Skip to content

Financial Dependence and Growth

Literature › Works

Raghuram G. Rajan; Luigi Zingales (1996). Financial Dependence and Growth. NBER Working Paper Series, No. 5758. National Bureau of Economic Research.
Type working paper
Scope Comparative and methods
Method econometric
Period studied 1980–1990
Regions 43-country cross-section (developed and developing; includes South Africa as one sample country), United States (benchmark for measuring industry financial dependence)
Themes comparative theory, methods data sources, macro fiscal finance debt
Citation key rajan1996

Argument

The paper asks whether financial development causes growth by testing one mechanism: that it lowers the cost of external finance to firms. Industries that depend more on external finance, measured from US listed firms, should grow faster where financial markets are better developed. The authors regress 1980-90 industry growth in value added on the interaction of US-derived industry dependence and country financial development, with country and industry fixed effects. They find a significant positive interaction and argue it is not explained by omitted variables, outliers or reverse causality. The claim is that ex ante financial development facilitates ex post growth of finance-dependent sectors.

Evidence and method

Compustat firm-level data for US firms in the 1980s (and 1970s) to compute industry external-finance dependence. UN General Industrial Statistics for value added, fixed capital formation and number of firms in manufacturing industries (ISIC). IMF International Financial Statistics for domestic credit, GDP and price indices. IFC Emerging Stock Markets Factbook for 1980 stock market capitalization. Barro-Lee data on schooling and investment share. La Porta et al. legal-origin and judicial-efficiency indices, and CIFAR accounting standards. Final sample of 43 countries and 1,217 industry-country observations.

Findings

  • Finance-dependent industries grow faster in financially developed countries; the effect is about 1 percentage point a year for a one-standard-deviation difference in both variables. (p. 5)

    the difference in growth rates between an industry whose financial dependence is one standard deviation above the mean and the average industry, is 1% more (annually, and in real terms) than in the average country

  • Cash-generating industries do relatively better where financial systems are underdeveloped, consistent with a wedge between internal and external funds. (p. 5)

    industries that generate substantial cash flow from operations grow relatively faster in countries with underdeveloped financial systems.

  • External dependence is defined from US firms' capital expenditure not covered by operating cash flow, taking US large listed firms as a frictionless benchmark. (p. 10)

    dependence on external finance is defined as the ratio of capital expenditures (Compustat # 128) minus cash flow from operations divided by capital expenditures

  • Equity market capitalization has a weaker effect on dependent industries than credit does. (p. 20)

    We find that a dollar of market capitalization is not the same as a dollar of credit; it has only 40% of the effect on the growth of financially dependent firms.

  • Credit to the public sector hurts relative growth of finance-dependent industries. (p. 22)

    the overall impact of a dollar of extra credit, if directed at the public sector, is to reduce the relative growth rate of financially dependent industries

  • The extra growth comes about equally from entry of new firms and expansion in average firm size. (p. 22)

    the incremental growth of financially dependent firms in more developed markets is almost equally driven by a growth in the number of firms and by a growth in average size.

  • The result is robust to using accounting standards as an alternative, more exogenous measure of financial development. (p. 25)

    When we include the interaction between financial dependence and accounting standards instead of our previous measure of financial development (see Table 6, column (IV)), the coefficient is positive and highly statistically significant

  • Financial dependence does not simply proxy for human capital intensity. (p. 26)

    the coefficient on the human capital interaction term is small and not statistically significant, while our fundamental interaction term is virtually unchanged.

  • Using legal origin and judicial efficiency as instruments does not remove the result, which argues against reverse causality. (p. 30)

    As Table 7, column (VI), shows the fundamental correlation persists even when we estimate it using instrumental variables.