| A. Comparison of means across samples . | |||||
|---|---|---|---|---|---|
| . | PS-matched sample year t-1 . | ||||
| . | Treatment group (obs. = 568) . | Control group one match (obs. = 568) . | Control group two matches (obs. = 1,026) . | Difference (treatment-control one match) . | Difference (treatment-control two matches) . |
| Propensity score | 0.560 | 0.563 | 0.550 | –0.003 | 0.010 |
| Implied contract | 0.664 | 0.658 | 0.724 | 0.005 | –0.060 |
| Public policy | 0.787 | 0.713 | 0.782 | 0.074 | 0.005 |
| ln(Assets) | 5.125 | 5.823 | 5.906 | –0.698* | –0.781* |
| Tobin’s q | 1.323 | 1.374 | 1.334 | –0.052 | –0.011 |
| Cash flow | 0.079 | 0.092 | 0.087 | –0.013 | –0.008 |
| Cash holdings | 0.094 | 0.087 | 0.090 | 0.007 | 0.004 |
| Book leverage | 0.272 | 0.283 | 0.287 | –0.011 | –0.015 |
| Firm age | 12.930 | 19.778 | 20.286 | –6.849 | –7.356 |
| P.C. GDP growth | 0.008 | 0.000 | 0.000 | 0.008 | 0.008 |
| ln(P.C. GDP) | 3.560 | 3.502 | 3.510 | 0.058 | 0.050 |
| Political balance | 0.641 | 0.678 | 0.661 | –0.038 | –0.020 |
| A. Comparison of means across samples . | |||||
|---|---|---|---|---|---|
| . | PS-matched sample year t-1 . | ||||
| . | Treatment group (obs. = 568) . | Control group one match (obs. = 568) . | Control group two matches (obs. = 1,026) . | Difference (treatment-control one match) . | Difference (treatment-control two matches) . |
| Propensity score | 0.560 | 0.563 | 0.550 | –0.003 | 0.010 |
| Implied contract | 0.664 | 0.658 | 0.724 | 0.005 | –0.060 |
| Public policy | 0.787 | 0.713 | 0.782 | 0.074 | 0.005 |
| ln(Assets) | 5.125 | 5.823 | 5.906 | –0.698* | –0.781* |
| Tobin’s q | 1.323 | 1.374 | 1.334 | –0.052 | –0.011 |
| Cash flow | 0.079 | 0.092 | 0.087 | –0.013 | –0.008 |
| Cash holdings | 0.094 | 0.087 | 0.090 | 0.007 | 0.004 |
| Book leverage | 0.272 | 0.283 | 0.287 | –0.011 | –0.015 |
| Firm age | 12.930 | 19.778 | 20.286 | –6.849 | –7.356 |
| P.C. GDP growth | 0.008 | 0.000 | 0.000 | 0.008 | 0.008 |
| ln(P.C. GDP) | 3.560 | 3.502 | 3.510 | 0.058 | 0.050 |
| Political balance | 0.641 | 0.678 | 0.661 | –0.038 | –0.020 |
| B. Effect of the adoption of the good faith exception . | ||||
|---|---|---|---|---|
| . | PS-matched sample (one match) . | PS-matched sample (two matches) . | ||
| . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . |
| . | (1) . | (2) . | (3) . | (4) . |
| Treatment | 2.11 | –2.39 | 1.71* | –3.01 |
| (1.48) | (8.63) | (0.98) | (4.46) | |
| Treatment |$\times$| Post | –1.65*** | 0.62 | –1.21** | –1.43 |
| (0.57) | (3.74) | (0.49) | (3.70) | |
| Post | 0.11 | 1.58 | –0.01 | 2.66 |
| (1.41) | (7.49) | (0.73) | (5.32) | |
| Control variables | Yes | Yes | Yes | Yes |
| Firm age FEs | Yes | Yes | Yes | Yes |
| Industry |$\times$| Year FEs | Yes | Yes | Yes | Yes |
| Firm FEs | Yes | Yes | Yes | Yes |
| Observations | 7,059 | 7,059 | 10,012 | 10,012 |
| Adjusted |$R^{2}$| | .654 | .391 | .674 | .448 |
| B. Effect of the adoption of the good faith exception . | ||||
|---|---|---|---|---|
| . | PS-matched sample (one match) . | PS-matched sample (two matches) . | ||
| . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . |
| . | (1) . | (2) . | (3) . | (4) . |
| Treatment | 2.11 | –2.39 | 1.71* | –3.01 |
| (1.48) | (8.63) | (0.98) | (4.46) | |
| Treatment |$\times$| Post | –1.65*** | 0.62 | –1.21** | –1.43 |
| (0.57) | (3.74) | (0.49) | (3.70) | |
| Post | 0.11 | 1.58 | –0.01 | 2.66 |
| (1.41) | (7.49) | (0.73) | (5.32) | |
| Control variables | Yes | Yes | Yes | Yes |
| Firm age FEs | Yes | Yes | Yes | Yes |
| Industry |$\times$| Year FEs | Yes | Yes | Yes | Yes |
| Firm FEs | Yes | Yes | Yes | Yes |
| Observations | 7,059 | 7,059 | 10,012 | 10,012 |
| Adjusted |$R^{2}$| | .654 | .391 | .674 | .448 |
This table reports the results from OLS regressions relating corporate investment and sales growth rates to the adoption of the good faith exception using a matched sample and the |$\pm$|3-year window around the adoption of the law. Treatment and control firms are selected based on the propensity that a firm will be in a state that adopts the good faith exception in the following year. The treatment group consists of all firms headquartered in states that adopt the good faith exception the following year that have at least one observation in both the pre- and post-treatment periods in the |$\pm$|3-year window around the adoption of this law. The control group consists of firms headquartered in states that never adopt the good faith exception as well as any firm headquartered in a state that adopts the good faith exception as long as these states did not adopt the law in the |$\pm$|3 years around when its matched state adopts the law. Control firms must have at least one observation in both the pre- and post-treatment periods, be headquartered in a state that borders the treatment firm’s state of headquarters, and be in the same 2-digit SIC industry as the treatment firm. Control firms are matched to treatment firms (with replacement) such that the logit of the treatment and control firms’ propensity scores are within 0.2 standard deviations of each other. In the first (second) matched sample, we match each treatment firm to one control firm (up to two control firms) with the closest propensity score. Panel A tabulates the means of the matched variables for the treatment and control groups in year |$t$|-1. *, **, and *** in the columns labeled “Difference” indicate significance at the 10%, 5%, and 1% levels, respectively, for a |$t$|-test of whether the two samples have equal means. To account for repeated measurement due to some control firms being matched multiple times to different treatment firms in our |$t$|-test comparing the means of the variables across the two samples, we cluster standard errors by firm if it is a firm-level variable and by state if it is a state-level variable. Columns 1 and 2 (3 and 4) of panel B present the results examining the effect of the adoption of the good faith exception on capital expenditures and sales growth using the sample in which each treatment firm is matched to one control firm (two control firms). Treatment is an indicator variable set to one if the firm is headquartered in a state that adopts the law and zero otherwise. Post is an indicator variable set to one in the years after the adoption of the law (same for matched control firms) and zero otherwise. Control variables include Implied contract, Public policy, ln(Assets)|$_{t-1}$|, Tobin’s q|$_{t-1}$|, Cash flow|$_{t-1}$|, Cash holdings|$_{t-1}$|, Book leverage|$_{t-1}$|, P.C. GDP growth|$_{t-1}$|, ln(P.C. GDP)|$_{t-1},$| and Political balance|$_{t-1}.$|Table 2 defines all variables. Industry fixed effects are defined at the 2-digit SIC level. Standard errors in parentheses are clustered by state. *|$p < .1$|; **|$p < .05$|; ***|$p < .01$|.
| A. Comparison of means across samples . | |||||
|---|---|---|---|---|---|
| . | PS-matched sample year t-1 . | ||||
| . | Treatment group (obs. = 568) . | Control group one match (obs. = 568) . | Control group two matches (obs. = 1,026) . | Difference (treatment-control one match) . | Difference (treatment-control two matches) . |
| Propensity score | 0.560 | 0.563 | 0.550 | –0.003 | 0.010 |
| Implied contract | 0.664 | 0.658 | 0.724 | 0.005 | –0.060 |
| Public policy | 0.787 | 0.713 | 0.782 | 0.074 | 0.005 |
| ln(Assets) | 5.125 | 5.823 | 5.906 | –0.698* | –0.781* |
| Tobin’s q | 1.323 | 1.374 | 1.334 | –0.052 | –0.011 |
| Cash flow | 0.079 | 0.092 | 0.087 | –0.013 | –0.008 |
| Cash holdings | 0.094 | 0.087 | 0.090 | 0.007 | 0.004 |
| Book leverage | 0.272 | 0.283 | 0.287 | –0.011 | –0.015 |
| Firm age | 12.930 | 19.778 | 20.286 | –6.849 | –7.356 |
| P.C. GDP growth | 0.008 | 0.000 | 0.000 | 0.008 | 0.008 |
| ln(P.C. GDP) | 3.560 | 3.502 | 3.510 | 0.058 | 0.050 |
| Political balance | 0.641 | 0.678 | 0.661 | –0.038 | –0.020 |
| A. Comparison of means across samples . | |||||
|---|---|---|---|---|---|
| . | PS-matched sample year t-1 . | ||||
| . | Treatment group (obs. = 568) . | Control group one match (obs. = 568) . | Control group two matches (obs. = 1,026) . | Difference (treatment-control one match) . | Difference (treatment-control two matches) . |
| Propensity score | 0.560 | 0.563 | 0.550 | –0.003 | 0.010 |
| Implied contract | 0.664 | 0.658 | 0.724 | 0.005 | –0.060 |
| Public policy | 0.787 | 0.713 | 0.782 | 0.074 | 0.005 |
| ln(Assets) | 5.125 | 5.823 | 5.906 | –0.698* | –0.781* |
| Tobin’s q | 1.323 | 1.374 | 1.334 | –0.052 | –0.011 |
| Cash flow | 0.079 | 0.092 | 0.087 | –0.013 | –0.008 |
| Cash holdings | 0.094 | 0.087 | 0.090 | 0.007 | 0.004 |
| Book leverage | 0.272 | 0.283 | 0.287 | –0.011 | –0.015 |
| Firm age | 12.930 | 19.778 | 20.286 | –6.849 | –7.356 |
| P.C. GDP growth | 0.008 | 0.000 | 0.000 | 0.008 | 0.008 |
| ln(P.C. GDP) | 3.560 | 3.502 | 3.510 | 0.058 | 0.050 |
| Political balance | 0.641 | 0.678 | 0.661 | –0.038 | –0.020 |
| B. Effect of the adoption of the good faith exception . | ||||
|---|---|---|---|---|
| . | PS-matched sample (one match) . | PS-matched sample (two matches) . | ||
| . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . |
| . | (1) . | (2) . | (3) . | (4) . |
| Treatment | 2.11 | –2.39 | 1.71* | –3.01 |
| (1.48) | (8.63) | (0.98) | (4.46) | |
| Treatment |$\times$| Post | –1.65*** | 0.62 | –1.21** | –1.43 |
| (0.57) | (3.74) | (0.49) | (3.70) | |
| Post | 0.11 | 1.58 | –0.01 | 2.66 |
| (1.41) | (7.49) | (0.73) | (5.32) | |
| Control variables | Yes | Yes | Yes | Yes |
| Firm age FEs | Yes | Yes | Yes | Yes |
| Industry |$\times$| Year FEs | Yes | Yes | Yes | Yes |
| Firm FEs | Yes | Yes | Yes | Yes |
| Observations | 7,059 | 7,059 | 10,012 | 10,012 |
| Adjusted |$R^{2}$| | .654 | .391 | .674 | .448 |
| B. Effect of the adoption of the good faith exception . | ||||
|---|---|---|---|---|
| . | PS-matched sample (one match) . | PS-matched sample (two matches) . | ||
| . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . | Capex|$_{t \times}$| 100 . | Sales growth|$_{t \times}$| 100 . |
| . | (1) . | (2) . | (3) . | (4) . |
| Treatment | 2.11 | –2.39 | 1.71* | –3.01 |
| (1.48) | (8.63) | (0.98) | (4.46) | |
| Treatment |$\times$| Post | –1.65*** | 0.62 | –1.21** | –1.43 |
| (0.57) | (3.74) | (0.49) | (3.70) | |
| Post | 0.11 | 1.58 | –0.01 | 2.66 |
| (1.41) | (7.49) | (0.73) | (5.32) | |
| Control variables | Yes | Yes | Yes | Yes |
| Firm age FEs | Yes | Yes | Yes | Yes |
| Industry |$\times$| Year FEs | Yes | Yes | Yes | Yes |
| Firm FEs | Yes | Yes | Yes | Yes |
| Observations | 7,059 | 7,059 | 10,012 | 10,012 |
| Adjusted |$R^{2}$| | .654 | .391 | .674 | .448 |
This table reports the results from OLS regressions relating corporate investment and sales growth rates to the adoption of the good faith exception using a matched sample and the |$\pm$|3-year window around the adoption of the law. Treatment and control firms are selected based on the propensity that a firm will be in a state that adopts the good faith exception in the following year. The treatment group consists of all firms headquartered in states that adopt the good faith exception the following year that have at least one observation in both the pre- and post-treatment periods in the |$\pm$|3-year window around the adoption of this law. The control group consists of firms headquartered in states that never adopt the good faith exception as well as any firm headquartered in a state that adopts the good faith exception as long as these states did not adopt the law in the |$\pm$|3 years around when its matched state adopts the law. Control firms must have at least one observation in both the pre- and post-treatment periods, be headquartered in a state that borders the treatment firm’s state of headquarters, and be in the same 2-digit SIC industry as the treatment firm. Control firms are matched to treatment firms (with replacement) such that the logit of the treatment and control firms’ propensity scores are within 0.2 standard deviations of each other. In the first (second) matched sample, we match each treatment firm to one control firm (up to two control firms) with the closest propensity score. Panel A tabulates the means of the matched variables for the treatment and control groups in year |$t$|-1. *, **, and *** in the columns labeled “Difference” indicate significance at the 10%, 5%, and 1% levels, respectively, for a |$t$|-test of whether the two samples have equal means. To account for repeated measurement due to some control firms being matched multiple times to different treatment firms in our |$t$|-test comparing the means of the variables across the two samples, we cluster standard errors by firm if it is a firm-level variable and by state if it is a state-level variable. Columns 1 and 2 (3 and 4) of panel B present the results examining the effect of the adoption of the good faith exception on capital expenditures and sales growth using the sample in which each treatment firm is matched to one control firm (two control firms). Treatment is an indicator variable set to one if the firm is headquartered in a state that adopts the law and zero otherwise. Post is an indicator variable set to one in the years after the adoption of the law (same for matched control firms) and zero otherwise. Control variables include Implied contract, Public policy, ln(Assets)|$_{t-1}$|, Tobin’s q|$_{t-1}$|, Cash flow|$_{t-1}$|, Cash holdings|$_{t-1}$|, Book leverage|$_{t-1}$|, P.C. GDP growth|$_{t-1}$|, ln(P.C. GDP)|$_{t-1},$| and Political balance|$_{t-1}.$|Table 2 defines all variables. Industry fixed effects are defined at the 2-digit SIC level. Standard errors in parentheses are clustered by state. *|$p < .1$|; **|$p < .05$|; ***|$p < .01$|.
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
