© Evanston Photographic Mitchell Petersen, the Glen Vasel Associate Professor of Finance

The information revolution in small business lending

By Mitchell Petersen, the Glen Vasel Associate Professor of Finance

Whether lending to a firm or an individual, bankers must evaluate the likelihood of repayment. Historically, the method for evaluating the two loan applications has been different. When an individual applies for a mortgage or a credit card, a computer, not a loan officer, often makes the decision based on analysis of the applicant's credit report. Automated data collection and lending decisions make these markets impersonal, national in scope and more competitive.

The lending markets for small firms are quite different. Firms do not have credit scores. Instead, a loan officer interviews the entrepreneur, and possibly customers and suppliers, to assess the proposed investment's merit. The lending decision is based on the officer's judgment; information that is difficult to computerize or automate. Historically, the small business lending market has been a local market, based on personal relationships.

The consumer lending market of the 1950s was described in the same terms. Lenders knew their borrowers personally, not through computerized data. Over the past three decades, the consumer lending market has been transformed, but has the small business lending market?

To document changes in small business lending, we examined the Federal Reserve's Survey of Small Business Finances. The survey describes firms' relationships with their banks, including loans, deposit accounts and their purchase of financial services. It also records the distance between a firm and their bank branch as well as how they communicate with their bank.

When examining the evolution of firms' banking relationships, two facts are striking. First, the distance between firms and their lenders has risen by 9 percent per year from the early 1970s to the 1990s; a 460 percent increase (Fig. 1). Figure 1 Correspondingly, firms' communication with their banks has become progressively more impersonal. The fraction of firms communicating with their banks in person falls from 68 to 34 percent.

What caused this dramatic change? Given U.S. banking history the answer may be obvious: deregulation and consolidation. There were more than 14,000 banks in the U.S. at the beginning of our sample. The number dropped to less than 11,000 by 1992 (Fig. 2). This reduction could explain the growing distance between firms and their lenders. This explanation, however, is too simple.

First, banking consolidation begins in the mid 1980s, yet the growth of distance and impersonal communication starts a decade earlier. Second, we measure distance between firms and their lender's branch; although the number of banks has been declining since the mid 1980s, the number of branches has consistently grown. Thus, other forces must be at work.

Technology transforms the way we store, transmit and process information, and the information-intensive banking industry has exploited this technology. Using credit scoring models to make lending decisions is an example of substituting capital for labor. Previously, loan officers would review an application — a labor-intensive and qualitative process. Lending decisions based on credit reports and analytic decision rules, however, require less of the loan officer's time. Personal intervention has not been eliminated, just focused on the most marginal decisions. Loan originations involve fewer people and more computers.

To verify that greater use of information technology has transformed the lending market, we examined how bank employment has changed. The ratio of bank employees to loans, a measure of the labor intensity, drops by almost 50 percent over our sample as banks automate the loan approval process. The reduction in employees explains the increase in the distance we documented. In regions where banks replaced employees with information technology early, distance increases at the same time.

As technology has sped up human ability to process information, it has altered information-intensive industries such as lending. The growing national scope of lending has the capacity to lower the cost and expand the availability of capital for small firms. Whether it does is an open question.

This essay is based on Petersen and Raghuram Rajan's "Does Distance Still Matter: The Information Revolution in Small Business Lending," Journal of Finance, Dec. 2002.