Table 4.

Random Audits and Business Failure – Administrative Costs

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Dependent Variable: Out of Business				Dependent Variable: Defunct
Audited	0.050**	0.070***	0.002	0.119***	0.006	0.018***	0.001	0.023***
	(2.30)	(3.25)	(0.08)	(5.47)	(1.52)	(2.75)	(0.17)	(3.52)
Intercept	0.331***	0.273***	0.328***	0.280***	0.007***	0.012***	0.008***	0.011***
	(22.01)	(18.44)	(21.33)	(19.29)	(2.69)	(3.89)	(3.30)	(3.33)
Sample	Low audit hours	High audit hours	Low audit length	High audit length	Low audit hours	High audit hours	Low audit length	High audit length
N	2,502	2,514	2,504	2,512	2,502	2,514	2,504	2,512
R-Squared	0.003	0.006	0.000	0.016	0.001	0.004	0.000	0.006
Cross Equation Tests on Audited Coefficient
(1) vs (2):	0.020		(5) vs (6):	0.011
	(0.66)			(1.46)
(3) vs (4):	0.117***		(7) vs (8):	0.002***
	(3.79)			(2.96)

This table presents the results of our going concern regression tests using the random audit sample. In each column, we run our main specification on a subset of the sample based on compliance and administrative costs (audit hours or audit length). The specification is a linear probability model, where the dependent variable is Out of Business in columns 1–4 and Defunct in columns 5–8. The independent variable of interest is Audited, which indicates whether a firm was audited as part of the National Research Program. The sample consists of 2508 audited firms and 2508 control firms (one observation per firm). Observations are weighted using the sampling weights (inverse probability of being selected) to generalize the results to the population. All variables are defined in Appendix 1. Robust t-statistics are presented below the coefficients in parentheses. ***, **, * denotes statistical significance at the 10, 5, or 1% level, respectively.