Coase Lecture - The Inverted-U Relationship Between Credit Access and Productivity Growth
GAICIL Insights
Cheap credit is usually sold as pure fuel for innovation.1 This paper finds it is also life support for firms that should be leaving. Coase Lecture - The Inverted-U Relationship Between Credit Access and Productivity Growth, by Philippe Aghion, Antonin Bergeaud, Gilbert Cette, Rémy Lecat and Hélène Maghin, identifies two counteracting effects of credit access on productivity growth. Better access makes it easier for entrepreneurs to innovate. Better access also lets less efficient incumbents stay longer in the market, discouraging entry of new and potentially more efficient innovators.
Here is the uncomfortable test for crisis-era lenders and regulators: the next round of emergency credit should be judged by whom it keeps alive as well as by how many. In an AI transition, rescuing the least productive firms may be the costliest form of help.
A model of firm dynamics and innovation-based growth with credit constraints turns these effects into an inverted-U relationship between credit access and productivity growth. On a comprehensive French manufacturing firm-level dataset, the authors find an inverted U at the sectoral level. At the firm level, incumbents with easier credit access experience higher productivity growth and lower exit rates, particularly the least productive among them. The same loan can sharpen a good firm and preserve a weak one.
Causality comes from the 2012 Eurosystem's Additional Credit Claims program, which generated an exogenous extra credit supply for a subset of incumbent firms. Treated firms saw higher productivity growth after the program than similar control firms, and lower exit rates, especially the least productive treated firms. Credit access causally raises incumbents' productivity growth while slowing the exit of weak firms.
The inverted U is a fundamental trade-off. Some credit access is essential for innovation; too much, at least of the kind that keeps unproductive incumbents alive, impedes the reallocation driving aggregate productivity. The optimal level balances these forces. Finance has a sweet spot, and past it, money starts protecting the past.
Artificial intelligence raises the stakes of that sweet spot. AI adoption and development require investment, and credit helps finance it. Credit supporting innovation raises productivity. Credit sustaining unproductive firms may slow the reallocation of resources toward firms using AI most effectively. Financial policy can influence the pace at which AI transforms the economy. Every zombie firm holds workers and capital that an AI adopter could have used.
Reallocation matters most during technological change. As AI raises adopting firms' productivity, resources should flow toward them and away from firms failing to adapt. Exit of unproductive incumbents frees resources and creates space for entrants. Credit policies delaying exit slow the process and reduce aggregate productivity growth.
Nuance is required. Restricting credit broadly would harm innovation by constraining firms that would invest productively. The aim is directing credit toward productive uses while allowing unproductive firms to exit. Better screening, bankruptcy procedures facilitating efficient exit and support for young innovative firms can achieve this balance. The tool is precision.
The quasi-experiment strengthens the case. Exploiting a program that expanded credit for some firms and not others identifies causal effects on productivity and exit, the kind of careful evaluation credit policy needs.
For AI, the financing of new entrants deserves attention. Startups developing AI applications often face credit constraints. Ensuring their access to finance while preventing misallocation of credit to unproductive incumbents supports innovation and the renewal of the economy. A banking system that lends easily to the old and grudgingly to the new is betting against the future.
Distilled to one claim: credit access affects productivity growth through two channels, innovation by incumbents and exit of unproductive firms. Too little credit stifles innovation; too much can impede reallocation. As AI drives a new wave of productivity growth, financial policies balancing these forces will help resources flow toward the most productive uses and give new innovators room to grow.
Introduction
We identify two counteracting effects of credit access on productivity growth: on the one hand, better access to credit makes it easier for entrepreneurs to innovate; on the other hand, better credit access allows less efficient incumbent firms to remain longer on the market, thereby discouraging entry of new and potentially more efficient innovators. We first develop a simple model of firm dynamics and innovation-based growth with credit constraints, where the above two counteracting effects generate an inverted-U relationship between credit access and productivity growth. Then we test our theory on a comprehensive French manufacturing firm-level dataset. We first show evidence of an inverted-U relationship between credit constraints and productivity growth when we aggregate our data at the sectoral level. We then move to firm-level analysis, and show that incumbent firms with easier access to credit experience higher productivity growth, but that they also experience lower exit rates, particularly the least productive firms among them. To support these findings, we exploit the 2012 Eurosystem's Additional Credit Claims programme as a quasi-experiment that generated an exogenous extra supply of credits for a subset of incumbent firms.
Takeaway
In this paper we have identified two counteracting effects of credit access on productivity growth: first, a positive direct effect of credit access on (incumbent) firms’ productivity growth which reflects the fact that better access to credit makes it easier for entrepreneurs to innovate; second, a negative reallocation effect of credit access working through the exit rate of incumbent firms and its effect on entry of potentially more efficient innovators. We developed a simple model of firm dynamics and innovation-led growth with credit constraints, to generate these two effects and then we showed that the combination of the two effects can result in an aggregate invertedU relationship between credit access and productivity growth. We then used a comprehensive French firm-level dataset to first provide evidence of such an inverted-U relationship in a crosssector panel regression. Then we moved to firm-level analysis to provide supporting evidence of the two counteracting effects of credit access on aggregate productivity growth, respectively working through the productivity growth and the exit rates of incumbent firms. Finally, we used the 2012 Eurosystem’s Additional Credit Claims (ACC) program as a quasi-experiment, to argue that the effects of credit access on the productivity growth and exit rates of incumbent firms are causal: namely, incumbent firms directly affected by this program - the treated firms experienced higher productivity growth post-ACC that similar firms not affected by the program - the control firms -; moreover, treated firms also experienced lower exit rates, particularly the least productive firms among them, compared to control firms between before and after the introduction of the ACC program.1
A first extension of the analysis in this paper, would be to look at firms and sectors in other countries: in which countries and/or sectors are we most likely to observe the downward-sloping part of the inverted-U? A second extension would be to analyze the implications of this inverted-U relationship for the conduct of fiscal and monetary policy over the business cycle. Our analysis points to a potentially negative growth effect of maintaining low interest rates in good times. Third, our analysis also has implications for the debate on secular stagnation. The decline in productivity growth in most advanced countries since the 1970s may indeed be partly related to an overall easier access to credit due to financial liberalization over the period. This mechanism may have been amplified by the decrease of interest rates and the capital abundance observed in the last decade. The increase of real interest rate expected in the recovery phase could then contribute to productivity gains through cleansing mechanisms. Hence, a natural next step would be to test the relationship between credit access and productivity growth over longer time-periods, and to look at how much of the observed productivity slowdown can actually be explained by the observed decrease in interest rates since the 1970s. These and other extensions of the analysis in this paper, are left for future research.
Notes
- Philippe Aghion et al., Coase Lecture - The Inverted-U Relationship Between Credit Access and Productivity Growth, 86 Economica 1 (2019), https://doi.org/10.1111/ecca.12297. Takeaway source: Conclusion, reproduced from a free author or repository copy. ↩
Cite as: Philippe Aghion et al., Coase Lecture - The Inverted-U Relationship Between Credit Access and Productivity Growth, 86 Economica 1 (2019), https://doi.org/10.1111/ecca.12297
Header image: Pont de Brooklyn de nuit - Octobre 2008 edit.jpg by Martin St-Amant (S23678), CC BY 3.0, via Wikimedia Commons.
