I'm looking for empirical evidence (optimally from natural experiments, as described below) of the impact of working capital constraints onto firm output, and the differentials of firm size.

I suppose that generally, stricter working capital constraints are bad for output, and typically these constraints are stronger for smaller firms (measured in output) and younger firms. However, I couldn't manage to find (a) paper(s) to establish that fact yet.

Where by working capital constraint I mean something along the lines of Kiyotaki&Moore (1997)

$$L, K: f(wL, rK) \leq \psi(F(K,L))$$

, for example, for some constant $\psi_0$,

$$wL + rK \leq \psi_0 \cdot F(K,L)$$

where $F(\cdot)$ is the production function, and $f$ and $\psi$ can be whatever.

I have, in other environments, seen that some people take bank liquidity to approximate $\psi$ and argue that for example the last recession was a natural experiment on a change of $\psi$.

So, one type of useful regression table would be (on the firm level)

  • dependent variable: output
  • independent variables:
  • Increase of financial constraint (here, decrease of $\psi_0$)
  • Dummy for firm size (or different size bins)
  • And/Or: Dummy for firm age (or different age bins)
  • additional controls

One example would be Chodorow-Reich (2014, QJE). The two downsides with that paper are that

  • He uses employment, not output as dependent variable (but that's tolerable)
  • He does not actually list the coefficients on the dummies for size. He just includes them as controls.
  • 1
    $\begingroup$ I'm having a hard time understanding what you mean by "working capital constraints." The constraint written seems really general and so I'm having a hard time imagining an example of what you might mean. Could you maybe give a more explicit example? $\endgroup$
    – jmbejara
    Jan 24, 2015 at 5:34
  • $\begingroup$ @jmbejara : I hope my update helped clearing it up $\endgroup$
    – FooBar
    Jan 25, 2015 at 18:21
  • 1
    $\begingroup$ With your equation $$wL + rK \leq \psi_0 \cdot F(K,L)$$, do you have in mind a setting where a firm would like to spend more on inputs to production but is constrained? Because that situation sounds a lot like "time to build" models of productive capacity. $\endgroup$
    – BKay
    Jan 26, 2015 at 19:16
  • $\begingroup$ @BKay Yes, they're constraint, but not due to technology, but rather a liquidity shock, if you will. Cash-in-Advance, with insufficient cash at hands. $\endgroup$
    – FooBar
    Jan 26, 2015 at 19:29
  • $\begingroup$ So, I understand that you're looking for empirical evidence for this kind of thing. Do you have any theory papers that describe in more detail the mechanism that might cause the drop in production? $\endgroup$
    – jmbejara
    Jan 30, 2015 at 4:44

1 Answer 1


I can think of at least one paper where binding accounting rules are shown to have negative consequences on real investment:

Can Tight Accounting Oversight Distort Investment? Evidence from Mortgage Restructurings after the JOBS Act

Abstract: I study the effect of accounting oversight on a bank’s level of troubled mortgage restructurings. If banks manage their accounting numbers (e.g. regulatory capital or earnings) in the wake of the financial crisis, then they may be hesitant to restructure troubled loans because such restructurings require the bank to mark the value of the loans down. I show that giving a bank more accounting discretion leads the bank to restructure more troubled mortgages. My identification strategy uses a provision of the Jumpstart Our Business Startups (JOBS) Act of 2012 that allows a subset of banks to deregister from the SEC. The first stage of the analysis shows that accounting oversight is of first-order importance in the deregistration decision; in particular, there is a sharp increase in the probability of deregistration from 7% to 30% for banks that can escape accounting oversight. By instrumenting for the decision to deregister with the bank’s newfound eligibility to deregister after the JOBS Act, I show that deregistration leads banks to slash their accounting and audit fees and make fewer provisions for loan losses. I next show that deregistration induces banks to double their investment in mortgage restructurings, which entail large upfront accounting expenses; this effect is driven by banks with low capital ratios, and estimates for non-mortgage troubled debt restructurings react similarly. These results suggest that banks manage their capital using real activities, and this is one reason for the low level of mortgage restructurings since the recession.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.