Does corruption ever help entrepreneurship?

The argument that corruption may improve the efficiency of a political process from a standpoint of economic efficiency has its roots in the fact that votes or political preferences sometimes do not accurately convey the full economic value of alternatives (see, e.g., Tidemann and Tulloc 1976).

While most articles approach the subject from the standpoint of viewing corruption as the exogenous determinant of entrepreneurial activity, Tonoyan et al. (2010) examine the choice of entrepreneurs to engage in bribe-paying schemes in these economies. They examine the role of both formal and informal institutions on entrepreneurs as bribe payers and find that the likelihood of engaging in this practice is influenced by the lower efficiency of financial and legal institutions and the lack of their enforcement.

See the literature discussed earlier, including Shleifer and Vishny (1993), Ehrlich and Lui (1999), Mauro (1995), La Porta et al. (1999), Treisman (2000), Busenitz et al. (2000) and Glaeser and Saks (2006).

PCA is a multivariate statistical technique used to examine relationships among different quantitative variables. Mathematically speaking, if there are n correlated variables, PCA generates uncorrelated indices or components, where each component is a linear weighted combination of the n variables. For example, for a set of variables \( X_{1} \ldots .X_{n} , \)

\( \begin{array}{*{20}c} {PC_{1} = w_{11} X_{1} + w_{22} X_{2} + \cdots \cdots \cdots + w_{1n} X_{n} } \\ \vdots \\ {PC_{m} = w_{m1} X_{1} + w_{m2} X_{2} + \cdots \cdots \cdots + w_{mn} X_{n} } \\ \end{array} \) where w _mn represents the weight for the mth principal component and the nth variable. These weights are the eigenvectors of the covariance matrix (since we have standardized our data; otherwise it is the correlation matrix). The eigenvalue of the corresponding eigenvector is the variance (σ) for each principal component. The first principal component, PC ₁, explains the largest possible variation in the dataset subject to the constraint, \( \sum\nolimits_{i = 1}^{n} {w_{1i}^{2} } \). Since the sum of the eigenvalues equals the number of variables in the original dataset, the proportion of total variation accounted for by each principal component is the ratio \( \frac{{\sigma_{i} }}{n} \).

Standardized values for each component have been used to generate the index—thus, the index ranges from negative to positive figures. A more negative number implies less business regulation.

According to Roodman (2009), GMM dynamic panel estimators are particularly suited for (1) small “T” (fewer time periods) and large “N” (many individual or country) panels, (2) a linear functional relationship, (3) a single dependent variable that is dynamic, depending on its own past realizations, (4) independent variables that are not strictly exogenous and are correlated with present as well as past realizations of the error, (5) country fixed effects and (6) heteroskedasticity and autocorrelation within countries.

SD of corruption = 1.04.

Based on World Bank definition, net enrollment ratio ‘is the ratio of children of official school age based on the International Standard Classification of Education 1997 who are enrolled in school to the population of the corresponding official school age’ (WDI 2013). Primary completion rate to the completion of primary education. The figures are expressed as a percentage of the relevant age group.

We further test our results to the inclusion of different indicators of economic freedom. Studies like Bjørnskov and Foss (2008) have shown that certain indicators of economic freedom affect entrepreneurial initiatives of nation. The different measures of economic freedom we consider from Heritage database are measures of property rights, fiscal freedom, government spending, labor freedom, monetary freedom, trade freedom and financial freedom. The coefficient on the interaction term remains positive and significant for all the alternate specifications. The coefficient of corruption is negative and significant for all the alternate specifications. The estimated marginal effects are similar to our benchmark results. The coefficients of most indicators of economic freedom are significant, indicating that they do affect entrepreneurship for nations.