substantial literature, dating back at least to Baumol [1959], Marris [1964], and Williamson [1964], holds that managers may undertake inefficient projects in order to extract private benefits. This problem is particularly severe when managers are entrenched and can resist hostile takeovers [Jensen and Ruback 1983; Shleifer and Vishny 1989]. Under this view, if capital expenditure increases following the adoption of new takeover defenses, this increase would be a net negative for firm value.
To examine the empirical relationship between capital expenditure and governance, we estimate annual median regressions for capital expenditure (CAPEX), scaled by either sales or assets, and net of the industry median. To control for the different investment opportunities available at value and growth firms, we include the log of the book-to-market ratio (BM) as a control variable in all specifications. Table X summarizes the results, with BM coefficients omitted. Columns (1) and (3) give results for the full sample, with G as the key regressor; columns (2) and (4) give results for the sample restricted to firms in the Democracy and Dictatorship Portfolios, with a Democracy dummy as the key regressor. The average coefficient on G is positive and significant in both sets of regressions. Consistent with these results, we find that the average coefficient on the Democracy dummy is negative and significant in both sets of regressions. We conclude that,