reflects differences in economic development and in the development of the private sector itself. Services and manufacturing are represented about equally and agriculture is underrepresented in all regions. Regarding geographical location, there is a certain bias towards the capital city, however, in many countries this reflects the distribution of firms. About two thirds of the surveyed companies do not have any foreign participation they are purely local. This contrasts with other earlier attempts of subjective measurement of investment climate that concentrate entirely on the perceptions of multinational firms. Company Size: Table 2. Description of responding companies by region (percent of all responses in the region) All countries High income OECD South and South- East Asia Middle East and North Africa Central and Eastern Europe Latin America and Caribb. Sub- Saharan Africa Commonwealth of Indep. States less than 50 employees 39 26 32 35 40 27 43 61 > 50 and < 200 employees 31 45 28 35 28 29 31 23 more than 200 employees 28 28 40 26 31 42 24 15 Industry: Manufacturing 49 69 52 51 48 41 46 35 Services 41 27 43 42 41 47 39 57 Agriculture 8 2 4 1 10 9 11 7 Location of management: Capital city 48 23 49 42 37 59 58 61 Large city 28 29 24 40 36 25 26 21 Small city or countryside 21 48 11 12 27 13 13 18 9

Foreign participation: yes 35 33 47 34 26 30 42 25 no 63 65 52 62 73 67 56 73 Exports: yes 49 73 55 37 51 44 46 28 no 51 27 45 63 49 56 54 72 2. 2. Possible problems with selection bias and measurement error Before turning to empirical results we discuss possible selection biases and measurement errors of our approach. In most cases we believe that they should not seriously affect the quality of the results. A first possible source of selection bias is that for most of the 60 countries surveyed over World Bank contacts, governments had to be asked if they agreed that firms in their country participate in the survey. This introduces the problem that the countries with low credibility and low growth could choose not to participate in the survey because their government might fear to have this fact exposed. This bias would exclude the worst cases of low credibility. Not all countries were asked in the first place because the most important constraint in determining which countries were covered was the internal administrative capacity of the World Bank to organize the survey in a short time. Of the countries that were asked only 5 explicitly chose not to participate and in 5 more there was no official response or the resident mission preferred not to conduct the survey; this bias, therefore, does not seem to be too strong. The fact that the questionnaire involved some delicate questions on the relationship to the government might be another source of selection bias. There could be two possible problems. Entrepreneurs which are completely exasperated with their government might take the opportunity to vent their anger while entrepreneurs who feel reasonably happy might choose not to answer to the survey. In this case the bias would be to consistently underestimate credibility. The other possibility is that entrepreneurs who are desperate have given up and do not even care to submit a questionnaire. This would lead to an overestimation of credibility. Similarly, some entrepreneurs might fear that governments find out about their responses and therefore present a too rosy picture. In order to temper 10

this fear we conducted the survey anonymously and asked for no company-specific data which would allow to identify the responder. All in all, the direction of a possible company-level bias is not evident: it could lead to under- as well as overestimation of our variable of interest. A more serious source of measurement error could be that purely local entrepreneurs might not have the experience to put their answers in relation to the situation in other countries. About 60 percent of the total sample of enterprises were purely local, i.e. they had no foreign participation and did not export. Of course entrepreneurs might still have had good knowledge of other countries (through imports, or they might even be nationals of other countries) but probably in the smaller enterprises there might be businessmen who were purely local. On the one hand this is exactly what we want, because when a local entrepreneur feels severely threatened by uncertainty this would affect his investment behavior in the country. On the other hand, if this problem was serious it would compromise the comparability of country surveys and we would not find any association between uncertainty indicators and economic performance, even though it might be there. The fact that below we do find strong associations between economic performance and indicators derived from the survey is, therefore, an indirect prove of the validity of the instrument. This same argument, raises an additional possible measurement problem, namely, it might be argued, that the survey is nothing but an indirect measure of the private investment rate. It is conceivable that entrepreneurs would respond to questions about the business environment with their general gut feeling ; that their responses would not reflect their opinion about the institutional framework but whether they invested or not. In this case we would expect that entrepreneurs would respond more or less the same to all questions. In other words, if the firm has recently invested in the country the entrepreneur would answer all the questions positively and visa versa. It seems that this is not the case. Businessmen seemed to clearly distinguish between, say, the perceived political stability and the level of corruption. An entrepreneur who would just be expressing his overall gut feeling on the country would tend to tick the same or very similar ratings for all questions; the degree of differentiation in answers for different questions from the same respondents is, therefore, comforting. 11

3. The credibility indicator and its components 3.1. Construction of the credibility indicator The multiple choice questions used in the survey had six standardized responses. For instance in question number 1 entrepreneurs were asked whether they had to cope regularly with unexpected changes in rules and regulations which could seriously affect their business. The six answers ranged from completely predictable to completely unpredictable. Based on such standardized answers indices could be constructed for every question. 19 The credibility-indicator was designed as a broad measure of the reliability of the institutional framework averaging the information from many such questions. It encompasses several different sources of uncertainty in the interaction of government and private sector and summarizes them into one global indicator. The credibility-indicator is composed of five subindicators. It is constructed as the simple mean of the average answers that make up these five subindicators. 20 (1) Predictability of rule making: Extent to which entrepreneurs have to cope with unexpected changes in rules and policies and whether they expect their governments to stick to announced major policies. The degree to which entrepreneurs are usually informed about important changes in rules and if they can voice concerns when planned changes affect their business. (average of questions 1-4) (2) Subjective perception of political instability: Reflects whether government changes (constitutional and unconstitutional) are perceived to be accompanied by far-reaching policy surprises which could have serious effects on the private sector. (average of questions 5-6) 19 20 Entrepreneurs were also asked how the situation was 10 years ago. The average of the response for 10 years earlier and the value of 1996 was used to construct a 10 year average (for the transition economies only 5 year averages were considered). For the indicators of security of property, judiciary enforcement and perceived political instability we asked directly in the questionnaire how the rating was 10 years ago. For the indicators of predictability and corruption we asked one overall question for several questions at the end of the block. The individual questions used for the construction of these indicators can be found in appendix 1. 12

(3) Security of persons and property: Reflects whether entrepreneurs feel confident that the authorities would protect them and their property from criminal actions and whether theft and crime represent serious problems for business operations. (average of questions 7-8) (4) Predictability of judicial enforcement: Captures the uncertainty arising from arbitrary enforcement of rules by the judiciary and whether such unpredictability presents a problem for doing business. (question 9) (5) Corruption: Asks whether it is common for private entrepreneurs to have to pay some irregular additional payments to government agents to get things done. (question 10) 3. 2. Regional statistics for the credibility indicator and its components Figure 2 shows regional averages of the credibility indicator. An indicator value of 6 means perfect credibility and a value of 1 no credibility at all. Figure 2: Regional averages of credibility index 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 High income OECD South and South-East Asia Middle East and North Africa Central and Eastern Europe Latin America and Caribbean Sub-Saharan Africa Commonwealth of Independent States 13

Table 3 gives more detailed information by showing the regional ratings for all of the five subindicators that are used to calculate the overall credibility indicator: Table 3: Regional averages of credibility and its components The high income OECD countries overall prove to have the most favorable institutional framework. Their firms clearly assign the highest credibility rating. The high growth South and South- East Asian countries have very good credibility ratings as well. At the lower end of the regional averages we find the Commonwealth of Independent States and the Sub-Saharan African countries. Credibility Predictability Political Stability Violence Reliability of Judiciary Lack of Corruption All countries 3,23 3,21 3,25 2,80 3,04 3,86 High income OECD 4,15 3,85 4,27 3,64 3,98 5,04 South and South-East Asia 3,69 3,55 3,56 3,28 3,94 4,12 Middle East and North Africa 3,28 3,36 2,86 3,57 2,61 4,01 Central and Eastern Europe 3,22 2,93 3,51 2,72 3,14 3,82 Latin America and Caribbean 3,12 3,17 3,60 2,43 2,63 3,79 Sub-Saharan Africa 2,91 3,06 2,57 2,59 2,76 3,55 Commonwealth of Independent States 2,69 2,87 2,91 2,16 2,35 3,16 Again, the high income OECD have the best rating for all of these indicators. This is especially apparent for corruption where this region has an extremely favorable rating of 5.04 which is much higher than the next best value of 4.12 for the South and South-East Asian region. For all of these indicators the region of the Commonwealth of Independent States has very low ratings. In particular political violence and the lack of a reliable judiciary seem to be major problems in this region. As instructive as such regional comparisons are, we cannot derive any strong conclusions as a lot of information on cross-country differences is averaged out. Therefore, in the next sections we turn to an econometric analysis of the data set that can take advantage of individual country ratings. 14

4. Specification and data sources for the empirical analysis In the empirical analysis we use cross-section regressions to evaluate the hypothesis that high credibility is associated with higher growth rates and higher rates of investment. Starting with the contributions by Kormendi and Meguire (1985) and in particular Barro (1991) this has become the standard method for analyzing the sources of cross-country differences in economic performance. Our indicator and subindicators of credibility will be added as additional explanatory variables in the most common specification of such growth regressions. This specification regresses the average rate of growth on the starting levels of per capita GDP and human capital. The first variable controls for the convergence effect predicted by neoclassical growth theory; the higher initial GDP per capita the lower will be the ceteris paribus growth rate as decreasing returns to capital reduce the growth effects of additional capital. According to this argument, a country with low starting level of GDP should grow faster and gradually converge to the levels of higher developed countries. The problem with this approach is that it does not work for country samples that include LDC s and DCs. 21 Mankiw, Romer and Weil (1992) have argued that the neoclassical growth model does not predict absolute but conditional convergence. Each country does converge, but not to a common steady state but to its own steady state that depends on country characteristics, most prominently the level of human capital. As a consequence, more recent cross-country growth regression analysis has included, as we also do, at least a measure of human capital as an additional right-hand variable in the basic specification. In addition to testing the credibility measures in this basic specification we will check whether the results are sensitive to adding individual additional explanatory variables that are frequently used in the empirical growth analysis. The specification we test, therefore, has the following form: Growth8092 = a 0 + a 1 GDP80 + a 2 SEC80 + a 3 Credibility + a 4 X + u 15

Growth8092 is the average per capita growth rate for the period 1980 to 1992 calculated from the updated data set provided by Summers and Heston (1991). 22 GDP80 is per capita GDP in 1980 from the same data set. SEC80 is the enrollment ratio in secondary school in 1980 from the UNESCO Statistical Yearbook. Credibility is the average indicator calculated from our survey approach for the last decade. X is an additional variable that is drawn from a sample of standard explanatory economic and political variables for economic growth. 23 As control variables we will include the following frequently used measures: The average rate of inflation ( Inflation ) from 1980 to 1992 calculated from World Bank sources, the average rate of government consumption per GDP ( Govern. Consump. ) for 1980-1992 provided by Summers and Heston, for the same period, the average degree of openness to international trade measured as the sum of exports and imports as a percentage of GDP ( Trade ) calculated from Summers and Heston and the average level of liquid liabilities in GDP ( Lly ) in the eighties from King and Levine (1993). In addition we will check how credibility affects economic growth. Credibility can influence growth by either affecting the accumulation of capital or by affecting the allocation of capital to different sectors. We can try to disentangle these effects by separately estimating investment regression in order to check the effect on accumulation and growth regressions that control for investment in order to check the effect on allocation. 24 5. Regression results 21 22 23 24 See e.g. Barro and Sala-i-Martin (1992). We use the period 1980-1992 because this is the most recent period for which Summers and Heston-data is available on the internet. As the institutional framework is unlikely to change very much in a short period of time we feel confident that the overlap of this period and the calculated credibility indicator is sufficient. One of the sensitivity test we perform is to use the credibility indicator in growth regressions for the longer period og 1970-1992. As in all cross-section growth-analysis this of course raises the issue of causality. Due to the notorious problem of finding adequate instruments this issue is very hard to address. For a discussion see Mankiw (1995). Precise definition of data and data sources can be found in appendix 3. See King and Levine (1993) or Fischer (1993). 16

Using specifications and data discussed in the previous parts, this section presents the results of the cross-country regression analysis. We proceed in five steps. In subsection 1 we present the basic results of the overall credibility indicator in growth- and investment-regressions for 52 countries for which we have reliable data (see appendix 2 for a country list). Subsection 2 discusses some problems of specification and tests whether they are related to the sample or the period covered. Subsection 3 individually tests each of the five subindicators that together make up the credibility indicator. In subsection 4 we compare the credibility indicator with other political variables. Finally, in the last subsection we do some exploratory analysis with data from 18 transition economies in our sample. 5.1 Basic growth and investment results We first test the relation between the aggregate indicator of credibility and average per capita growth rates for the period 1980-1992. A higher the value of this indicator means a more credible institutional framework. Therefore we expect a positive relationship. The simple scatterplot is shown below (Figure 2): 17

Figure 3: Scatterplot of credibility and GDP per capita growth 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00-0.02 1 2 3 4 5 6-0.04-0.06 Credibility Table 4 displays multivariate regression results. The first regression shows that the sign of the coefficient is positive in the basic specification that contains GDP per capita and secondary school enrollment as additional right-hand variables. The coefficient is significant at the 5 percent confidence level. Regressions (2 ) to (5) test whether this result is sensitive to the inclusion of additional explanatory variables. Controlling for the rate of government consumption the indicator of credibility has the expected positive sign and is significant on the 1 percent level. If we include the rate of inflation and the extent of international trade the coefficient of the indicator is significant only at the 5 and 10 percent confidence level, respectively. In the last regression we control for the level of financial depth, the level of liquid liabilities of the banking system (Lly). In this case both Lly and credibility are insignificant. However, credibility and Lly are highly correlated (simple correlation of 0.78). 25 This high correlation might be no coincidence because the two variables might be measuring the same phenomenon. 25 See Appendix 5 for the correlation matrix of the survey indicators with all economic variables. 18

Table 4: OLS growth regressions for credibility indicator Dependent variable: GDP per capita growth 1980-92 (1) (2) (3) (4) (5) Constant -0.07 *** -0.04-0.07 ** -0.06 ** -0.06 * (-2.86) (-1.44) (-2.35) (-2.47) (-1.93) GDP80-2.2 E-6-2.8 E-6-2.4 E-6-1.6 E-6-2.3 E-6 (-1.23) (-1.57) (-1.27) (-0.89) (1.26) Sec80 0.027 0.012 0.031 0.026 0.035 (1.56) (0.45) (1.15) (1.02) (1.28) Gov. Cons. -0.001 (-2.10) Inflation -0.01 (-1.05) Trade 0.0001 * (1.86) Lly 0.006 (0.27) Credibility 0.023 ** 0.024 *** 0.022 ** 0.017 * 0.017 (2.62) (2.82) (2.29) (1.81) (1.53) Numb. Obs. 52 52 46 52 47 Adjusted R 2 0.21 0.26 0.21 0.25 0.15 T Statistics in Parentheses * Significant at 0.1 level; ** Significant at 0.05 level; *** Significant at 0.01 level Clague, Keefer, Knack and Olson (1996) have suggested that the depth of the financial system can also be interpreted as a variable of contract enforcement. The less confidence there is in contract enforcement the less there will be intermediation through the banking system. If this interpretation is correct, the fact that credibility becomes insignificant when controlling for liquid liabilities of the banking system is not a cause of great concern. In the following regressions, we, therefore, exclude this measure as an additional control variable. We proceed to check whether credibility has a positive impact on growth through higher rates of investment. Figure 4 shows that investment and credibility are highly correlated. 19

20

Figure 4: Scatterplot of credibility and investment 40 35 30 25 20 15 10 5 0 1 2 3 4 5 6 Credibility Table 5 presents regresssions of the average rate of investment per GDP on the same set of variables as used in the growth regression. Regression (1) shows that the coefficient of credibility has the expected positive sign and is significant at the 1 percent level. Together with the variable for initial human capital and GDP per capita this minimal specification explains 63 percent of the cross-country variations in investment rates. The result proves to be quite robust as can be seen in the extended specifications tested in regressions (2) to (4). When controlling for government consumption and inflation credibility remains positive and significant at the 1 percent confidence. If we included the extent of international trade, the significance of the credibility indicator drops to the 5 percent level. 21

Table 5: OLS investment regressions for credibility indicator Dependent variable: Average Investment rate 1980-92 (1) (2) (3) (4) Constant -7.29-2.70-7.22-5.55 (-1.59) (-0.51) (-1.38) (-1.21) GDP80-8.5 E-5-1.7 E-4-6.3 E-5-2.7 E-5 (-0.25) (-0.50) (-0.17) (-0.89) Sec80 14.22 *** 11.88 ** 13.98 ** 13.92 *** (2.96) (2.42) (2.70) (2.96) Gov. Cons. -0.19 * (-1.68) Inflation -0.11 (-0.06) Trade 0.025 * (1.78) Credibility 5.15 *** 5.28 *** 5.16 *** 4.02 ** (3.13) (2.42) (2.83) (2.32) Numb. Obs. 52 52 52 52 Adjusted R 2 0.63 0.64 0.60 0.65 T Statistics in Parentheses * Significant at 0.1 level; ** Significant at 0.05 level; *** Significant at 0.01 level The investment regressions tested for the effect of credibility on the accumlation of ressources. In order to test the allocation channel we included investment as a control variable in the growth regression. Credibility keeps its positive sign in these regressions but it becomes insignificant (not shown). This result suggests that higher credibility affects growth mainly through investment, that is by raising the rate of capital accumulation. 22

5.2 Specification issues with standard economic variables in the growth regressions In this section we discuss some peculiarities of the results obtained in the growth regression for the other control variables; in Table 4 most standard economic variables do enter with the expected sign but they are mostly not significant. The GDP per capita has a negative sign but it mostly not significant, secondary school enrollment is positively associated with growth, but again the coefficient is not significant. 26 Similar problems arise for the other economic variables, government consumption and inflation are insignificant and trade is only marginally significant. This differs from the results usually obtained in the empirical growth literature with larger country samples and longer time periods. One possibility is that the period we study is too short to find the standard long run relationships. Also, there could be some peculiarity in the decade under consideration which may lead to different results than in longer series. After all, the period 1980-1992 includes the lost decade in which the consequences of the debt crises were born out. Alternatively, our country sample countries might systematically differ from the larger country samples usually studied. For instance, in our sample the proportion of African countries is quite large and this might lead to different results. To test if there is something peculiar about our country sample we first exclude one large outlier by including a dummy in the growth regression. The first regression in Table 6 shows that this outlier is indeed responsible for the atypical behavior of the two control variables. Once the dummy is included, GDP per capita and secondary school enrollment turn significant. This country is an outlier in the 1980-1992 period, because it a had a very large exogenous drop in per capita GDP during this time. Including the dummy clearly strengthens the result for credibility (the t statistic is rising 3.17) and the adjusted R 2 rises from 0.21 to 0.35 compared to the basic regression. The second regressions shows that with respect to investment the outlier-country with respect to growth is not a clear outlier. The dummy is insignificant, the fit of the regression is only marginally improved. We also conducted extended 26 For a discussion of the problems of finding adequate measures of human capital for cross-country analysis see Pritchett (1996). 23

regressions including the dummy and obtained much better results both for the control variables as well as for credibility. This leads us to believe that there is no large country sample bias in our data set. Table 6: Outlier and period test, OLS regressions Dependent var Growth8092 Invest8092 Growth7092 Growth7092 Growth7092 Growth7092 Constant -0.08 *** -8.10-0.082 *** -0.023-0.083 *** -0.063 *** (-2.86) (-1.78) (-3.25) (-0.80) (-2.75) (-2.70) GDP -4.1E-6 ** -2.5 E-4-9.1E-6 *** -9.9E-6 *** -9.4E-6 *** 6.2E-6 *** (-2.38) (-0.72) (-3.87) (-4.65) (-3.72) (-2.70) Sec 0.054 ** 16.58 *** 0.094 *** 0.077 *** 0.093 *** 0.078 *** (2.19) (3.30) (2.87) (2.59) (2.74) (2.58) Gov. Cons. -0.002 *** (-3.54) Inflation -0.024 (-1.11) Trade 0.0003 *** Dummy for outlier -0.09 *** -7.79 (-3.34) (-1.44) Credibility 0.026 *** 5.36 *** 0.03 *** 0.03 *** 0.03 *** 0.02 ** (3.17) (3.28) (3.58) (3.52) (3.36) (2.15) (3.38) Numb. Obs. 52 52 51 51 46 51 Adjusted R 2 0.35 0.67 0.37 0.49 0.37 0.48 T Statistics in Parentheses * Significant at 0.1 level; ** Significant at 0.05 level; *** Significant at 0.01 level In a next step we test whether the time period matters for our results. The last 4 columns in Table 6 show results of regressions using longer period averages of per capita growth rates. For the longer period the growth regressions do not suffer from the mentioned specification problems. We obtain the standard results for the base control variables and other economic policy variables become significant (with the exception of the inflation rate). Regarding the credibility indicator these results are also very positive. Credibility is significant on the 1 percent level in almost all regressions. We also 24

conducted investment regressions for this longer period (not shown in the table) and obtained strong results for the credibility indicator: it is signficant on the 1 percent level in all specifications. It is possible to argue that institutional factors are deep societal fundamentals that do not change much over time. If this is true, the regressions over the extended period are the better test of the long term relationship between credibility and growth. The fact that credibility is significant in these standard regressions give some confidence in our results. 5.3 Subcomponents of the credibility indicator Table 7 presents results for the individual subcomponent of the credibility indicator for the basic growth regressions in both periods, 1980-92 and 1970-92. We have shown above that credibility affects growth mainly through the accumulation channel. Therefore, table 7 also shows results for investment regressions. Three out of the five subcomponents are significant in all regressions. The most robust one is the predictability of judiciary enforcement indicator which is significant on the 1 percent level in all regressions. The security of property rights indicator is highly significant in both growth regressions and remains significant in investment regressions, albeit at lower levels of confidence. The reverse is true for the indicator of perceived political instability. The indicator of predictability of rule making is the most fragile of all subcomponents. It is only significant in the growth regression for the longer period. The opposite pattern applies to the corruption indicator: it is significantly associated with investment but not with growth. It is interesting to note that the same result was obtained by Mauro (1995). He used business expert data and a different sample of countries, but also found that corruption affected investment but not growth directly. 25

Table 7: Coefficients and t-values for subcomponents of credibility in OLS growth and investment regressions Dependent variable Growth8092 Growth7092 Invest8092 Invest7092 Predictability 0.016 0.028 ** 2.77 3.74 of Rule Making (1.25) (2.08) (1.15) (1.56) Percieved 0.010 * 0.015 ** 2.79 *** 3.55 *** Political Instability (1.69) (2.38) (2.56) (3.35) Security of Persons 0.018 *** 0.024 *** 2.43 * 2.96 ** and Property Rights (2.62) (3.35) (1.76) (2.20) Predictability of 0.016 *** 0.018 *** 2.76 *** 2.92 *** Judiciary Enforcement (3.33) (3.46) (3.06) (3.26) Corruption 0.001 0.011 2.24 * 3.66 *** (0.16) (1.57) 1.79 (3.13) T Statistics in Parentheses * Significant at 0.1 level; ** Significant at 0.05 level; *** Significant at 0.01 level Other control variables (not shown) are GDP per capita in 1980 and schoolenrollment 1980 5.4. Comparison with other political variables Our next step is to compare the indicator of credibility with other political variables which are frequently used in cross country growth analysis. Table 8 shows results of base growth regressions for the period 1980-1992. Each regression includes four right hand variables: initial level of GDP per capita and secondary school enrollment (coefficients are not shown in the table), credibility and one other political variable. 26

Table 6: OLS growth regressions for credibility and other political variables Dependent variable: GDP per capita growth 1980-92 Political Variable Credibility Numb.Obs Political Rights -0.005 0019 ** 51 (-1.43) (2.17) Assassinations -129.7 0.021 ** 50 (-0.73) (2.32) Coups 0.011 0.020 ** 49 (0.79) (2.24) Wars 0.002 0.023 *** 51 (0.22) (2.57) Inst. Quality 0.010 0.009 42 (1.65) (1.05) T Statistics in Parentheses * Significant at 0.1 level; ** Significant at 0.05 level; *** Significant at 0.01 level The table shows that credibility remains significant in all but one regression. The other political variables we introduced, instead, are all insignificant. The first political variable is the indicator of political rights compiled by Freedom House. It is used as a measure of the level of democracy and ranges from a high of 1 to a low of 7. A negative sign of the coefficient therefore means that more democracy is associated with higher growth. From a theoretical point of view it is not clear whether a more democratic system necessarily leads to more growth than a less democratic system. Empirically, the recent cross country analysis has found no significant association between the level of democracy in a country and its long term growth performance. 27 This result is reproduced in our sample, the coefficient of the indicator is not significant at the 10 percent level. When we control for political rights, credibility remains significant and of the expected sign. The simple correlation between credibility and political rights is fairly high (-0.66) which indicates that democracy and credibility often go together. 28 27 28 See Brunetti and Weder (1995) for a survey of the literature on democray and growth. See Appendix 4 for the correlation matrix of all political variables. 27

The next three variables are objective indicators of political stability taken from Easterly and Levine. Here, we would expect negative signs of all coefficients. However, all three variables are not even close to significance. Credibility, on the other hand remains significantly associated with growth. The simple correlation between credibility and these three indicators is very low, ranging from 0.06 to - 0.18. The last other political variable is a subjective indicator of institutional quality compiled by ICRG, a professional risk rating company. 29 This is the indicator which is most readily comparable with the credibility indicator. It tries to capture similar problems, albeit by asking country experts rather than the local private sector. The results of this regression is not fully comparable with the ones above because the ICRG indicator was available only for 42 countries in our sample. The regression shows that institutional quality has the expected sign but is not significant. Credibility also looses significance in this specification. This is because the two indicators are very highly correlated. The simple correlation is 0.81. This is an interesting result in itself because it provides an indirect check on the validity of the survey method. It shows that the local private sectors have on average expressed similar views as country experts. The advantage of the surveys are that they provide much more disaggregated information than risk assessments companies do. 5.5. Growth results for an extended sample of transition economies In this last section we present results for a set of transition economies for which data could be gathered. The figure shown should be regarded as tentative mainly because of data limitations in transition economies. The results are not directly comparable with the previous sections because we had to work with different growth-data than for the sample of 52. In addition, given that 10 year averages are not very sensible in the case of transition economies, we look at the average growth rate 29 See Knack and Keefer (1995) for details. The indicator we use is an average of corruption, rule of law and quality of the bureaucracy. 28

for 1990-1995 (provided by the World Bank). The scatterplot of credibility and growth is shown in Figure 5. Figure 5: Scatterplot of credibility and GDP per capita growth for transition economies 4.0 3.5 3.0 Credibility Index 2.5 2.0 1.5 1.0 0.5 0.0-30 -25-20 -15-10 -5 0 5 GDP Growth, 1990-95 The scatterplot indicates that credibility may also contribute to explaining differences in growth performance in the transition economies. However, the results for this sample have to be interpreted with caution mainly because of the short observed time period, as well as intrinsic problems of measuring and explaining growth in countries that went trough such a major systemic change. Therefore, we do not perform a more formal econometric analysis. Conclusions and directions for further research This paper has presented a new data set and some results which indicate a close association between indicators of institutional uncertainty derived from this data set and economic growth. The two 29