Until 1974, firms could capitalize or expense all or part of their research and development (R&D) costs. Managerial choice between these two alternatives is hypothesized to be affected by the existence of debt covenants which employ accounting numbers relating to leverage, interest coverage, and ability to pay dividends. In addition, the use of public versus private debt is hypothesized to affect the accounting choice due to differential renegotiation costs. Lastly, a political cost hypothesis is tested. This study uses a multivariate statistical technique, the generalized jackknife. The results suggest that firms which capitalized R&D costs were more highly levered, used more public debt, were closer to dividend restrictions, and were smaller than firms which expensed R&D costs.

Firm size has been used as a proxy for the firm's political costs and hence managers' proclivity to choose income reducing accounting procedures. This study provides additional evidence on this topic by examining the association between firm size and effective corporate tax rates. The latter are one component of a firm's political costs. The roughly fifty largest U.S. exchange-listed firms, in particular oil and gas companies and manufacturing firms, have significantly higher worldwide tax rates than other firms. These higher tax rates are observed primarily after the implementation of the U.S. 1969 Tax Reform Act and after the OPEC countries raised their tax rates on U.S. oil producers. The findings, which are insensitive to alternative sources of data, alternative measures of firm size, and alternative measures of effective tax rates, are consistent with the use in previous studies of firm size as a proxy for the firm's political costs.