III. Business Environment, Legal Institutions and Economic Performance: Cross-country Evidence and the Case of Italy30
Core Questions, Issues, and Findings.
- How robust and applicable are the insights from the fast-growing literature on the role of institutions in economics? Despite many studies of the links between institutions and economic performance, data imperfections elicit fundamental concerns. The subjective nature of many institutional measures is one key problem. The World Bank’s “Doing Business in 2005” dataset may overcome some of these concerns.
- What do the latest business environment data suggest about the links between institutions and growth? Slower, costlier, and more rigid procedures are correlated with lower levels of per capita income, across both high-income and all countries. While the correlations between institutions and income (and long-term growth) could reflect reverse causality, the pattern of relative strengths of the relationship between growth and different elements of the business environment suggests this may not be the case.
- What do the new data say about Italy? A number of the indicators included in the World Bank data set raise concerns about aspects of Italy’s business environment. In particular, the data suggest that some types of legal procedures in Italy take much longer to conclude than in other advanced economies. Although it is possible that the limited number of indicators included in the data set do not accurately summarize the overall state of the business environment, they are broadly consistent with the findings of surveys conducted by other entities. The results suggest that reforms to strengthen the business environment in Italy could have an important growth dividend.
50. This chapter will investigate the role of institutions and the judicial system in overall economic performance, with special reference to the case of Italy. Despite a general consensus that institutions have important implications for aggregate economic activity, demonstrating this has proven difficult, both theoretically and empirically. The chapter will focus on the growing global and Italy-specific academic literature on these linkages and will explore some empirical issues, including the role of cross-country evidence in general and in explaining developments in Italy.
51. More substantively, the paper will advance the following arguments. First, the recent comparative literature on the role of institutions may benefit from systematization. The paper will review the main contributions, qualifying the tentative agreement that institutions cause growth outcomes. Second, it will examine the empirical side of the problem in a cross-country setting (with reference to Italy), by using the World Bank’s “business environment” data. This section will illustrate the interaction of specific institutions with economic performance, emphasizing the dimensions of time, rigidities, and cost. Third, given evidence that Italy is an outlier in terms of some of the characteristics of its legal system, the paper will review some properties of that system. The conclusion will draw some policy implications.
B. Survey of the Literature.
General and cross-country studies
52. The “institutional” determinants of economic performance have attracted increasing attention from researchers over the last decade. Following widespread recognition of the work by Douglas North in the early 1990s, a number of recent papers (see Acemoglu et al. (2004), Rodrik et al. (2002), IMF World Economic Outlook, April 2003 (WEO, 2003) and references therein) have forcefully made a case for the primacy of institutions in understanding economic growth and income differences across time and countries. In particular, it has been argued that institutional factors appear to be “fundamental” causes of differences in income, compared to the “proximate” causes (such as labor and capital accumulation, technological progress, etc.) emphasized in much of the preceding mainstream theoretical literature on growth (Solow (1956), Barro and Sala-i-Martin (1995)).
53. Several competing theoretical frameworks for “fundamental” institutional determinants of economic outcomes have been suggested. In addition to the longstanding, but largely self-explanatory, themes of “efficient institutions” and ideology, two concepts have showed some, albeit still limited, promise recently. The legal origin theory (La Porta and others, 1997, 1998, 1999, 2000), and Levine (1997, 1998, 1999) argues that the Anglo-Saxon common law fosters financial – and hence economic – development distinctly better than do other legal systems.31 This theory has had some empirical validation and may explain why some features of particular economies have persisted for a very long time. However, it cannot explain the substantial and uneven economic and financial progress both within and across countries after the “historical accident” of legal adoption took place. The “political economy (social conflict)” theory (proposed, with variations on the theme, by Rajan and Zingales (2003), Pagano and Volpin (2001), and Acemoglu et al. (2004)) rejects the above deterministic view, and instead offers a “theory of institutional change.” The latter paper argues that financial and economic development is largely a product of the evolving comparative powers of different political interest groups, with a delicate dynamic balance between “de-jure” and “de-facto” forces. The theory gives a plausible interpretation of historical evolution, including of instances of substantial shifts in legal systems. However, it (i) does not explain persistence of institutions; (ii) has so far focused on only a limited field of institutions (political and economic); and (iii) has been difficult to formalize and test empirically.
54. Cross-country regression studies of institutional causes of economic performance have grown exponentially since the early 1990s. The initial motivation was a surge in the focus on growth empirics, whereby variables proxying the “rule of law” and related factors appeared to add explanatory power to cross-country regressions (see Barro (1991), Barro and Sala-i-Martin (1995)). Since then, continual progress has been made in extending research agendas and developing data and analytical tools for such analysis (see WEO, 2003). In addition to the focus on macroeconomic outcomes, a large body of research concentrated on the determinants and effects of specific institutional factors (legal, political, firm-level, etc.), at various levels of detail and disaggregation.
55. The empirical research on the macroeconomic effects of institutions could be summarized by the following points:
- With a few exceptions (legal origin, aspects of political economy),32 the studies have focused on “proximate” rather than fundamental institutional factors. However, the analysis of some proximate factors (i.e., financial systems) has developed into an influential and policy-intensive body of research. [See Rajan and Zingales (2003)].
- The more widely-used data for such analysis have been subjective measures of institutional quality [see Kaufman, Kraay and Zoido-Lobaton (1999), as well as various indexes constructed by EBRD (for transition economies), Transparency International, Heritage Foundation, etc.)]. Many of the indicators are compiled on the basis of surveys or expert panels, whose respondents may – often subconsciously – factor outcomes (including GDP) into judgments on the relative merits of the particular issues in a given economy. Furthermore, some of the data have a limited range of scores, which do not permit differentiation between countries.
- These aggregate measures of “institutional quality” have been correlated with indicators of economic performance. The correlation is positive with economic growth and per capita income, and negative with volatility. Furthermore, available indicators of institutional quality have been correlated among themselves. There is some evidence of causality between institutions and aggregate economic outcomes, but much of it remains precarious. While various methods have been used to test for - and have claimed - causality (panel/time-series techniques, two-stage least squares, instrumental variables, and, most recently, identification through heteroscedasticity (Rigobon and Rodrik (2004)), they do not appear to overcome fundamental subjectivity problems with the data on institutions [(see Glaeser et al. (2004)].
- Cross-country regressions have generally been based on the widest country sample possible, in part for the sake of greater statistical power. However, it is commonly recognized that the institutional fabric varies enormously across countries, and the profile of this variation has not yet been explored systematically.
56. The vast field of analysis of specific institutions is virtually impossible to summarize in terms of substance. However, in practice it has been largely confined to a few broad areas, such as macroeconomy, financial development; firm behavior/performance, and labor markets. Table 1 classifies selected recent cross-country contributions (most are empirical but some are descriptive) by plotting principal economic outcomes against main groups of institutional factors. The analysis is often interdisciplinary, and sometimes the data do not permit separation of single institutional factors. Thus, the empirical studies that use composite “rule-of-law” measures effectively test political, regulatory, and/or legal institutional factors jointly. Similarly, descriptive and empirical literature on financial systems and corporate structures [see references in Pagano and Volpin (2001)] posits interrelationships at the intersection of political, regulatory, and corporate governance factors.
|Political economy||Regulatory Quality||Financial system||Corporate governance||Legal “efficiency”||Legal tradition|
|Macroeconomy/Growth/Volatility/General economic efficiency||Perotti (1995)||Central Bank Independence; Fiscal rules, etc.||Marchesi (1998, 2003), EBRD, Kaufman,||Levine (1997-1999)|
|Financial development||Pagano and Volpin (2001) Rajan and Zingales (2001)||Pagano and Volpin (2001)||Zingales (2004)||Pagano and Volpin (2001)||Zingales, Rajan and Kumar (2001)||Levine (1997-1999) La Porta et al. (1997-2000)|
|Firm structure, behavior and performance||Roe (2003)||Scarpetta et al. (2002) Bartelsman et al. (2003)||Zingales (2004)||Zingales, Rajan and Kumar (1999)||Roe (2003)|
|Labor markets||Scarpetta et al. (2002)||Ichino, Ichino and Polo (1998)|
|Determinants of specific institutions||Botero et al. (2004)||Djankov et al. (2002)Botero et al. (2004)||Djankov et al (2004); La Porta et al. (1998)||Roe (2003)||Djankov et al (2003)||Djankov et al. (2003)|
57. Importantly, the above body of research has examined not only the effects of particular institutions on economic outcomes, but also their determinants, which gave rise to more objective measures of institutions. The literature on the determinants of specific institutional factors is highlighted in the last row of Table 1. Most of the papers in question have analyzed a given institutional factor in detail, often checking consistency of predictive power through regressions of the relevant measure on a set of exogenous or quasi-exogenous variables, such as legal origin, geographical and country dummies, GDP (which is posited as a proxy for the country’s level of development), etc. As an input or by-product of this analysis, more detailed and objective measures of institutions, mostly related to the business environment, have been constructed.
58. A number of international surveys report the perception of problems with Italy’s business environment. According to the Global Competitiveness Reports (GCR), published by the World Economic Forum on the basis of a survey of business executives, Italy ranked only 40th on average in 2002-2004 among some 100 countries in terms of the synthetic “growth competitiveness” index, which includes many institutional factors; the country slipped to 47th place in 2004. As Figure 1 shows, subcomponents relating to perceptions about the quality of public institutions (notably contracts and law and corruption) were a drag on Italy’s rankings in 2002-2003. (Given that the number of countries has increased in the 2003-2004 report, these rankings have been adjusted to be comparable to those in the 2002-2003 report). At the same time, a number of Italy-specific GCR scores provoke questions. It is puzzling that the assessment of the Italian legislature changes so much between 2002 and 2003, and is probably a result of excessive short-term swings in the judgment of respondents. Also, Italy’s 2003 ranking on “red tape” (18) was much better than that on the extent of administrative burden (73) and government favoritism (51), which seems inconsistent. The broad GCR’s conclusions are echoed by Transparency International reports, wherein 41 countries are perceived to suffer less from corruption than Italy.
Figure 1.Italy’s Competitiveness Rankings, 2002-2003.
59. Many economists, within and outside Italy, have argued that institutional factors have contributed to slow growth over the past decade. The last few years have seen a further intensification of the debate over the main causes of Italy’s slow-growth trap (see Ciocca (2003)). A number of recent comprehensive reports, notably by the Bank of Italy, Italy’s statistical agency ISTAT, Confindustria, and various Italian think-tanks, have emphasized a broad range of complementary institutional and economic factors, under the label of “sistema-paese,” or factors that affect the country’s national “competitiveness” in the broad sense of the term. These have included institutional factors behind rigidities in product and labor markets, particular features of the financial system, corporate governance and firm performance, lack of spending on research and development, and other factors, including those related to the pressures on external competitiveness [see Bank of Italy (2004) and ISTAT (2004)].
60. There is no consensus however on the specific structural factors that may have constrained overall economic performance. While not long ago the emphasis was placed on the reforms of labor market rigidities [see Decressin (2000)], the persistence of low growth in the last few years against the background of increasing employment and some recent liberalizing labor market reforms have prompted researchers to pay more attention to factors affecting productivity. In this respect, much recent research (including by the Bank of Italy) has tended to emphasize supply-side factors (society’s “productive fabric”), notably the combination of reasons affecting firm size and incentive structure, quality of infrastructure and investment, level of education, and mechanisms to deal with regional disparities. At the same time, from an operational perspective, there has been a need to put in sharper focus structural aspects related to the quality of institutions, compared to other factors (e.g., external shocks, neoclassical-type analysis of productivity dynamics, etc.).
61. Many of Italy’s institutional features have been well-documented, including their effect on the economy. This research has been partly facilitated by the wide regional variations. With substantial oversimplification, Table 2 classifies (a small subset of) Italy-specific literature on the role of institutions along the lines of the sectoral breakdown of Table 1 (note that the “legal tradition” column is obviously not applicable to a single-country case). In particular, some the main themes of this literature have been: (i) interaction of regulatory rigidity and firm characteristics; (ii) the effects of a relationship-based financial system (in part associated with the lack of transparency or political economy) on financial and corporate development; and (iii) aspects of Italy’s legal system. In what follows, some of these themes will be further explored through an analysis of Italy’s business environment and of some features of its legal system.
|Financial system||Guiso (2004)|
|Lombardo and Pagano (2000)||Bianco et al. (2002)|
Fabbri and Padula (2001)
Bank of Italy (2004)
|Schivardi and Torrini (2004)|
Bank of Italy (2004);
|Bianco and Giacomelli (2005)|
Marchesi and Pappalardo (2004)
|Labor markets||Decressin (2000)||Ichino, Ichino, and Polo (1998)|
C. Analysis of the Business Environment.
Introduction and data.
62. The analysis of the business environment is becoming an important part of the assessment of institutions. Some researchers have emphasized this channel between institutions and growth.33 Promisingly, a number of objective measures of the business environment have recently become available and have been studied extensively in a series of contributions (most of the entries in the last row of Table 1). As yet, these data have not been combined to evaluate macroeconomic effects of the aggregate business environment.
63. The World Bank’s most recent “doing business” dataset offers a broad-brush picture of institutional variations across countries. The database, available at http://rru.worldbank.org/DoingBusiness/ contains 25 indicators for over 140 economies (although for some countries the list is incomplete). The indicators reflect the situation as of January 2004, with the exception of two “control variables” (informal economy and GNI per capita), which refer to full-year 2003. The dataset includes topics such as business opening and closure, labor markets, property registration, financial development, and contract enforcement, thus covering all but two (broad political economy and corporate governance factors) of the institutional groups of factors highlighted in Table 1. Despite the extensive use of surveys in data compilation, most (but not all) of these indicators are not the outcome of subjective rankings, but rather represent a study of rules and regulations, with outputs being measured in terms of an objective characteristic (time, cost, etc.). Still, these data are far from perfect and sometimes may not be fully representative of the broader institutional processes.34 However, there is no better set of comprehensive cross-country indicators. Importantly, most of these data (except, possibly, “cost” indicators) would not have a built-in tendency to be biased by the countries’ GDP level, which could be the case for the other widely-used subjective ranking-based indicators.
64. This paper proposes to investigate the relationship between institutions and macroeconomic performance based on the characteristics of business environment.Such approach offers the following relative advantages:
- Reliance on “more objective” measures of institutions would alleviate the apparent bias in existing aggregate measures of institutional quality, potentially rendering results more meaningful.
- While the business environment per se has not been mentioned as part of the “deeper” theories of institutions, as a “proximate” cause it fits well with any of the “fundamentals,” by capturing at least some of the effects of political economy and the legal system on society’s economic performance.
- Furthermore, the business environment has its own “regulatory” policy dimension, as the indicators permit one to structure the analysis in terms of objective policy-dependent criteria or dimensions, such as the speed and cost of regulatory action, as well as specific rigidities (i.e., hiring/firing in the labor market).
- The wide choice of indicators allows the relative influence of each specific indicator (or group of indicators) to be assessed separately.
- One possible drawback of this approach, however, could be that the specific institutional indicators, even if more objective, could be measured too narrowly and prove less useful for interaction with aggregate economic outcomes. Furthermore, by restricting the analysis to a specific type of transactions it could fall prey to pathological cases in some countries.
65. The literature summary of Table 1 prompts a simple four-sector framework for tracking the significance of the World Bank data. It is proposed to classify the indicators along four main groups, relating to: (i) firm performance (start-ups and closure); (ii) labor markets (difficulties and costs in firing and hiring); (iii) financial sector (credit, collateral, disclosure); and (iv) legal/contract enforcement. This partly, but not fully, corresponds to the World Bank’s own-classification. For simplicity, we subsume issues related to property registration in the financial sector (given the clear relationship to the ability to borrow and claim collateral), although conceivably property registration could also be considered as one of the elements of the legal framework.
66. From another standpoint, the indicators could classified according to the objective metric used for their evaluation. In the case of World Bank data, the groupings would reflect: (i) cost of a given institutional transaction; (ii) time required to complete the transaction; and (iii) bureaucratic rigidity (number of procedures) and (iv) miscellaneous institutional indicators. The latter division proves a convenient organizing framework for presenting the results.
Italy’s business environment.
67. The business environment data reinforce concerns over Italy’s business environment arising from the survey-based indicators. Figure 2 shows Italy’s rank (among 145 countries)35 across the 25 specific indicators of the business environment, highlighting the main aggregate groups of such indicators. In this figure, countries with faster, less expensive, and less bureaucratic environment are ranked ahead of the others. Italy’s average rank across all 25 indicators is 58, which is consistent with its placement in most subjective surveys (see below).
Figure 2.Italy’s Business Environment Rankings, 2004.
68. At the same time, these indicators reveal significant asymmetries in Italy’s relative institutional characteristics, both compared to OECD peers and to all countries in the sample. Figure 2 and panels in Figure 3 prompt some conclusions. In particular:
- In terms of broad sectors, Italy’s indicators are somewhat weak in labor markets,36 as well as in a few key aspects of contract enforcement and firm closure. In firm startups, property registration and financial development Italy’s scores are generally slightly below those for other OECD countries, but significantly above those of most other countries.
- Regarding cost indicators, Italy’s results are almost uniformly above the world average, and moderately above those of OECD countries on most measures – notably firm start-ups and closure, workforce firing, and contract enforcement.
- The speed of most processes is on par or even slightly shorter than in OECD countries (and thus much better than the world average), with the crucial exception of the number of days to resolve debt-related contracts, whereby Italy comes out as second-longest in the whole dataset, by far Italy’s most extreme ranking across the whole realm of business environment indicators.
- Italy’s regulations appear more rigid (in terms of the number of administrative procedures needed to complete an action) than both the OECD and the world averages. The exception is contract enforcement, where Italy is on par with the former and much less rigid than the latter.
- Regarding available financial development and transparency measures, Italy’s scores are generally in line with those of other OECD countries, with the notable exception of creditor rights, on which Italy scores below the world and OECD averages.
69. Italy-specific data issues may qualify some of the conclusions from this analysis of the business environment. As noted above, many of the indicators in the World Bank’s cross-country dataset are compiled only on the basis of the information for the most populous cities. When there are significant differences in institutions across cities, this may give a misleading picture of overall institutional quality. Regional statistics indicate that the average length of civil justice proceedings of the first degree in 2001 (a measure of the speed of contract enforcement) was 1,082 days in Italy. Rome did somewhat better, at 960 days, but in Milan these proceedings took 813 days. In general, for most available law-based indicators of contract enforcement, procedures were longer in Rome than in the Northern provinces and Milan in particular.37 However, the apparent difference between Rome and the North of Italy on these measures does not seem very large. Other potential data problems with the “doing business” Italy-specific indicators may regard inertia in capturing some of recent reforms, especially in the labor market and in reducing the cost of business start-ups. The dataset is presumably designed to reflect the situation as of January 2004, but it is unclear whether the scores have fully incorporated the substantial improvements in the flexibility of labor contracts undertaken in Italy in the last few years.
Figure 3.Italy: Business Environment Indicators, 2004
Source: World Bank “Doing Business in 2005” dataset.
70. In view of the many indicators and the limited number of observations (for industrialized economies) and of the lack of a clear theoretical model to link all of these variables in a particular relationship, simple correlations offer some clues to the nature and direction of the evidence. In the following analysis, the sample is split into two parts: (i) all countries in the dataset and (ii) 27 high-income economies. Looking just at the latter offers the advantage of concentrating on a fairly uniform group of economies with reasonably mature and independent institutions.
71. Table 3 shows the correlation matrix with respect to the “cost” indicators. As is intuitive, in all sectors the costs of various institutional procedures are positively correlated, while all of the cost indicators are negatively correlated with per capita GDP.38 The results do not differ in qualitative terms whether measured in terms of the whole sample of 140 countries or 27 high-income nations. Thus, in countries where some institutions are relatively “costly,” the other institutions are costlier as well, while GNI per capita ceteris paribus, smaller. Interestingly, the indicator pertaining to the labor market (cost of firing workers) exhibits the strongest negative relationship with per capita GDP, followed by firm start-ups. The correlation between the financial and legal cost indicators on the one hand and per capita GDP on the other is somewhat weaker, but remains non-negligible. Also notable is the significant positive 3-way correlation for high-income countries (all higher than 0.5) between the costs of firm start-ups, firing workers, and creating collateral.
|Sectors||Variables||Firm startups||Firm closure||Labor markets||Financial sector||Contract enforc.||GNI per capita|
|Labor||Labor markets (firing)||0.66/0.14||0.32/0.24||1.00|
|Financial||Financial sector (creating collateral)||0.61/0.49||0.21/0.08||0.54/0.23||1.00|
|Legal||Contract enforcement (debt collection)||0.27/0.35||0.23/0.25||0.35/0.27||0.02/0.21||1.00|
|Macro||GNI per capita||-0.48/-0.37||-0.17/-0.32||-0.65/-0.31||-0.31/-0.29||-0.33/-0.41||1.00|
72. Table 4 presents the correlation matrix for the “time” indicators, with the results being similar to those of the cost dimension. There are also some differences in coverage. Still, the indicator relating to firm start-ups appears to be the most negatively correlated with GNI per capita among the other institutional indicators (for both high-income and all countries). The signs for the contract enforcement variable are intuitive, but again its correlation is weaker than that of the other institutional variables with the exception of the variable pertaining to firm closure, with the latter basically uncorrelated with GNI per capita.
|Firm start-ups||Firm closure||Registering property||Contract enforcement||GNI per capita|
|Days to register property (financial/legal)||0.28/0.12||0.41/0.02||1.00|
|GNI per capita||-0.45/-0.38||0.02/-0.37||-0.33/-0.20||-0.27/-0.31||1.00|
73. With respect to “other” institutional indicators, the correlation matrix (Table 5) shows a similarly intuitive picture. All pairwise relationships have the “expected” signs. Of note is the strong correlation within the labor market institutional indicators, particularly for the high-income countries, though this is not surprising. As expected, the data denoting “rigidity” (labor market indicators, number of various procedures) are strongly negatively correlated with per capita GDP, which indirectly supports the hypothesis that the same negative relationship between cost and per capita GDP is not driven by the construction of the cost data.
|Indicator||Firm min. capital||Diff.hires||Rigid hours||Diff. firing||Creditor rights||Credit inf.||# of proc.||GNI pc|
|Firms||Firm minimum capital||1.00|
|Labor||Difficulty of hiring||0.07/0.08||1.00|
|Rigidity of hours||0.28/0.09||0.57/0.27||1.00|
|Difficulty of firing||0.20/0.10||0.60/0.34||0.81/0.21||1.00|
|Legal||Number of procedures for debt collection||0.29/0.24||0.07/0.24||0.40/0.21||0.39/0.21||-0.06/0.33||-0.39/0.25||1.00|
74. A full table of correlations for the World Bank data (not shown) has also been calculated and is similarly in line with the basic intuition. Among other things, it helps check any changes in the results if one departs from the sectoral partitions proposed above. For example, one may look at the signs and the relative strength of the “within-sector” relationship between time, cost, and other institutional factors, compared to the “between-sector” linkages investigated above. The within-sector correlations are also intuitive, in that there is positive correlation between the “inefficiency” measures such as time, cost, and those “other” measures that should be inversely related to efficiency (difficulty in hiring, etc.). In particular, the intrasectoral correlation between time and cost is quite large (about 0.5) both for firm start-ups and contract enforcement, but is much smaller for property registration. Dummies for the “legal origin” were added to the correlation matrix; they support the “legal tradition” argument for the superiority of the common law system compared to the French system in terms of supporting growth of GNI. Finally, the share of the informal economy is positively correlated with cost, time and rigidity indicators, and negatively correlated with GNI per capita.
75. The above evidence is more suggestive than conclusive, but its conformity with intuition is reassuring. Slower, costlier, and more rigid specific business environment procedures tend to be correlated, across both high-income countries and the full set of countries, between and within different sectors, and are associated with lower levels of per capita output. This suggests that many of the highlighted sectors could potentially be important for economic performance, and possibly indicating complementarities among the reforms in case of very high pairwise correlations, for example between regulations that encourage firm start-ups and labor market regulations. The firm sector and the labor market appear to be the broad fields that have the strongest negative correlations between specific inefficiencies and per capita GDP.
76. While causality is not addressed by the correlations, the latter may be meaningful inasmuch as the data capture objective factors. It is still informative that, even assuming the absence of a causal link from efficiency of institutions to GNI per capita, richer countries deem it worthwhile to devote more resources to making their institutional frameworks speedier and less costly. It is also possible that the data collection process may have introduced additional subjective elements, in view of some particular features of the procedures that were chosen to achieve country comparability. This may in particular concern the coverage of the legal system, since the “contract enforcement” variable may be somewhat narrowly defined, being based on a particular sequence of debt collection transactions, rather than on the workings of the core civil law institutions and processes.
Business environment and growth.
77. In the absence of a consistent long-term time series, the above business environment indicators do not permit a rigorous and comprehensive investigation of the more fundamental question of causality from institutions to growth. Although some of the individual indicators have been available for a few years, they were presented in the current reasonably comprehensive form only during the 2004 exercise. In addition, even where the data existed prior to 2004, there have been some methodological changes in the compilation of individual indicators. Thus, it cannot be ruled out that the causality underlying the high correlation of these indicators with GNI per capita may run in either or both directions. A further complication is that the potential effects of many institutions on growth outcomes are likely long-term, which requires a fairly long time series for formal quantitative tests. When such a series is established on a consistent basis, the business environment data may be a promising avenue for formal empirical research of growth.
78. However, one may throw some light on the link from business environment data to growth by exploiting the persistent nature of some institutions. Given that institutional factors are likely differ in the extent of their persistence, it would be reasonable to expect that, at least for the relatively stable factors, the relative data patterns of 2004 would be roughly the same as in the not very distant past. This could be particularly true for such measures as contract enforcement (and perhaps property registration), which depend on the grassroots workings of the legal system more than on short-term policy initiatives. Some other institutions, for example those relating to labor markets or firm start-ups, could be more fluid, whether de-facto or de-jure. (A case in point are the recent rounds of labor market reforms in some countries of the European Union). If so, and assuming the causal link from institutions to growth, in a regression of past growth rates on current institutional factors one would expect a greater impact of relatively persistent institutions.
|Time indicators||Cost indicators||Rigidity indicators|
(number of procedures)
|Rigidity of employment||N/a||N/a||-0.00(-0.127)||N/a|
|# of observations||96||89||100||100|
79. Indeed, regression analysis based on these assumptions indicates that longer-term institutional factors such as contract enforcement have a more robust and logical relationship with past growth rates than other factors. Results reported in Table 6 suggest that all dimensions of contract enforcement (length, cost, and rigidity) have a (small) negative impact on growth, with two dimensions statistically significant.39 Other “grassroots legal” factors such as burdens associated with property registration and creating collateral also generally have negative and significant signs. The results are much more indeterminate, even in terms of signs, for such factors as firm start-ups and closure and rigidity of employment. This does not mean that these factors do not affect growth, but rather that the regulations and institutions might have changed significantly over the course of the sample period. (We know for a fact that they changed in many European countries during 1985-2003). Note that the beginning-of-period level of institutional quality was somewhat controlled for by including the log of the countries’ GDP in 1985. Interestingly, the inclusion of the legal origin dummies is insignificant for the regression results (and for the dummies) implying that the latter may affect institutions (as demonstrated by La Porta et al. (1997-2000), but do not appear to directly cause growth outcomes. In essence, these results appear to lessen concerns that the link between institutions and income/growth is driven by reverse causality from growth to institutions, although these issues, including the possible “simultaneity bias” have to be investigated more formally.40
80. The statistical significance is especially strong for the recovery rate variable (last column of Table 6). This is not surprising, given that it (i) seems a more synthetic, partly outcome-based, indicator of the overall quality of the business environment, possibly reflecting cumulative influences of a number of, but by far not all, factors; and (ii) is mostly related to the long-term workings of contract enforcement, which should interact with (and apply to) the fairly time-consuming process of bankruptcy, rather to the more short-term factors like firm entry or labor regulation. Thus, this result appears consistent with an intuitive story that “broader” business environment factors have a stronger (statistical) relationship to growth than “narrower” factors, implying that cumulative progress on a number of fronts would be most effective in affecting economic outcomes.
81. The pattern of potential dependence on the recovery rate varies with income level. An analysis of income-sorted data (e.g., by gradually adding each country observation analogous to the recursive analysis of the time series) indicates that the value of regression coefficient on the recovery rate is ambiguous for a dozen of highest-income economies, but resolves itself and becomes irrevocably positive as countries at Italy’s income level are added to regression. The coefficient becomes significant at a 5 percent level after a few more observations (around Greece’s and Portugal’s income levels) are added. Also, the coefficient value roughly stabilizes with a dozen of additional observations (in the vicinity of higher-middle-income countries). Thus, while a few countries (particularly at very high income levels) have selected idiosyncrasies related to role, or measurement, of the recovery rate,41 this indicator may be important in explaining variation in growth performance between other economies, including for many high-income countries similar to Italy.42
82. Robustness checks indicate that the cross-country link between “long-term” business environment indicators and growth is not dependent on outliers. Table 7 presents the results by splitting the dataset, in alphabetical order, into two non-overlapping subsets with 50 country observations each. The thrust of results generally carries over to both subsets of data, although the significance is always stronger in the “lower” subsample. An examination of the two subsamples suggests the main reason for the difference: while the upper one tends to have more middle-income countries, the other subsample includes relatively more extreme observations (both very high-income and very low-income countries). Thus, more income (and hence institutional) variation in the data tends to produce stronger statistical results, although the difference in the coefficients between the two subsamples is not very large for most factors. Excluding particular country outliers from the sample does not affect results. In particular, this is the case for two country outliers in terms of length of debt-related contract enforcement (Guatemala and Italy), both of which are, incidentally, in the upper subsample.
|Time of contract enforcement||Cost of contract enforcement||Rigidity of contract enforcement||Recovery rate|
|Full sample (100 obs.)||-0.000(-1.70)*||-0.003(-1.69)*||-0.010(-2.65)***||0.010 (4.59)***|
|Upper subsample (50 obs)||-0.000(-1.03)||-0.002(-1.26)||-0.014(-1.92)*||0.009 (2.61)***|
|Lower subsample (50 obs)||-0.000(-2.34)**||-0.005(-1.56)||-0.016(-3.02)***||0.012 (4.31)***|
83. To conclude, Italy’s rankings by the World Bank “doing business” indicators suggest significant scope to transform aspects of its business environment in line with best practices. Some institutional changes seem more compelling than others. Italy’s most dramatic disparity (among the 25 indicators) is the time required for contract enforcement, which exceeds six times the OECD average. With respect to save other indicators, Italy could benefit from undertaking a broad-based effort to match most OECD countries; while on each individual score the distance is not substantial, it cumulates across the many factors. Italy’s overall environment is perhaps best reflected by the “synthetic” indicator of the recovery rate during business closures as this indicator, implicitly or explicitly, incorporates the effects of a number of other variables. At 43.5 cents on the dollar, Italy would rank only 40th in the world, which is roughly consistent with the broader survey-based rankings reported above.
D. The Legal System.43
84. While a number of studies have documented the link between the legal system and economic performance (often in a cross-country setting), the treatment of legal system characteristics has typically been cursory. In particular, these studies have either been exclusively based only on subjective legal rankings data, or have been mainly limited to the investigation of the role of legal origin, which, ex-ante, appears too coarse for distinguishing among the many essential modern characteristics of legal systems. Furthermore, even if the investigation of the legal origin patterns yields some results, the natural “experiment” of adoption of legal systems clearly did not occur randomly and could also lead to biased results.44 Some authors have focused on the above “contract enforcement” data akin to those contained in the business environment dataset, but, as per above, these refer to debt-related transactions and may not fully capture all aspects of the legal process.
85. An alternative stylized, albeit also cursory, approach would be to eschew any subjective data and focus only on objective characteristics of the broader legal/judicial system. In this respect, civil law emphasizes three key dimensions of legal efficiency (Zuckerman (1999)): (i) objectivity or correctness of judgment; (ii) length of the legal process; and (iii) cost (public and private). While the degree of correctness of judicial decisions is necessarily a subjective matter, the length and cost could be documented by the data, which may be usefully compared in practice. Moreover, as noted by Marchesi (2003), in countries at a sufficiently high stage of judicial development, where judge impartiality and citizen equality before the law are overriding principles, correctness of judgment is addressed by guaranteeing both sides the right to present all of their arguments for the attention of the judge. For such countries, it is sensible to presume the objectivity of judgment as a rule, with material differences in the judicial systems being largely reduced to the length and cost of the legal process. Although the length and cost may well reflect differences in the legal tradition, focusing on these should permit a richer analysis of system’s effects on economic performance, including within a group of countries with the same legal origin.
Selected characteristics of Italy’s legal system.
86. Survey results suggest the perception of problems with Italy’s legal institutions. Figure 4 plots the main components of the GCR’s ranking of the “contracts and law” subindex, which is sub-divided into several areas. The first area, comprising the subjective scores for judicial independence, legal efficiency in resolving disputes, and property rights, could be considered a proxy for perceptions about the civil justice process. The second block measures opinions about reliability of police services, the business costs of crime and violence, and spread of organized crime, and is a rough proxy for the respondents’ views on the effectiveness of the criminal justice process and enforcement. The third block measures perceptions regarding the extent of regulatory burden, government transparency, favoritism of government officials, and bureaucratic red tape. Figure 4 shows that Italy’s subjective civil and criminal law-related rankings were essentially in line with those in other institutional dimensions. Figure 5 indicates an improvement in 2 out of 3 measures of the civil justice process between 2002 and 2003, but Italy’s position vis-à-vis other countries has remained basically unchanged.
Figure 4.Italy: Selected Contracts and Law Components, 2002 Figure 5.Subjective Indicators of Italy’s Civil Legal System, 2002-2003.
87. The data on Italy’s core civil law system confirm that the lengthy legal procedures are one of its defining characteristics. Table 8 shows that Italy’s ordinary civil judicial inquiries (comprising 3 degrees) were the longest among the European countries in 1996, and averaged some 116 months, compared to 69 for the EU average. Furthermore, as noted by Marchesi (2003), this aggregate indicator likely understates Italy’s difference with respect to other countries, in that the gap is particularly significant in the first two degrees of the civil process, which have more general applicability than the third stage.45 These data on the length of the legal process are different -- and more comprehensive -- than the World Bank data, since the latter reflect only the particular category of debt-related disputes.46 As can be seen from Figure 6, individual country scores may be substantially different on these two measures of the duration of the legal process, but Italy ranks largest on both counts among the European countries. The lengthly legal process has been exacerbated by Italy’s backlog of unresolved disputes
|average||17||21||31||69|Figure 6.Length of Legal Procedures in European Countries 1996 and 2002
88. Currently, there is a public debate in Italy on the factors accounting for these characteristics of the legal system. Marchesi (2003), while recognizing some moderate supply-side issues, has argued that the long legal procedures and the backlog of court cases reflect primarily excessive demand rather than insufficient supply. The number of legal professionals in Italy, on a per capita basis, has been among the highest in Europe (third-highest after Germany and Austria in the mid-1990s), and public expenditure on the legal system has grown very rapidly in the course of the last decade. The same author notes that the total financial resources devoted to the operation of the legal system in Italy is not below the European average, while the level of the judges’ preparation and productivity does not appear to be a major concern. At the same time, Zan (2003) has highlighted the organizational problems of courts as meriting particular attention in the context of lengthy procedures.
89. With respect to the costs of legal proceedings, Italy appears to be more competitive, but lower legal costs may not be an unqualified advantage if the legal process is excessively lengthly. Figure 7 shows that in 1996 the costs of legal assistance and procedural expenses in Italy were on average lower than in the European Union by some 45 percent for disputes with a value of €50,000 and by over 60 percent for disputes with a value of more than €200,000. As in many countries, the bulk of these costs in Italy are represented by attorneys’ fees (93 percent). At the same time, the low costs have co-existed with possibly excessive demand for legal services, as the proceedings involve fairly small contested values on average. Thus, about 60 percent of all first-degree judicial inquiries in economic matters (property and contracts) pertain to disputes with a contested value of less than €5,000, of which only 25 percent regard disputes with a contested value above €2,500. As a result, the Italian judicial system appears to be clogged with small cases, which may contribute to the length of the legal process.
Figure 7.Legal Costs in European Countries
Some effects of lengthy legal procedures: a conceptual framework.
90. A simple conceptual framework may account for the stylized set-up of a time-consuming legal system. The focus is on the length of the legal process and its interaction with costs and other institutional variables. The long legal processes could correlate with poor “enforcement,” and some empirical studies (Bianco et al. (2002) and Fabbri (2001)) assume the two are equivalent. The extent of this enforcement/slowness has mostly been assumed as exogenous in these studies. The assumption of exogeneity may be appropriate for some situations, but in general equilibrium this characteristic is endogenous, reflecting the dynamics of demand and supply of legal services.
91. Marchesi (2003) provides some ideas needed for an endogenous derivation of the pace of the legal process, but a comprehensive formal model for the interaction of supply and demand for legal services has yet to be studied carefully in this context. The supply and demand for legal services could be represented as a function of the length and cost of trials, respectively. Panel 1 of Figure 8 shows a possible (partial equilibrium) relationship with respect to the length of trials (n), whereby both the supply and demand curves would be upward sloping, the latter reflecting “pathological” demand, which stems from the fact that more extended trials, combined with the low rate of legal interest and the de-facto splitting of the legal assistance fees, provide incentives for the losing side to prolong as much as possible the legal process in courts.47 The supply curve would also be upward sloping, as, other things equal, more resources would be devoted to longer trials. The supply curve could alternatively be assumed flat (for example reflecting the fact that judges operate in a ‘fixed market”), but this would actually strengthen the argument. More generally, the argument requires that the supply curve be flatter than the upward-bending demand curve for very long trials.
Figure 8.Supply and Demand for Legal Services
92. Panel 2 shows the corresponding supply and demand curves for the legal services with respect to the cost of trials (c). Here the shape of the curves is more “normal,” with supply upward-sloping and demand downward-sloping. The supply curve reflects an assumption that, other things equal, higher cost of trials would attract greater supply of legal professionals (for example through greater remuneration of lawyers, which could well spill over to the rest of the legal profession).48
93. A persistent excess demand for legal services would be consistent with the combination of relatively low legal costs and slow legal procedures. Thus, the length of the legal process is greater than “optimal” (shift to the right from point O in Panel 1), while cost is lower than that in a would-be equilibrium (leftward shift in Panel 2). The excess demand creates a backlog of unresolved court cases, which exacerbate the imbalance in future periods (not shown in the static set-up of Figure 8). As is apparent from the Figure, a (limited) policy effort to increase the supply of legal services (e.g., judges), represented by an upward shift in the supply curve, may have a relatively minor effect on the degree of disequilibrium if the demand curve is strongly upward bending. Furthermore, to the extent the disequilibrium is not corrected fully, the continued give-and-go between snowballing of court cases and the pathological demand could reverse this progress.
94. The efficiency impact of the justice system on the economy partly operates through the enterprise sector. Following Marchesi (2003), one may assume that the economic agents pursue two basic strategies: (i) comply with their contractual obligations, or (ii) “hit and run,” (sign contracts with the intention of reneging on their obligations). Those that comply benefit from a good reputation and would be able to stay in the market. Those that renege acquire a poor reputation and have to exit the market, and also incur a penalty through the legal system, which eventually forces them to pay the procedural (legal) costs and deliver on the contractual obligations. Marchesi shows that in a simple setting characterized by perfect competition and rationality, the condition that would induce the agents to comply would be given by:
where n is the length of the judicial proceedings, S is the ratio of the procedural costs to the value of the contractual commitment, and r is the interest (or discount) rate. Essentially, if the legal procedures are slow enough, with the expected length exceeding n*, there would be no incentives to abide by the contractual obligations.
95. There are two main reasons why a market economy could support contractual transactions in countries with very slow legal systems, but both may involve governance and competitive inefficiencies. The first channel, emphasized by Marchesi (2003) is the recourse to “alternative justice,” for example in the form of arbitration. This would effectively reduce the average time spent on the legal process, i.e., the value of n.49Marchesi and Pappalardo (2004) analyze some of the implications of such alternative justice in the context of Italy’s industrial districts. However, recourse to “alternative justice” mechanisms is not a panacea and in certain situations may interfere with an orderly administration of justice as a public good. For example, the general scope for alternative justice could be exploited by organized crime. The second channel is acceptance of some competitive inefficiency in the system, whereby prices charged by producers would be higher than those under perfect competition. Established enterprises enjoy a reputational advantage that corresponds to the premium they can charge in the market, and which also operates as a barrier to entry. In this respect, lengthy legal procedures could be a factor deterring entry and limiting competition. Lack of competition is widely regarded as a substantial problem for Italy, especially in the services sector. Demirguc-Kunt and Maksimovic (1998) provide some indirect cross-country evidence on the negative effect on competition, showing that the return on capital depends negatively on the “law and order” variable.50
96. According to researchers, a situation of a lengthy legal process due to excess demand could be alleviated through a number of low-cost policy measures. While the diagnosis of Italy’s overall situation in terms of the relative importance of supply and demand factors is a difficult empirical issue, there is evidence of some instances of excess demand (Marchesi (2003)). Work by the author suggests that Italy could still benefit from some specific reforms, including: (i) unification of the legal interest rate with that of the market and (ii) de-linking attorneys’ compensation from the number of “activities” such as court sessions, in favor of a flat-fee principle. A broader simplification of legal and administrative procedures related to the legal process, including in the context of specific laws (e.g., bankruptcy procedures), would also be consistent with the objective of rationalizing the demand for legal services and speeding up transactions. “Supply-side” issues such as incentives for judges to boost productivity and/or work hours should also not be neglected. However, the merits of particular proposals should be assessed by experts, and is beyond the scope of this paper.
Cross-country evidence on the European Union.
97. Several applied studies have examined empirical evidence on adverse effects of the length of Italy’s legal processes on its financial and aggregate economic activity. Thus, Fabbri (2001) has shown, on the basis of panel data for 1970-1995, that Italian regions with lengthier legal enforcement have had lower capital stocks and lower value added. The author traces this effect to the enterprise level, as in regions with lengthier legal enforcement firms face tighter financing constraints. The latter conclusion is shared by Bianco, Japelli and Pagano (2002), who demonstrate that lending-to-GDP ratios are negatively related to the length of trials, based on panel data for 95 Italian provinces from 1984 to 1995. In what follows, we explore the length of legal enforcement in a cross-country setting, but, as distinguished from the above business environment analysis, focus on the length and costs dimensions of the broader legal system instead of the narrower “debt contract enforcement.”
98. We illustrate some tentative relationships between “objective” legal and other institutional factors, as well as economic performance, for a cross-section of European countries. In addition, we compare results obtained with the widely-used subjective indexes on the quality of the legal process with “objective” indicators derived from the length and cost of the legal proceedings as compiled by the European commission (see above). We focus on the long-term indicators, assuming that the characteristics of the legal system are relatively stable over time. The underlying data are taken from a variety of sources, including from various academic publications by La Porta et al. (1997-2000) and Beck et al. (2001). The subjective indicators have been mostly drawn from the World Economic Forum’s Global Competitiveness Report. The small number of country observations for the legal data do not allow meaningful statistical tests, but permit consideration of some broad patterns.
99. In European countries, the legal system’s efficiency in terms of cost and length of proceedings is correlated with indicators of economic and financial performance. The first panel in Figure 9 shows that there is a negative, albeit weak, relationship between market capitalization51 and the length of civil legal proceedings. Interestingly, this negative relationship becomes much stronger (Figure 9, second panel) if the “length” variable is replaced with the ratio of length to cost (expressed as a percentage of case value, including legal assistance and procedural expenses). As argued above, increasing the cost of legal proceedings decreases the total demand for legal services, including “pathological” or “petty” demand. To the extent higher legal costs deter pathological demand, the higher cost may be “efficient.” The above evidence appears to suggest that this “efficiency” effect may be pertinent to the cross-country variation, in that for a substantial subset of countries the legal procedures may be “long enough” for higher costs to be “efficient.” A similar negative relationship emerges between the length/cost ratio and intermediary credit (Figure 9, third panel). Given only 14 observations, it is pointless to examine statistical significance or more complex relationships. However, the nature and the basic strength of these relationships is not dependent on any single outlier observation.
Figure 9.Judicial Efficiency (1996) and Long-term Financial and Economic Indicators
100. There is also a negative relationship between the length/cost ratio and per capita GDP (bottom panel of Figure 9). As with the above analysis of the business environment indicators, there are no clues about causality, but it is still instructive that higher-income countries in practice tend to have a fairly small ratio of the length of legal proceedings to their costs.
101. For the same sample of European countries, the subjective indicators of the legal system appear to play a similar role than that of objective indicators. The positive relationship of the GCR’s “judicial efficiency” and the “rule-of-law” variables, as well as the extent to which cross-country variation is accounted for, is similar for these subjective variables with respect to market capitalization and intermediary credit (results not shown). At the same time, the correlation with GDP per capita (Figure 10) is greater for the subjective judicial indicators than for the objective ones. Again, this may be coincidental, but if not, one possible reason is the suspicion that economic outcomes like GDP are implicitly factored into subjective data.
Figure 10.Subj. Judicial Efficiency (2002) and GDPpc
102. These correlations suggest that the evolution of objective legal system characteristics could be pertinent to understanding the relationship between institutions and economic performance. Obviously, no statistical significance or causality can be established on the basis of a small sample and lack of consistent time series data. Still, the evidence is suggestive and, importantly, not very sensitive to the timing of particular cross-country comparisons, as most of these relationships tend to be stable over time. Further research on this matter is intuitively promising in view of (i) robust empirical results yielded by the “legal tradition” and (ii) the more long-term nature of the legal system compared to other institutional factors, as suggested in the arguments above. As a practical matter, Italy’s very lengthy legal processes could thus be quite taxing in terms of economic performance, not least with respect to its EU peers.
103. This chapter has investigated the role of Italy’s institutional and business environment, including the legal system, in its broad economic performance, against the background of a recent surge in academic contributions. A literature survey has been necessary to put the problem in context, as the theme is becoming ever more “macro-relevant.” The survey prompts the following conclusions: (i) despite growing interest, so far the studies have yet to converge on a mechanism linking institutions and aggregate economic outcomes; (ii) at a theoretical level, the political economy view appears the most plausible, but is not operational, (iii) the “legal tradition” view has had some empirical validation, but it underestimates even the legal system’s role in economic performance; (iv) for a better understanding of this role, one at the very least needs also to capture the dynamics and modern characteristics of various institutions, not least the business environment and the legal system; (v) in doing the latter, more empirical work is necessary, which should be based on objective data and cover as wide a field of institutional issues as possible.
104. An analysis of the recent cross-country data highlights the importance of institutions related to the business environment. It is generally the case that slower and costlier procedures tend to be correlated across all, as well as among high-income, economies between and within different sectors, and are associated with lower levels of per capita output. The effect of the negative role of “contract enforcement” inefficiencies is also detectable, but is somewhat weaker. There is also some evidence that, while the business environment depends on exogenous fundamentals such as the legal origin (see La Porta et al. (1997-2000)), it may also affect growth outcomes regardless of the legal origin. For example, “more stable” business environment factors appear to have some statistical power in explaining cross-country variation in past growth rates, which is clearly not the case for more “short-term” factors. The causality issue and the particular elasticities still need to be approached formally as time series data become available.
105. Measures of Italy’s business environment suggest there is a case for progress in some areas. Given the cross-country evidence of the link between the business environment and economic performance, Italy could gain from institutional reforms. Progress is needed in many directions, partly in view of the likely complementarities, with the length of contract.
106. Italy’s civil legal procedures are characterized by their extreme length. In the European Union, the ratio of length to the cost of trials is negatively correlated with various measures of economic and financial development. The evidence is not compelling due to data availability problems, but to the extent there was a causal link from legal institutions to economic performance, the data would enhance arguments for giving policy priority to reducing the length of trials while preserving the rights of participants and other core principles. Among other things, progress in this area could limit “pathological” demand for legal services. The difficult empirical issues of the relative weights of normal and pathological demand for justice, as well as feasibility of steps to speed up legal proceedings, would need to be addressed carefully by experts to assess particular trade-offs in this context.
107. A number of specific and generally low-cost reforms in all areas (some of which are ongoing) could help improve Italy’s economic potential. Regarding the business environment, a key issue is a sweeping simplification of administrative procedures. Such steps may build on progress made with the three “Bassanini laws” of 1997-1998. The forthcoming discussion of the draft “competitiveness” legislation anticipated in the 2005-2008 Documento di Programmazione Economica e Finanziaria provides a valuable opportunity to advance and complete the substantial legislative agenda. In addition, a draft “simplification” law would include important measures to liberalize many firm-related procedures, while the second part of the Biagi reform would solidify progress in the labor market. Another critical issue is the adoption of a long-delayed modern bankruptcy law that would speed up and streamline firm closure procedures. Regarding the legal system, work by a number of researchers, including Marchesi (2003) – mostly focused on the demand side – and Zan (2003) on the organizational side, suggests some specific options to deal with inefficiencies at various levels, although an evaluation of their merit and the likely impact must be left to the domain of legal scholars.
AcemogluD., JohnsonS., and J.Robinson (2004) “Institutions as the Fundamental Cause of Long-Run Growth,” NBER Working paper, 10481, May.
Bank of Italy (2004) “Relazione del Governatore sull’esercizio 2003,” Rome, May.
BarroR. (1991) “Economic growth in a Cross-Section of Countries,” Quartrely Journal of Economics, 106 (2), May, 407–443.
BarroR. and X.Sala-i-Martin (1995) “Economic Growth,” McGraw Hill, N.Y.
BeckT., Demirgüç-KuntA and R.Levine (2001) “Legal Theories of Financial Development,” Oxford Review of Economic Policy, Vol. 17, No 4, 483–501.
BeckT., Denirgüç-KuntA., and R.Levine (2004) “Law and Firms’ Access to Finance,” NBER Working paper, 10687, August.
BiancoM. and S.Giacomelli (2005), Efficienza della giustizia e imprenditorialita’: il caso italiano, forthcoming inEconomia e Politica Industriale.
BiancoM., JapelliT., and M.Pagano (2002) “Courts and Banks: Effect of Judicial Enforcement of Credit Markets,” CEPR Discussion paper, No 3347, April.
BoteroJ., DjankovS., La PortaR., Lopez-de-SilanesF. and A.Shleifer (2004) “Regulation of labor,” mimeo, May.
CioccaP. (2003) “The Italian Economy: A Problem of Growth,” Banca d’Italia, Economic Bulletin, 37, 145–158, November.
DecressinJ. (2000) “Puzzling Out Italy’s Growth Performance,” IMF Staff Country Report 00/82, July, 6–27.
DjankovS., La PortaR., Lopez-de-SilanesF., and A.Shleifer, (2002) “Regulation of Entry” Quarterly Journal of Economics, 117, 1–37, February.
DjankovS., La PortaR., Lopez-de-SilanesF., and A.Shleifer, (2003) “Courts,” Quarterly Journal of Economics, May.
DjankovS., McLeishC. and A.Shleifer (2004), “Private Credit in 129 Countries,”mimeo, November.
FabbriD. (2001) “Legal Institutions, Corporate Governance, and Aggregate Activity: Theory and Evidence,” CSEF Working Paper, No 72, University of Salerno, October.
FabbriD. and M.Padula (2001) “Judicial Costs and Household Debt,” CSEF Working Paper No 65, University of Salerno, August.
GlaeserE., La PortaR., Lopez-de-SilanesF. and A.Shleifer (2004) “Do Institutions Cause Growth?” NBER Working paper, No 10568, June.
GuisoL. (2004) “Quanto e difficile per una piccola impresa italiana accedere al finanziamento esterno?,” Chapter 3 in de Caprariis G and L. Guiso “Finanza, Legge e Crescita,” Il Sole 24 Ore, Milano, April.
HausmanR., PritchettL. and D.Rodrik (2004) “Growth Accelerations,” NBER Working paper No 10566, June.
IchinoA., IchinoP., and M.Polo (1998) “Il mercato del lavoro e le decisioni dei giudici sui licenziamenti” inS.Cassese and G.Galli (Eds.) L’Italia da semplificare. Le istituzioni. Bologna. Il Mulino.
ISTAT (2004) “Rapporto Annuale: La situazione del paese nel 2003”ISTAT, Rome, May.
KaufmanD., KraayA. and P.Zoido-Lobaton (1999) “Aggregate Governance Indicators,” World bank Policy Research Working Paper
KaufmanD. (2004) “Governance Redux: The Empirical Challenge,” Chapter 2.5 (137-164) in Global Competitiveness Report, 2003-2004, Oxford University Press, New York, N.Y.
KlapperL, LaevenL. and R.Rajan (2004) “Barriers to Entrepreneurship,”mimeo, November.
LaPortaR., ShleiferA., and R.Vyshni (1997) “Legal Determinants of External Finance,” Journal of Finance, 52(3), 1131–50.
LaPortaR., ShleiferA., and R.Vyshni (1998), “Law and Finance,” Journal of Political Economy, 106 (6), 1113–55.
LaPortaR., ShleiferA., and R.Vyshni (1999), “The Quality of Government,” Journal of Law, Economics, and Organization, 15(1), 222–79.
LaPortaR., ShleiferA., and R.Vyshni (2000) “Investor Protection and Corporate Governance,” Journal of Financial Economics, 58(1/2), 3–29.
LevineR. (1997) “Financial Development and Economic Growth: Views and Agenda,” Journal of Economic Literature, 35, 688–726.
LevineR. (1998), “The Legal Environment, Banks, and Long-Run Economic Growth,” Journal of Money, Credit, and Banking, 30(2), 596–620.
LevineR. (1999) “Law, Finance, and Economic Growth,” Journal of Financial Intermediation, 8(1/2), 36–67.
LevineR. and S.Zervos (1998) “Stock Markets, Banks, and Economic Growth” American Economic Review, 88(3), 537–58.
LevineR., LoyazaN., and T.Beck (2000) “Financial Intermediation and Growth: Causality and Causes,” Journal of Monetary Economics, 46(1), 31–77.
LombardoD. and M.Pagano (2000) “Law and Equity Markets: A Simple Model,” Csef Working Paper, N. 25.
MarchesiD. (1998) “L’inefficienza della Giustizia Civile. Conseguenze sull’Economia,” inS.Cassese and G.Galli (Eds.) L’Italia da semplificare. Le istituzioni. Bologna. Il Mulino.
MarchesiD. (2003) “Litiganti, Avvocati e Magistrati: Diritto ed Economia del processo civile,”Il Mulino, Bologna, Italy.
MarchesiD. (2004) “Regole per l’insolvenza e incentivi per la crescita dimensionale delle PMI,” in de Caprariis G and L. Guiso “Finanza, Legge e Crescita,” Il Sole 24 Ore, Milano, April.
MarchesiD. and C.Pappalardo (2004) “Competitivita delle imprese distrettuali e la rilevanza del enforcement dei contratti,” Chapter 5 in “Rapporto trimestrale: priorita nazionali: trasparenza, flessibilita, oportunita,”April.
PaganoM. and P.Volpin (2001) “The Political Economy of Finance,” Oxford Review of Economic Policy, Vol. 17, No 4, 502–518, Winter.
PerottiE. (1995) “Credible Privatization,” American Economic Review, 85(4), 847–59.
RajanR. and L.Zingales (2003) “The Great Reversals: The Politics of Financial Development in the 20th Century,” Journal of Financial Economics, Vol 69, 1, 5–50, July.
RigobonR. and D.Rodrik (2004) “Rule of Law, Democracy, Opennes and Income: Estimating the Interrelationships,” NBER Working Paper 10750, September.
RodrikD., SubramanianA. and F.Trebbi (2002) “Institutions Rule: The Primacy of Institutions over Geography and Intergration in Economic Development,” NBER Working Paper, No 9305, November.
RodrikD. and A.Subramanian (2004) “From Hindu Growth to Productivity Search: The Mystery of the Indian Growth Transition,” NBER Working Paper No 10376, March.
RoeM. (2003) “Political determinants of Corporate Governance: Political Context, Corporate Impact,”New York, Oxford University Press.
ScarpettaS., HemmingsP., TresselT., and J.Woo (2002) “The Role of Policy and Institutions for Productivity and Firm Dynamics: Evidence from Micro and Industry Data,” OECD Working paper No 329, April.
SchivardiF. and R.Torrini (2004) “Firm Size Distribution and EPL in Italy,”Bank of Italy, mimeo.
SolowR. (1956) “A Contribution to the Theory of Economic Growth,” Quarterly Journal of Economics, 70, 1 (February), 65–94.
WEO (2003) “Growth and Institutions” in Chapter 3 in IMF World Economic Outlook95–128, April.
ZanS. (2003), Fascicoli e tribunali, Il processo civile in una prospettiva organizzativa, Il Mulino.
ZingalesL. (2004) “L’evoluzione del settore finanziario in Europa e le sfide per il sistema Italiano,” Chapter 1 in de Caprariis G and L. Guiso “Finanza, Legge e Crescita,” Il Sole 24 Ore, Milano, April.
ZingalesL, RajanR. and M.Kumar (1999) “What Determines Firm Size?,” CEPR discussion paper No. 221.
ZuckermanA. S. (1999) “Civil Justice in Crisis,”Oxford, Oxford University Press.
Prepared by Bogdan Lissovolik.
According to this view, the French civil law system offers the least support for financial development, with the German and Scandinavian systems falling into an intermediate range.
Empirical efforts at discriminating between the political-economy and legal origin arguments have not been fully conclusive, highlighting both the pervasive data problems. Kaufman (2004) infers that the legal tradition appears to matter, but the share of cross-country variation explained by the legal origin is low. He concludes that there are important lessons “from a comparison across legal systems that focuses on more than the (admittedly important) distinction between civil and common law systems.”
For example, Rodrik and Subramanian (2004) note, in the context of India, that “growth was triggered by an attitudinal shift on the part of the national government towards a pro-business … approach. (p. 1)” Hausman et al. (2004), interpret their comprehensive cross-country study of growth accelerations as likely pointing in a similar direction (although the hypothesis was not tested directly).
The work underlying the compilation of many indicators has been restricted to, or dominated by, the most populous cities. The indicators also reflect various transaction-specific adjustments to permit comparability, and often are restricted to some specific types of transactions. Several of the indicators are subjective and do not permit a wide range of scores.
Actually, rankings based on the business environment data may have a slight favorable bias, since for many indicators a few country observations are missing.
However, as mentioned in the staff report for this Article IV consultation, other indicators give a more favorable assessment of Italy’s labor market environment. As explained below, although the World Bank data refer to January 2004, some of the most recent reforms may not have been captured in the data. Also, the staff report compares Italy to the Euro area, which on average has somewhat lower scores than the whole group of OECD countries.
One exception to this pattern has been the time of closing a business, which was shorter in Rome than in much of the country. This may partly explain why Italy’s position within the dataset in terms of the time of business closure appeared overly favorable to independent experts.
The measurement of costs is indicator-specific, usually it is in terms of the value of the specific asset involved in the transaction, but also in terms of per capita income (i.e, for firm start-ups). It is however possible that the correlation properties may derive from explicit or implicit measurement of most cost indicators in terms of per capita GNI. In this case, if there is a fixed element on the cost side of the indicators across all countries, it could by itself determine the negative correlation with per capita GNI. However, it is not clear that such a fixed-cost element would be present for these cost indicators, it appears that the most cost-intensive procedures related to registration and licencing are non-tradables and thus would not have a fixed cost element.
In fact, it would be difficult to expect a large value of the coefficient for each separate institutional factor, partly because a given factor refers to a relatively narrow area and partly because the limitations of data availability over time may well weaken the measured statistical link. In any case, we do not want to emphasize the particular elasticities at this point, but rather focus on the signs and the relative strengths of the relationships.
Another important caveat is an empirical verification (based on as yet unavailable time series data) of whether the intutitively “long-term” factors such as contract enforcement and the recovery rate are indeed more “long-term” than such factors as firm start-ups and labor regulations.
For example, it is puzzling why Switzerland appears to have the recovery rate of only 37 cents on the dollar while doing very well on the remainder of World Bank indicators. Brazil is another country for which the measured recovery rate (close to zero) appears highly questionable.
Incidentally, the recovery rate is very high (around 90 percent or higher) for a good share of established “economic miracle” countries, including Korea, Singapore, Ireland, Taiwan, etc. For other important miracle cases (China, Thailand, Malaysia) it is not high, but perceptibly higher than for their comparators in the same per capita income group.
In the law and economics literature the term “legal system” has been used ambiguously. On the one hand, as in this paper, it is used to denote the thrust of civil law institutions, mainly the system of basic legal principles and their enforcement (and thus it will be closer to the term “judicial system”). In other contexts, the term “legal system” is understood to also include specific legislation (company law, bankruptcy law, etc.), which generally goes beyond the scope of this paper.
For example, countries with a particular legal system have been disproptionately represented within some geographical regions and cultures (French civil law system in Latin America).
In France, for example, many agents decide to bypass the Cassation Court (third degree) and refer to alternative means of settling disputes (arbitration, etc.).
Unfortunately, there have been no consistent cross-country updates of these data, which appear also in part survey-based, but again measured in “objective” time units.
Total demand for legal transactions would be given by the sum of a downward-sloping normal (physiological) demand for transactions and pathological demand. The final shape of the aggregate demand curve is a difficult and outstanding empirical issue, but in any case it may be reasonable to assume that for very long trials the demand curve would “bend” and become upward-sloping.
Again, if the supply curve were horizontal, for example, assuming that judges could only operate in a “fixed setting,” the basic story would not change.
In practice, alternative justice arrangements are asymmetric, as they affect certain stages of the legal process more than others. In particular, it is believed that they would mostly concern enterprise bankruptcies, which are by far the most time-consuming among all civil procedures.
A firm-level cross-country study by Klapper et al. (2004) emphasized Italy’s bureaucratic barriers to entrepreneurship as another causal reason. A deeper investigation of the empirical evidence (i.e., firm turnover) is hampered by the data. In particular, the information on enterprise entry and exit does not generally permit to discriminate between various specific reasons for these activities (reorganization due to tax reasons versus genuine entry/exit).