2002 Japan Conference: A Summary of the Papers

It is often argued that financial distress was mainly responsible for the long stagnation of the Japanese economy in the 1990s. This paper is an empirical attempt at examining this argument. Specifically, I estimate the extent to which employment in Japanese firms is affected by high leverage in the corporate sector and bad-loan problems in the banking sector.

There are two elements to this study. First, I analyze the relationship between debt accumulation and employment using firm-level panel data. I construct a panel dataset from the Annual Report of Financial Statements of Incorporated Business, or Hojin Kigyo Tokei Nenpo, of the Ministry of Finance. The virtue of this dataset is its extensive coverage of corporations of various sizes, including unlisted small firms, for all industries except finance and insurance. The sample period is 1993 to 1998 and includes the time of financial turmoil in Japan.

Second, I deal not only with financial leverage in the corporate sector but also with bad loan problems in the banking sector. Lingering bad loans on a bank's balance sheet might lead to fewer bank loans, which might in turn affect employment in bank-dependent firms. In general, it is quite difficult to estimate the effect of loan supply on employment simply from observed data of bank loans because of problems in the identification problem of supply versus demand conditions. Therefore, it is necessary to select a variable that purely represents the supply of loans. I rely on the Short-term Economic Survey of Corporations, called Tankan, by the Bank of Japan. This survey reports a diffusion index of "banks' willingness to lend" which can be a good proxy for the supply of loans. Using this data, I examine the impact of the supply conditions of loans on employment.

The impact of financial distress can be examined empirically by estimating a dynamic labor-demand equation for a firm. The firm maximizes the present value of its earnings, net of the quadratic adjustment cost of hiring/firing labor. It turns out that the demand for labor is a function of the real wage rate, output, financial distress, and lagged employment. Incorporating the degree of financial distress into the model requires two variables: one corresponds to the leverage of the firms and is represented by the ratio of debt to total assets. The other is a proxy for the bad loan burden on banks, which is represented by the lending attitude of commercial banks. One would expect that the harder the bad loan problem hits the bank, the more severe the bank's lending attitude becomes.

The degree of financial distress affects employment through two channels: first, by changing the external finance premium facing the firm; high debt outstanding relative to total assets and/or a severe lending attitude on the part of banks raises the external finance premium under asymmetric information between lenders and borrowers; this in turn leads to an increase in the effective interest rate or a decrease in the discount factor. Second is the disciplinary role of debt. Firm managers have an incentive to cut employment when the debt-asset ratio is high. Faced with increasing debt, managers will make every effort to cut back labor and raise efficiency, because in the case of bankruptcy, those managers may be fired. In other words, managers realize that the adjustment cost of labor is less costly, thus prompting the managers to adjust employment quickly. The financial distress variables affect not only the employment level but also the adjustment process of employment.

Using the panel dataset constructed from the Annual Report of Financial Statements of Incorporated Business, I estimate a dynamic labor demand equation using the two-step GMM estimation proposed by Arellano and Bond. My estimates are for four cases classified by industry (manufacturing and non-manufacturing) and firm size (small to medium-sized firm and large firm groups). The large firms are defined as those whose equity capital in 1993 is larger than 1 billion yen.

I find that financial distress has a negative effect on the firm's employment. The effects are notably larger for small firms. As for corporate leverage and employment, the debt-asset ratio has a significantly negative effect on employment for small firms in both manufacturing and non-manufacturing industries. However, the manner in which the debt-asset ratio affects employment differs between these two industries. For manufacturing industries, a higher debt-asset ratio increases the speed of adjusting employment toward equilibrium. On the other hand, the debt-asset ratio directly and negatively affects the current level of employment for non-manufacturing industries.

Lending attitudes also have a statistically positive effect on employment for all firm groups, irrespective of industry. A severe lending attitude raises the adjustment speed of employment for all firm groups in manufacturing industries and for large firm groups in non-manufacturing industries, while similar attitudes lead to a direct reduction of current employment for small firm groups in non-manufacturing industries. Further, the impact of financial distress on the overall adjustment process of employment is not quantitatively large, but its effect on small firms is quite large. This finding may be interpreted as follows. The external finance premium might be raised by the debt-asset ratio for small firms because they do not have large collateralizable net worth helping them to diversify unobservable idiosyncratic risk as large firms do. Moreover, a number of large firms in Japan belong to industry groups known as keiretsu, where a main bank plays a central role in mitigating the informational asymmetry between lenders and borrowers. Small firms have only relatively loose ties with main banks. As is emphasized by Masahiko Aoki, the main bank system is institutionally complementary with the Japanese employment system in which employees embodying firm-specific training are kept within a firm over quite a long term. Moreover, large firms tend to retain a higher proportion of employees with firm-specific training. Therefore, the labor resources of large firms, embodying firm-specific training, fluctuate less with the temporary adverse shocks since the main banks financially support even troubled firms.

In addition, the disciplinary role of debt may be more potent for small firm groups: the managers of small firms on the verge of bankruptcy feel threatened by the reduction in bank loans and/or are fired easily by their parent firms. Therefore, these managers have good reasons for making every effort to reduce employment and to improve production efficiency. Finally, small firms are more bank-dependent, so the lending attitude of banks has a much stronger effect on the employment of small firms.

My findings that financial distress has an adverse effect on employment, notably for small firms, have important policy implications. Since labor resources in a firm embody new technology that raises production efficiency, reducing corporate debt and wiping out banks' bad loans is urgent if the Japanese economy is to attain sustained long-run growth.