Finance and growth : the unpleasant burden of evidence

A growing body of post-global financial crisis (2007-2008) literature documents several undesirable effects of enlarged financial sectors. One of these effects is the 'growth cost' of excessive finance, which reports that the finance-growth relationship is non-monotonic, and that a credit threshold of above 100% of GDP costs economic growth. Since most industrialized countries exceed this threshold by a large margin (reaching as high as 200% in some cases), the policy implications of these findings can hardly be overstated. Moreover, if a tipping point in the finance-growth relationship could be established beyond reasonable doubt then this would be a pathbreaking finding. Extensive, rigorous, and widely replicative empirical evidence-gathered through a unified approach across wide-ranging analytical trajectories-could serve as the 'burden of evidence' and minimize the odds of false positives. We offer such scrutiny regarding three propositions of finance-growth nexus: (i) the inverted U-shaped relationship, (ii) the relevance of financial development for growth, and (iii) the 'vanishing effects.' We analyze fourteen measures of financial development across twenty-two panels-two global datasets and a further twenty country panels based on distinct geographic, economic, and the relative financial development characteristics. The 'burden of evidence' from more than 7,000 well-structured cross-sectional and panel estimates fails to show any robust support to any of these three propositions. We document several other bizarre findings, viz., that the advanced economies need to scale back their relative levels of financial development to those of Eastern Europe to avoid the growth costs associated with overdeveloped financial systems. Surprisingly, the burden of evidence does not support even the widely reported results that financial development is significant under log-linear specifications. This study is unique in cross checking a set of well-accepted empirical results against the 'burden of evidence', and it certainly contests the mainstream cross-country literature. However, this does not disprove the potential role of finance for growth. Alternative approaches that analyze finance and growth at more disaggregated levels by linking sectoral and/or firmlevel production initiatives to their sources of finance may be more effective in unraveling the finance-growth nexus.