Institutions, Productivity, and the Shadow Economy

This is an interesting paper by Axel Dreher, Pierre-Guillaume Méon, & Friedrich Schneider: 

This paper assesses the relationship between institutions output, and productivity when official output is corrected for the size of the shadow economy. Our results confirm the usual positive impact of institutional quality on official output and total factor productivity, and its negative impact on the size of the underground economy. However, once output is corrected for the shadow economy, the relationship between institutions and output becomes weaker. The impact of institutions on total (“corrected”) factor productivity becomes insignificant. Differences in corrected output must then be attributed to differences in factor endowments. These results survive several tests for robustness. 

They explain:

The main rationale behind our results is that weak institutions not lead only to less factor accumulation, but also encourage participation in the shadow economy. The observed negative correlation between weak institutions and official output is therefore driven both by a reduction in production and a switch from the formal to the informal sector. Using official output to estimate the relationship between institutions and output implies that the production of countries with weaker institutions will be underestimated, thereby inflating the observed relationship. As a result, when shadow output is added to official output the correlation weakens. Using official output figures to compute TFP leads to the same bias. Thus, correcting official figures for the shadow economy also weakens the relationship between institutions and TFP, or even goes as far as removing it altogether. The essence of our results suggests that part of the observed relationships reported in the previous literature is not due to a reduction of output, but instead due to a switch from the formal to the informal sector. (p. 138). 

And the paper comes at a time when there seems to be a revival of discussions about GDP as an important statistic of the economy.