The labor productivity competitiveness of U.S. versus foreign commercial banks.

ABSTRACT

The productivity of labor, measured as the ratio of employees to total assets, has since the 1970s been a common measure of bank performance. Using what this article argues is a better indicator of the competitiveness of commercial banks--the number of employees required to generate one million U.S. dollars in annual revenues--this study concludes that U.S. banks are being outperformed by their foreign counterparts.

INTRODUCTION

The increasing globalization and integration of financial markets and the rapid pace of technological change in the financial services industry are compelling the world's financial intermediaries in general, and commercial banks in particular, to implement innovative ways to cut costs and improve performance in order to meet the challenge of keener domestic and international competition.

Mergers, consolidation of operations, new product and market development and moves toward paperless retail banking are some of the means through which banks around the world are pursing gains in their competitive position. In the midst of these developments, interest on issues of bank performance, efficiency, and competitiveness remains high among academicians as well as practitioners.

This study focuses on the specific issue of labor productivity as a measure of performance and as an indicator of the competitiveness of United States banks versus their foreign counterparts. Section two of this paper reviews the literature on the labor productivity approach to evaluate bank performance. Section three details the methodology to measure and compare the competitiveness of American and foreign commercial banks, based on the labor efficiency yardstick. Discussion of the results in part four is followed by the conclusions at the end of the paper.

REVIEW OF THE LITERATURE

A common approach to measure and rank the performance of the largest commercial banks in the world has been to analyze their levels of employment and patterns of labor utilization. Generally, under this approach, "production" of financial services is quantified in terms of a given bank's total assets and the institution's performance is evaluated via the productivity of its labor as measured by the number of employees needed to generate a given level of production of financial services (total assets).

The method of using labor productivity as the yardstick to evaluate bank performance was pioneered by G. G. Kaufman using 1967 data for large banks in industrialized and developing countries (Kaufman, 1970). He concluded that the efficiency of labor was highest in banks based in the United States and lowest in Japanese banks. Similar conclusions were reached by B.K Short also using data for 1967 (Short, 1971).

In 1984, Sang-Rim Choi and Adrian Tschoegl used 1979 data to reach the significantly different conclusion that the labor efficiency of Japanese banks was now higher than that of banks in the United States. Results of their study showed that American banks required more workers to produce the same level of financial services than their counterpart institutions in Japan and in other industrialized nations (Choi and Tschoegl, 1984). More recently, with figures for the 1980-1986 period, WC. Hunter and S.G. Timme, updating the previous studies by Kaufman, Short and Choi and Tschegl and continuing to use total assets as the measure of production of financial services, reaffirmed Choi and Tschegl's findings and concluded that large U.S. banks were being outperformed by banks in other industrialized countries, based on labor productivity rates (Hunter and Timme, 1990).

Bank mergers in the 1990s, including those of BankAmerica and Security Pacific, Chemical Bank and Manufacturers Hanover Trust, First Chicago and N.B.Detroit and lastly Chase Manhattan and Chemical Bank, have contributed to the increasing consolidation in the U.S. banking industry that has resulted in a 36 percent reduction in the number of independent banking organizations--from 12,380 in 1980 to 7,926 in 1994 (Ritter, Silber and Udell, 1997).

Arguably, this continuing trend towards further consolidation reflects bankers' belief that economies of scale do exist in their industry. Such an argument states that production costs fall as output increases, implying that banks would become more efficient as they grow larger. However, in a recent study, A.N. Berger and D. Humprey have concluded that such cost savings associated with economies of scale have not materialized for banks whose assets exceed $100 million, although they do seem to exist for banks smaller than $100 million in total assets (Berger and Humphrey, 1993).

METHODOLOGY

Measuring the Productivity of Labor in Commercial Banking

In the studies reviewed in the prior section, total assets was the variable used as a measure of the level of production of financial services offered by commercial banks. The number of employees per given level of total assets was used as a measure of labor productivity. This paper contends that since all businesses, including commercial banks, are in business not primarily to increase their assets, but to generate profits, focusing on a Balance Sheet item (total assets) may not be the most appropriate way of assessing the performance of businesses in the banking industry. Rather, since total revenue is more directly associated with sales than total revenue, an Income Statement item, is more appropriate as a measure of production of financial services during a given year. Similarly, the number of employees per given level of total revenue is a better variable to measure labor productivity during the period in question.

In this study, labor productivity is specified as the amount of labor (number of employees) needed to generate one million U.S. dollars in annual revenue for a given bank. The source of data is the Fortune Magazine 500 list for the top forty-four foreign banks ranked on the basis of annual revenues for the 1997/98 period, for the top forty-four non-Japanese foreign banks and for the top forty-four U.S. banks, ranked based on the same criterion during the same period. Complete ranking of the sixty eight international banks along with data on their respective annual revenues, number of employees, total assets and labor productivity, (measured in terms of the number of employees required to generate 1 US$ in annual revenue) is available from the author upon request.

The labor productivity mean was calculated for each of the three samples (foreign banks, non-Japanese foreign banks, and U.S. banks). Establishing the extent to which statistically significant differences in labor productivities occur, amongst the three groups of banks, was accomplished via the statistical testing of differences of sample means.

ANALYSIS OF RESULTS

The Labor Efficiency Rates of the World's Top Revenue Producing Banks

By far, Japanese banks head the list of the world's top revenue producing commercial banks ranked in terms of the productivity of their employees. Eight of the foreign banking institutions with the lowest labor input requirements per million dollars in annual revenues are Japanese. Table 1 below presents the top ten foreign banks ranked based on the labor productivity criterion.

In the category of non-Japanese foreign commercial banks, German institutions outperform their counterparts in this classification. Five of the top ten firms in this ranking are based in Germany. Table 2 shows the ten foreign non-Japanese commercial banks with the highest labor productivity rates.

Only one U.S. bank (J.P. Morgan & Co., Inc.) exhibits labor productivity figures that would rank it among top in the world. Table 3 shows the ten commercial banks in the United States with the highest labor efficiency rates (lowest employees per million dollars in revenues).

Data on labor productivity rates for the forty-four commercial banks included in each sample for the three different categories are summarized below:

HYPOTHESIS TESTING ON THE RELATIVE LABOR PRODUCTIVITIES OF U.S. BANKS VS. FOREIGN BANKS

Based on the results discussed above, the following hypothesis can be formulated about the relative labor efficiency rates of U.S. banks and foreign banks:

Hypothesis I: Labor productivity in U.S. banks is lower than that in foreign banks. That is, the mean number of employees required to generate one million dollars in revenues is higher (at a given statistical level of significance) for U.S. banks than for foreign banks.

Hypothesis II: Labor productivity in U.S. banks is lower than that in non-Japanese foreign banks. Or expressed differently, the mean number of employees needed to produce one million dollars in revenues is higher for U.S. banks than for their non-Japanese foreign counterparts at a given statistical level of significance.

Testing each of the hypothesis above, involves a statistical test on the difference between the means of two populations. For the purpose of these tests, it is assumed that the populations are normally distributed and given the large sample size (44 in each case), it is also assumed that the unequal sample variances approximate their respective population variances.

The corresponding t values are then calculated for each of the two hypothesis and then compared to their respective critical t values in order to validate or reject the given hypothesis (Daniel and Terrell, 1992). These calculations are summarized below:

Since for both hypotheses the computed t values are greater than their respective critical t values at the .01 level of significance, both hypotheses are validated.

CONCLUSION