Our primary measure of leverage is the long-term debt ratio (LT Debt/TA), defined as the amount of long-term debt scaled by
the book value of total assets.6 Demirgüç-Kunt and Maksimovic (1999) and Booth et al. (2001) also use the long-term debt ratio in
their research. For robustness, we first consider the total debt ratio (TD/TA), defined as the sum of short-term debt and long-term
debt scaled by the book value of total assets. We choose to use the long-term debt ratio rather than the total debt ratio as our main
leverage measure in part because of the well-documented observation that total leverage is driven largely by the amount of
long-term debt (see, e.g., Barclay and Smith, 1995; Johnson, 2003). Moreover, firms use short-term debt primarily to finance
current assets or as part of working capital management, and thus leverage measures that include short-term debt may be less
sensitive to factors that affect firms' corporate leverage policy. We also test the robustness of our results by considering the
market debt ratio (LT Debt/MA), which scales long-term debt by the market value of total assets. We use the book value-based
variable as our main measure of leverage because firms have been shown to be more concerned about book leverage ratios than
market leverage ratios,7 and hence the former should be more useful in capturing the relative importance of demand-side versus
supply-side forces in the debtor–creditor relation.