The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
FEDERAL RESERVE BAN K O F S T . L O U IS J«iy W» LITTLE Vol. 5 6 , No. 7 A Primer on the Consumer Price Index 2 Income and Expenses of Eighth District Mem ber Banks — 1973 .................... 8 Branching, Holding Companies, and Banking Concentration in the Eighth District.................................... 11 A Primer on the Consumer Price Index D E N IS S. KARNOSKY H a R D LY a day goes by without mention o f the effects o f inflation on the econom ic well-being o f the average citizen, as a worker and as a consumer. Most references to inflation, in turn, are in terms o f the con sumer price index (C P I) and the way many prices, prominently led by food and petroleum products, have shot up in the last year and a half. The C PI is often cited b y the media as a measure o f changes in the cost o f living. It is incorporated as an escalator in labor contracts covering over 5 million workers and is now used to adjust Social Security benefits for almost 29 million people. There is increasing talk o f indexing all contracts to some measure o f general prices, and the consumer price index presumably w ould play a role. Given its w ide use, and even misuse, it is important to understand the construction o f the index and some o f its major shortcomings. The stated intent o f the builders o f the index at the Bureau o f Labor Statistics ( B L S ) is quite limited. The CPI is designed to meas ure changes in the average price o f a representative sample o f goods and services purchased b y typical w age earners and clerical workers in urban areas in the United States. It is interpreted m uch more broadly, however. The Mechanics of the CPI In the jargon o f economists, the consumer price in dex is a modified Laspeyres index. W hat this simply means is that the CPI measures changes in the total dollar cost o f a specific com bination o f goods and services.1 For example, if a person kept track, month to month, o f the total dollar cost o f buying a dozen Grade-A large white eggs, a one-pound loaf o f white bread, and a 16-ounce box o f cornflakes, the numbers 'A Laspeyres index is a fixed-weight index, where the weights are the relative quantities as of some base period. The formula for such an index is: It — Sptq0/2 p 0q0 where It is the value of the index in the current period, pt are the various component prices in the current period, p0 are the component prices in the base period, and q0 are the component quantities in the base period. The formula actually used by the BLS is somewhat diiferent, but algebraically equivalent. The index is constructed by a chain computational procedure, which can be written in simplified form as: Page 2 CPIt = 2 p 0q0 /2 p 0q0. w ould provide the basis for constructing a little index o f food prices, m uch like the CPI. All that is left to do is divide the total cost in each month b y the cost in the first month, to get a measure o f the relative change in the cost. This is essentially the method used to construct an index like the CPI. The quantities of the goods and services are held constant, and the in dex measures the effect on total dollar cost o f changes in prices o f the components. The first problem in constructing any price index is to determine the items to be priced and just how m uch o f each to include in the bundle o f goods. In the example above, the index might be interpreted as measuring the month-to-month changes in the total cost o f breakfast foods. Should w e also include sugar, fruit juice, coffee, or milk? Should w e include bacon, or should it be sausage, and, if so, in what quantities? An index derived from a bundle containing three dozen eggs w ould be different from one containing only one dozen. Yet another series w ould result if oatmeal were substituted for cornflakes. One guide to the appropriate relative proportions, or weights, w ould be the actual amounts w hich comprise a “typical” breakfast. The index w ould then measure changes in the average price o f a particular breakfast, instead o f breakfast foods in general. The problem is to deter mine what makes up a “typical” breakfast. One person probably w ould have little trouble with this problem, but constructing an index appropriate to the spending patterns o f over 200 million people is very difficult. Consumers spend on a w ide variety of items. Some items, like food, are bought regularly and immediately consumed, while others, like houses and automobiles, are bought irregularly and yield services over a long period o f time. Some scheme for deter mining the relative quantities o f each o f these goods and services is required, and, if the index is to be use ful as an aggregate measure o f the prices o f consumer goods, the relative amounts should be representative o f those actually purchased in the econom y.2 The m ethod used in deriving weights for the C PI is based on periodic surveys o f consumer spending pat2Unfortunately no perfect means of determining the weights has yet been developed. For a presentation of some of the many schemes which have been suggested, see Irving Fisher, Making of Index Numbers (N ew York: Houghton Mifflin Company, 1922). F E D E R A L R E S E R V E B A N K O F ST. L O U I S terns. These surveys were undertaken in 1917-19, 1934-36, 1950-51, 1960-61, and 1972-73.3 The results of the 1972-73 survey, along with other major changes, are scheduled to be incorporated into the index in 1977. The survey seeks to determine the proportion of consumer spending that is devoted to various kinds of goods and services. These proportions are then used to determine the relative importance o f the various prices in the index. On the basis o f the survey con ducted in 1960-61, estimates were made that, on aver age, typical wage earners and clerical workers in ur ban areas devoted 22.4 percent of their spending to food, 33.2 percent for housing, 10.6 percent for ap parel and upkeep, 13.9 percent for transportation, 5.7 percent for m edical care, 2.8 percent for personal care, and 5.1 percent for other goods and services.4 These are the weights that these various prices receive in the com putation of the current consumer price in dex. The weights were introduced in January 1964 and have been held constant since. Prices of over 400 separate items are currently used to construct the CPI. The list o f items whose prices are sampled ranges from diapers through funeral serv ices and includes such things as cornflakes, roof shingles, cough syrup, basketballs, and two-year-old Chevrolets and Fords. The prices are sales prices and thus include excise and sales taxes. In addition to the prices o f commodities and services, the sample in cludes such items as real estate taxes on ow ned homes, utility rates, and mortgage costs. Incom e taxes are not included and neither are Social Security taxes. Trained representatives collect price quotations and the BLS uses strict statistical procedures for processing the data into the CPI. :1These surveys are conducted in numerous metropolitan areas. The 1960-61 survey was conducted in 66 Standard Metropolitan Statistical Areas and smaller cities. The sample included 4,343 urban families of two or more persons and 517 single workers. These single persons are not neces sarily unmarried, but are classified as being financially in dependent. Of the areas included in the survey 56 are cur rently sampled for price movements. Population weights for these 56 areas are used to combine the data into a city average for the United States. This city average is reported as the CPI. Price indexes for some of the individual cities are also published. For more details on the survey procedure, see Marvin Wilkerson, “ The Revised City Sample for the Consumer Price Index,” Monthly Labor Review (O ctober 1960), pp. 1078-83. Also see U.S. Department of Labor, Bureau of Labor Statistics, BLS Handbook of Methods, Bul letin 1711 (1971), pp. 59-67. 4These weights were introduced in January 1964 and were adjusted for changes in prices between the date of the survey and December 1963. The weights represent an estimate of how the typical urban wage earner would allocate a spending budget in December 1963 if the same items were bought as reported in the 1960-61 survey, but at the prices prevailing in December 1963. JULY 1974 Some Problems and Shortcomings Construction o f a price index as comprehensive as the CPI is a very com plex, difficult, and expensive task. On the one hand are the statistical problems re lated to sampling and processing o f data. Quotations on the prices involved in all transactions are almost impossible to record. Instead, samples are designed to yield results which h „ve a high probability of repre senting price behavior. On the other hand are the conceptual difficulties, the most prominent being the handling o f changes in the quality o f commodities and services, and changes in people’s tastes and prefer ences. The BLS is able to collect price quotations on automobiles, for example, but they are unable to price the services rendered by a car. It is the services o f an automobile that are valued by consumers, however, not just the auto itself.5 T o take another example, how m uch more service, in dollars and cents, does a color television set yield com pared to a black and white set? Even without this difference, there is the problem o f changes in quality stemming from the programming policies o f television networks and station owners. A decrease in the overall pleasure derived from a tele vision set, either as entertainment or as a source of information, increases the cost of its services just like an increase in the dollar price of the set. It is im pos sible for anyone, other than an individual viewer, to measure objectively changes in the quality of a given com m odity. A similar problem arises when new com modities are introduced. A related problem is that the CPI is constructed as a fixed-weight index. Essentially, the CPI attempts to measure the percentage change in the amount that consumers w ould have to spend to purchase goods and services in the same quantities and o f the same quality that they purchased when the survey was taken. Currently, the CPI measures changes in the dollar cost o f items that the average urban consumer bought in 1960-61. It says nothing at all about the 5Consider a hypothetical case based on the mandatory safety devices now built into cars. To the extent that they are effec tive in reducing the probability of bodily injury, the services of automobiles are apparently increased. It is not clear, how ever, that the increase in the price of a car that these safety devices represent should be discounted as reflecting an in crease in quality. Other things equal, effective safety devices will result in less injury in automobile accidents and, pre sumably, lower insurance premiums. The price of automobiles goes up and the price of insurance goes down. The result of treating the safety devices as increasing the quality of auto mobiles is a decrease in the index of the price of consumer goods and services. One would conclude, incorrectly, that the mandatory safety program had decreased consumer prices. In this example, all that actually happened was that the program tended to transfer resources from one industry to an other, leaving average consumer prices unchanged. Page 3 F E D E R A L R E S E R V E B A N K O F ST. LO U I S JULY 1974 relative quantities or quality o f the bundle o f goods that consumers actually buy today. It is in this context that the CPI is not an accurate gauge o f changes in the cost of living.6 himself to the limit o f his ability to absorb the services being rendered. Resources are limited, however, and the most binding constraint on an individual is his ability to command goods and services — that is, his purchasing power. CPI and the Cost of Living W ithin the context of a given level of incom e and ability to borrow, a person must decide where his dollars will probably yield the most satisfaction. The factors which determine this choice are each individ ual’s subjective valuation of various items, his income, and the price of each item relative to prices o f other goods and services, as well as some expectations about future incom e and prices. Changes in individual tastes, income, relative prices, and expectations w ould alter the way that incom e is allocated among various com m odities.8 The CPI attempts to measure the cost o f consumer goods to the “ average” urban w age earner. Being an average, the price index is only a rough approximation o f the prices paid by any one individual or family. Rising food prices, for example, get a weight o f about 22 percent in the index, but this understates the effect o f increases in food prices on the cost o f consumer goods to low incom e groups w ho devote more than 22 percent o f their spending to food. At the same time it overstates the effect on someone whose spending on fo o d accounts for only 10 percent o f their total spending. An additional problem is that consumers do not spend their incom e in the same manner year after year. They do not buy the same kinds o f things, or even if they do, they do not buy them in the same relative amounts. However, the CPI, as a fixed-weight index, is based on the presumption that consumer spending patterns change little over time. Thus the actual average price o f consumer goods is not cap tured in the index. W hat are the factors which determine the manner b y which people allocate their income among various goods and services? The foundation of econom ic analy sis is that people attempt to maximize their ow n w ell being. That is, they behave so as to derive the most satisfaction from their limited resources. People buy things which they believe ( not always correctly, since we do not have perfect information about the char acteristics of all goods and services) will yield them the greatest satisfaction per dollar.7 The decision is a very personal one, based on each individual’s subjec tive valuation of things he or she likes best among the available alternatives. If resources were unlimited there w ould be no problem , as everyone could indulge eThis shortcoming is recognized and emphasized by the Bu reau of Labor Statistics, which continuously reminds readers in its publications that the CPI cannot be used as an estimate of current spending patterns or as an indicator of changes in consumer spending. Despite this persistent warning, however, the CPI continues to be so applied. 7Some people interpret this lack of information about product characteristics as justification for governmental intervention to prohibit “ shoddy products” in the market. While no one wants to be disappointed in a product he buys, this argument fails to distinguish between purchases made in ignorance of a product’s true quality and those made precisely because of “ inferior” quality, and often associated lower price. Page 4 W e can get some feel for the way consumers change their spending patterns by com paring the proportion o f spending devoted to the various classes o f goods and services as reported in the 1960-61 survey of con sumer spending to those of the 1950-51 survey. Table I shows the com position of spending reported in each survey since the mid-1930s.° There were substantial shifts in spending patterns, highlighted by a sharp reduction in the proportion of total purchases devoted to food, and large increases in the proportion going for transportation services, m edical care, and reading and recreation. This does not mean that over the decade of the 1950s the average urban wage-eam er decreased spending on foods and increased spending on the other items. Consumer spending for all goods and services increased 70 percent between 1950 and 1960. Spending for some items, like transportation, rose faster than spending on other items, like food. As a result the proportion spent on transportation rose and the proportion spent on food decreased. For the entire interval from D ecem ber 1952 through D ecem ber 1963, the CPI was com puted on the basis 8The problem of comparing the satisfaction derived from con sumption of a commodity today to the satisfaction derived yesterday is not trivial. To a style-conscious person it makes a great deal of difference whether last year’s clothes are worn last year or this year. In the case of the CPI, this type of effect would be manifested, for example, in changes in the age-composition of the population. Presumably tastes change with age. For example, in 1973, the proportion of the popula tion under the age of 25 was estimated at 44.9 percent, up sharply from the 35.6 percent estimated in 1960 when the survey of consumer spending was taken. See Franklin Fisher and Karl Shell, Economic Theory of Price Indices (N ew York: Academic Press, 1972). "The metropolitan areas sampled changed from survey to sur vey. In addition, some commodities were added and others were dropped from one survey to the next. Differences in the reported proportion reflect, in part, these changes, and not changes in consumer spending patterns. F E D E R A L R E S E R V E B A N K O F ST. LOUI S JULY T a b le I 1974 C o n s u m e r P ric e I n d e x Percentage Distribution of Consumer Expenditures 1 9 3 5 -3 9 (b ased on 1 9 3 4 -3 6 su rv e y ) D ecem ber 1 9 5 2 (b a s e d on 1 9 5 0 -5 1 su rv e y ) Decem ber 1 9 6 3 (b a s e d on 1 9 6 0 -6 1 su rv e y) Food 3 5 .4 % 2 9 .6 % 2 2 .4 % H o u s in g 3 3 .7 3 2 .5 3 3 .2 A p p a re l & U p k e e p 1 1 .0 9 .2 1 0 .6 1 1 .3 1 3 .9 T ra n sp o rta tio n 8.2 M e d ic a l C a re 4 .0 5.1 5 .7 P e rso n a l C a re 2 .4 2 .0 2.8 R e a d in g & Recreation 2 .9 5 .3 5 .9 O th e r G o o d s & S e rv ic e s* 2 .4 5 .0 5.1 ♦Includes tobacco products, alcoholic beverages, and personal expenses such as funeral, bank, and legal services. o f weights determined in the 1950-51 survey.10 In terms o f the way people purportedly allocated their expenditures on commodities in the 1960-61 survey, changes in food prices had a smaller effect on their spending budget in late 1963 than reported in the CPI, and all other nonfood components were more im por tant than reported in the CPI. Changes in the price of food had an exaggerated effect on the CPI, but it is impossible to determine just when in the 1950-60 period consumer spending pat terns between food and other commodities changed. The pattern of spending could have changed slowly over the period. However, the change might have com e very soon after the survey was taken, for the 1950-51 period was marked by “scare-buying,” as con sumers sought to stockpile various commodities in an ticipation of price controls and rationing. The Korean W ar had just started and the memories o f the W orld W ar II experience were fresh. Alternatively, the change in spending patterns might not have com e un til 1960-61 when the new survey was taken. Whereas consumer spending was rising rapidly in 1950-51, the econom y was in a recession during the 1960-61 period, with unemployment rising to 7 percent o f the labor force. This w ould be expected to have an effect on the way consumers spend. The effect o f this weighting problem on the index can be seen by comparing the estimated consumption 10Prior to January 1953, the CPI was based essentially on weights determined in the 1934-36 period. Some interim adjustments were made during World War II and in the early Korean War period. The 1950-51 survey served as the basis of the CPI from January 1953 to December 1963. The 1960-61 survey has been used since, and is not scheduled to be replaced by the results of the 1972-73 survey until 1977. 1966 1967 1968 1969 1970 1971 Percentages are annual rates of change for periods indicated. Latest d ata plotted. M a y 1972 1973 1974 patterns in 1963 from the 1960-61 survey, with those im plied by the CPI based on the 1950-51 survey.11 If, in D ecem ber of 1963, consumers w ould have bought goods and services in the same quantities and of the same quality as they had in 1950-51, the bundle would have cost 15.6 percent more than it did in D ecem ber 1952. At D ecem ber 1963 prices, food w ould have ac counted for 28.2 percent o f total spending, 30.7 per cent w ould have gone for housing, 10.6 percent for apparel and upkeep, 11.6 percent for transportation, and 18.1 percent for health and recreation. Comparing these implied numbers to those o f the 1960-61 survey reported in Table I we can see that the CPI overstated the influence of food prices on household budgets and understated the importance of all other types of con sumer goods. The problem stems from the fact that the CPI, as a fixed-weight index, cannot account for changes in relative prices. A fixed-weight index presumes that the com position of their spending remains unchanged as relative prices change. W hen some prices rise faster than others, however, people substitute consumption of some items for others. There is no way, other than frequent surveying, to determine the extent to which n The CPI is not a measure of the price level, but instead is a measure of changes in the level of prices from some arbi trarily selected reference point. This presents a special prob lem when the Bureau of Labor Statistics introduces the results of new surveys and changes the weights. They must decide a reference point from which to compute changes in prices using the new weights. The procedure they use is to link the new series to the level of the CPI of the month prior to the weight revision. There is no reason to assume that this is the appropriate price level. In fact, comparison of the 1960-61 survey data and the relative importance in Decem ber 1963 shows clearly that it is not. Thus while a fixedweight price index loses economic meaning when relative prices change, periodic weight revision to account for the changes in relative prices destroys the validity of the CPI as a statistical time series. Page 5 F E D E R A L R E S E R V E B A N K O F ST. L O U I S people are willing or able to switch tlieir consumption patterns when some prices change relative to otners. From 1963, when the current weights were intro duced into the consumer price index, until 19/3, the consumer price index increased about 45 percent. Over that same period per capita after-tax personal incom e in the country increased by about 96 per cent.1- The difference in these two magnitudes repre sents the gain in “real incom e” per person over the decade, as suggested by the CPI. Such an increase in real incom e w ould be expected to generate substantial shifts in die spending patterns of the average Ameri can. For example, as income increases rapidly, the demand for “necessity” items such as food w ould not be expected to increase as fast as the demand for some other items. “Luxury” goods, such as recreational ve hicles, becom e increasingly attractive to families, either because o f higher incomes or because of a shift in preference toward more active leisure. A fixedweight index does not account for these shifts. CPI and the Value of the Dollar In the words o f the BLS, “The [consumer price] index represents price change for everything people buy for living . . . ” 13 If the statement of the BLS is interpreted literally, the CPI is intended to measure changes in the value of money. After all, prices are just exchange rates between money and other assets, including goods and services; if the CPI captures the average change of all prices, it necessarily would serve as a gauge of change in the purchasing pow er of money. An index of the purchasing pow er of money would have to be all inclusive; that is, it would have to ac count for the prices of all things that can be exchanged for money. The list w ould include, in addition to goods and services, bonds, stocks, and investment goods. The CPI, which incorporates only prices o f current consumer goods, is far short of incorporating a suffi cient number of prices to be used as a measure of the purchasing pow er of money. JULY 1974 current consumption. Buying a house, setting up a college fund for the children, and contributing to a re tirement plan reflect plans to consume in the future. For the most part, the prices o f assets which represent future purchasing pow er are not included in the C PI.14 W hat are the assets that people can buy today in order to consume tomorrow? The most obvious are durable goods, such as home appliances, automobiles, houses, and clothes. These all yield continuing service and can be bought today for consumption in the fu ture. Many of these items are included in the CPI, but many others are not. Excluded from the index are financial assets, such as bonds, savings accounts, pen sion plans, and retirement funds. W hile they yield little direct service through ownership, they can be exchanged in the future for dollars, which in turn can be exchanged for goods and services. In considering the purchasing pow er of money, we must take account o f the amount of future dollars that a dollar will buy today.15 Many of these assets are not included in the CPI and, therefore, it is not a good measure of the purchasing pow er of money. CPI and the Causes of Inflation A fixed-weight index, like the CPI, is particularly susceptible to misinterpretation during short periods when the prices of some o f the com ponent parts change dramatically.16 Analysis o f econom y-w ide de velopments requires a price index which measures changes in the average prices being paid in the econ omy. W hen some relatively autonomous event, like the recent oil embargo or the increase in the Russian demand for our grain, contributes to intense pressure on prices in a few markets, the CPI incorrectly trans14There is no guarantee that a person will be able, in the fu ture, to buy as much as was planned. If prices increase faster than expected, purchasing power will be less than an ticipated. W e know nothing about what prices will actually be in the future. W e are limited, instead, to the effect today of expected future prices. Armen A. Alchian and Benjamin Klein, “ On a Correct Measure of Inflation,” Journal of Money, Credit, and Banking, Part 1 (February 1973), pp. 173-91. One must keep in mind that people do not only make decisions about what to consume today, but they also make plans for consumption tomorrow and years into the future. People can and do trade-off b e tween consuming today and making provisions for consuming tomorrow. Eating a meal at a restaurant is 15It is popular to deflate the money stock by the CPI to get a measure of the amount of “ real money balances” in the economy. On this basis, real money balances have declined over the past year. It is interesting to construct a similar series where the money stock is deflated by the market price of Aaa-rated corporate bonds. The picture is very different. This latter series admittedly is arbitrary, but is it any more so than the series using the CPI? See “ Real Money Bal ances: A Misleading Indicator of Monetary Actions,” this Review (February 1974), pp. 2-10. i-This includes all persons, in addition to urban wage earners and clinical workers. " ’Everyone who deals with data should be aware of the pit falls. For a sobering discussion of the problems, see Oskar Morgenstern, On the Accuracy of Economic Observations (Princeton: Princeton University Press, 1963). 13Bureau of Labor Statistics, Handbook of Methods, p. 76. Page 6 F E D E R A L R E S E R V E B A N K O F ST. L O U I S JULY T a b le II Sources of Recent C h anges in the C P I1 ( M a j o r Exp e n d itu re C la sse s a n d Selected S u b -C o m p o n e n ts) A n n u a l Rates o f C h a n g e S ou rces of C h a n g e s in C P I2 1 2 / 7 3 -5 / 7 4 1 2 / 7 2 -1 2 / 7 3 8 .8 % 1 2 .7 % 1 0 0 .0 % 1 0 0 .0 % 2 2 .1 % 1 4 .7 % 4 3 .9 % 2 2 .7 % C e re a l a n d b a k e ry products 2 8 .8 2 7 .5 M e a ts, p ou ltry & fish 2 6 .4 — 10.1 1 8 .0 D a ir y products 2 2 .5 16.8 6.9 4 .0 Fruits a n d v e ge ta b le s 14.1 62.1 4.8 1 5 .4 O th e r 17.5 17.8 7.1 5 .4 1 2 .7 10.5 7.2 4.3 7.2 1 2 .4 2 7 .7 3 2 .5 1 2 / 7 2 -1 2 / 7 3 *1 (A l! Item s) Food at hom e Food a w a y from hom e H o u sin g 7.1 1 2 / 7 3 -5 / 7 4 3.5 -5 .6 1 1.5 2 3 .9 6.2 9.1 T ra n sp o rtatio n 4.5 2 1 .9 6 .7 2 1 .7 G a s o lin e 1 6 .7 Fuel & utilities 1 9 .7 72.1 6 .0 A p p a r e l a n d upkee p 4 .4 8.5 5 .2 6 .7 M e d ic a l care 5.2 10.1 3.8 5 .0 P e rso n a l care 6 .3 10.9 1.8 2.2 R e a d in g & recreation 2.9 8.5 1.9 3 .6 O th e r 3.8 5.8 2.2 2.2 1 0 .4 1 4 .7 7 3 .9 7 3 .4 1974 at inflation, as contributing to growth of aggregate demand. T o the extent that demand was not curtailed in some other market, pressure was put on the aggre gate price level. The increase in the price o f fo o d reflected the response in the mar ket for food to this increase in aggregate demand. The rise in food prices no more caused the inflation than a crowing rooster causes the sun to rise. If one falls into the trap of consider ing fo o d prices, or oil prices, or auto m obile prices as causes o f inflation, the logic of the position leads to the conclu sion that the way to stop inflation is to decree that henceforth these individual prices shall not rise — if you do not want the sun to com e up, shoot the rooster. Conclusion The major econom ic problem o f the day is inflation. The only proven perma D u ra b le 2.4 1 0 .6 1 3 .0 4 .5 nent cure for this problem is a program 6 9 .4 N o n d u r a b le 16.1 6 0 .4 13.3 designed to keep the growth of aggre Services 2 6 .6 6.2 9 .6 26.1 gate demand in line with productive ca 1Not seasonally adjusted pacity. Some might argue that such an 2The proportion o f the change in the CPI due to changes in the prices o f particular com ponents. For example, increases in the prices o f housing services between December approach is “ all right in theory, but it 1972 and December 1973 accounted for 27.7 percent o f the rise in the CPI over that same period. does not work in practice.” This position, though logically absurd, is understand lates these individual price increases into general in able, given the w ide circulation of the notion that our creases in the average price o f consumer goods. inflation is caused by special factors, such as the oil C om m o d ities W e know that the amount of food items, like beef, that consumers purchased last year decreased as the price o f beef rose. The rise in the CPI reflected the increase in beef prices, but not the decrease in the amount o f beef purchased. As beef prices rose, people switched to other food sources. The same phenomena occurred in the markets for gasoline and other petro leum-based fuel. Total consumption of refined petro leum products in the United States decreased by 7.4 percent from O ctober 1973 to March 1974. This de crease in quantity was not captured in the CPI, which held the quantity constant. The rapid increases in oil prices were added in with fixed weights. Inflation, as a persistent increase in the average level o f prices, is everywhere a problem of excess ag gregate demand, stemming from any o f a number of sources. The huge increase in the demand for grain by the Russians, while manifested directly in the general food market, is better analyzed, for purposes of looking embargo. It is an easy matter to com pute the portion o f the rise in the CPI that was due to increases in food prices, or oil prices. It is also easy, but incorrect, to take one further step and say that the increases in, say, food prices accounted for 44 percent o f the inflation. The prices of the com ponents can cause the price index to rise, but that says nothing about the causes of inflation. If shortcomings o f the CPI are kept in mind, it can serve as a gauge of price pressure in a significant por tion of the econom y. It does not tell us why prices are rising, just that some of them are going up. Our cur rent inflation is little different from those o f the past, except that it has been allowed to continue longer. R e sponsible action to keep aggregate demand in check has been and still is the only answer. Page 7 Income and Expenses of Eighth District Member Banks—1973 W ILLIAM LEPLEY E T IN C O M E of the 431 Federal Reserve mem ber banks in the Eighth District rose by 9.8 percent in 1973, substantially higher than the 3.5 percent in crease that occurred in 1972. This increase in Eighth District net incom e compares with increases for all member banks in the nation of 17.3 percent in 1973 and 7 percent in 1972. The total operating incom e of member banks in the District rose by 28.2 percent during the past year, while total operating expenses increased at an even faster rate of 31.4 percent. The com parable figures for member banks in the nation were 33.1 percent and 36.6 percent, respectively. Many o f the incom e and expense items increased substantially in 1973. These figures should be consid ered in light o f the monetary expansion, the high in flation rate, and the rising interest rates which o c curred in 1973. Operating Income Total operating income is largely determined by investments in various earning assets and the rates of return realized on these assets. Loan incom e makes up the largest portion of bank revenue. Various security holdings, trust services, service charges on deposit accounts, and other miscellaneous items provide the remaining sources of income. Incom e from loans, including that resulting from Federal funds sold and securities purchased under agreements to resell, increased faster than all other categories of operating incom e in 1973 (see Table I ) . Receipts from loans increased by 37.5 percent in 1973, follow ing the 10.6 percent increase o f 1972. One reason for the gain in loan incom e was the larger share of loans in the asset structure. W hile total assets rose by 12.9 percent during the year, the amount of total loans increased by 21 percent. As in Page 8 dicated in the accom panying chart, loans as a percent age of total assets increased from 53.1 percent in D e cem ber 1972 to 56.9 percent in D ecem ber 1973. The rates o f change of the major loan classifications varied substantially. The loan category which showed the most dramatic growth was Federal funds sold and securities purchased under agreements to resell — it increased b y 56.4 percent during 1973 and constituted 8.1 percent of total assets at year end. The largest type of loans, com mercial and industrial, registered an 18.9 percent increase in outstandings and accounted for F E D E R A L R E S E R V E B A N K O F ST. LOUI S JULY 1974 T a b le I IN C O M E A N D EXPENSES OF M EM BER B A N K S IN THE EIGHTH FEDERAL RESERVE DISTRICT __________ T h o u s a n d o f D o lla rs _________ 1973 Total O p e r a tin g Incom e ........................................................... .... 1972 Percent C h a n g e 19 7 1 1 9 7 2 -7 3 9 2 8 ,0 5 0 2 8 .2 % 1 0 .0 ' 1 0 .6 $ 1 ,3 0 8 ,3 9 5 $ 1 ,0 2 0 ,8 9 7 9 0 4 ,1 2 6 6 5 7 ,6 5 0 5 9 4 ,4 9 6 3 7 .5 Incom e from Securities ......................................................... 2 7 2 ,9 1 9 2 4 6 ,2 5 4 2 2 4 ,5 2 8 10.8 U.S. T re asu ry Securities ..................................................... 1 0 9 ,9 5 4 1 0 7 ,7 6 6 1 1 0 ,7 2 0 2 .0 O th e r .................................................................................. Incom e from L oan s ................................................................ $ 1971- 9 .7 - 2 .7 16 2 ,9 6 5 1 3 8 ,4 8 8 1 1 3 ,8 0 8 1 7 .7 2 1 .7 Trust D epartm ent Incom e ....................................................... 2 9 ,0 6 7 2 6 ,5 6 8 2 3 ,6 5 1 9.4 1 2.3 Service C h a rg e s on D ep osit A ccts............................................ 2 9 ,4 8 4 2 7 ,9 4 7 2 7 ,0 5 1 5 .5 3.3 O th e r O p e r a tin g In c o m e ......................................................... 7 2 ,7 9 9 6 2 ,4 7 9 5 8 ,3 2 6 1 6.5 7.1 7 3 5 ,3 6 5 3 1 .4 12 .9 Total O p e r a tin g Ex p e n se s ......................................................... .... S a la rie s, W a g e s , a n d Interest on $ 1 ,0 9 1 ,3 5 8 $ 8 3 0 ,4 4 9 $ Benefits ............................................. 2 4 2 ,1 9 5 2 1 4 ,3 3 2 1 9 7 ,8 4 0 1 3.0 8.3 D ep osits ............................................................. 4 5 7 ,6 8 2 35 1 ,6 7 9 3 0 7 ,8 3 3 30.1 1 4.2 2 4 .4 O th e r Interest Exp e n se s ......................................................... 1 3 5 ,6 8 1 4 6 ,5 9 3 3 7 ,4 6 9 1 9 1 .2 O th e r O p e ra tin g Exp e n se s ..................................................... 2 5 5 ,8 0 0 2 1 7 ,8 4 4 1 9 2 ,2 2 2 1 7.4 Incom e Before Incom e T a xes a n d Securities G a in s (or Losse s) ... 2 1 7 ,0 3 7 1 9 0 ,4 4 8 1 9 2 ,6 8 6 1 4 .0 Less A p p lic a b le Incom e T axes ................................................ 4 8 ,6 5 8 4 3 ,3 6 0 5 1 ,2 7 6 12.2 ......................... 1 6 8 ,3 7 9 1 4 7 ,0 8 9 1 4 1 ,4 1 0 (o r L osse s) A fte r T axes ......................... 311 5 ,3 7 1 5 ,8 7 6 738 605 498 Inco m e B efore Securities G a in s N et Securities G a in s (o r Lo sse s) Extra C h a rg e s or C redits A fter T a xe s ...................................... - 13.3 - 1 4 .5 - 9 4 .2 1.2 15.4 4 .0 - 8 .6 1 2 2 .0 2 1 .5 7 6 .5 2 2 6 .9 Less M in o r it y Interest in C o n so lid a te d S u b s id ia r ie s ............... 20 85 26 N et Incoime ................................................................................ 1 6 7 ,9 3 2 1 5 2 ,9 7 9 1 4 7 ,7 5 8 9 .8 C a sh D iv id e n d s Paid .................................................................. 6 0 ,2 7 7 5 6 ,7 6 2 6 1 ,2 6 6 6 .2 - 7.4 N u m b e r of B a n k s ........................................................................ 431 430 432 0 .2 - 0 .5 - 3 .5 NOTE: The boundaries o f the Tenth Federal Reserve District were expanded on January 24, 1972 to include several counties in western Missouri which had been in the Eighth Federal Reserve District. The income and expense data for 1971 have been adjusted to con form to the January 24, 1972 revision in district boundaries. 16.2 percent of total assets. Loans to individuals for personal expenditures were up 12.4 percent from the previous year, while all real estate-secured loans ex perienced a 22.5 percent increase. counts increased 5.5 percent, and other operating in com e increased 16.5 percent. Taken together, these three categories accounted for only 10 percent o f total operating incom e in 1973. The average rate o f return on loans increased sub stantially during 1973, another reason for the gain in loan income. The average return was 8.67 percent, up from 8.09 percent in 1972. Operating Expenses Income from securities also boosted the banks’ total revenue. W hile income from all securities increased 10.8 percent in 1973, most of this increase resulted from securities other than those of the U.S. Treasury. These “other” securities, including obligations o f states and political subdivisions, have increased in im por tance in the asset structure in recent years. The aver age return on U.S. Treasury securities increased from 5.69 percent in 1972 to 6.23 percent last year. The obligations of states and political subdivisions earned an average rate o f 4.25 percent in 1973, up from 4.14 percent in 1972. The remaining sources of incom e all increased over their 1972 amounts, but accounted for much less of total operating income. Trust department incom e in creased 9.4 percent, service charges on deposit ac Total operating expenses rose by 31.4 percent in 1973 to almost $1.1 billion. The largest expense cate gory was the interest paid on deposits. This item rep resented 41.9 percent of total operating expenses in 1973. Interest payments on deposits increased 30.1 percent during 1973 as a result o f higher average rates paid as well as greater volume. There has been an upward trend in the ratio o f time and savings deposits to total deposits in recent years, which partially explains the increasing interest ex pense on deposits. At year end 1973, time and savings deposits made up 49.6 percent o f total deposits, hav ing increased from 47.7 percent in D ecem ber 1972. Furthermore, regulatory changes in 1973 permitted higher interest payments on some types of deposits. On July 1, maximum rates were raised on certificates of deposit less than $100,000 and interest ceilings were Page 9 F E D E R A L R E S E R V E B A N K O F ST. LOUI S JULY expenses was the cost o f Federal funds purchased and securities sold under agreements to repurchase, which almost tripled from 1972 to 1973. Distribution of Liabilities, Reserves, and Capital Accounts E ig h t h 1974 D is t r ic t M t m b a r B a n k s Remaining classifications o f expenses grew at rela tively slower rates. Salaries, wages, and benefits in creased 13 percent, and all other expenses, w hich in cludes such items as occupancy, furniture and equip ment, depreciation, and provision for loan losses, in creased 17.4 percent. Net Income 1969 1970 1971 1972 1973 suspended for certificates of deposit in minimum de nominations o f $1,000 with maturities of at least four years. A maximum rate o f 7V4 percent was imposed on the latter type of deposit on N ovem ber 1. In addi tion, maximum rates on all time deposits of $100,000 or more were phased out during the year. Conse quently, the average rate paid on time and savings deposits increased from 4.78 percent in 1972 to 5.07 percent in 1973. The largest percentage increase among the expense items in 1973 was the 191.2 percent increase in “other” interest expenses. The most significant part of these Page 10 The result o f these changes in revenues and ex penses for the Eighth District member banks was an increase in income before taxes and securities gains of $26.6 million, or 14 percent over the 1972 figure. This compares with a 17.3 percent increase in this item for all m ember banks in the nation. Although total operat ing expenses rose at a slightly faster rate than total operating incom e in 1973, the absolute increase in o p erating incom e was sufficiently large to cause an in crease in incom e before taxes and securities gains. After higher incom e taxes, low er after-tax gains on securities, and extra charges, net incom e increased 9.8 percent to $168 million in 1973. Bank Capital Capital o f the m ember banks in the Eighth District consists primarily o f equity capital. Capital notes and debentures accounted for only 4.6 percent o f total capital accounts as o f D ecem ber 1973. For the year ending D ecem ber 1973, equity capital rose 8.2 per cent and capital notes and debentures increased 5.7 percent, resulting in an increase o f 8.1 percent in total capital accounts. Net incom e as a percentage o f equity capital increased slightly, from 11.3 percent in 1972 to 11.5 percent in 1973. Branching, Holding Companies, and Banking Concentration in the Eighth District GERALD P. DW YER, JR., and W ILLIAM C. NIBLACK R a n k expansion through branching and bank holding com pany acquisitions has been the subject o f nationwide discussion in recent years. Pressure has com e from larger banks for fewer restrictions on branching and bank holding companies. The smaller banks have generally resisted such pressure and at the same time have pressed for greater restrictions on bank holding companies. Both groups have been ac tive in the Eighth Federal Reserve District, as indi cated by a number o f recent changes in state laws. The often-heated debates about branching and multiple bank holding companies are concerned with the effect o f these multi-office organizations on bank ing structure — the number and size distribution of banking organizations in an area. O f particular in terest is the effect on concentration — the extent to which bank deposits are held in a few relatively large banking organizations in a market or a state. The debate, however, is fundamentally about the effects o f increased concentration. Those who favor branching or bank holding com pany expansion typi cally argue that any increase in concentration results from greater efficiency of these multi-office organi zations and leads to improvement in services. One proponent of multi-office banking noted that in one state with state-wide branch banking: . . . there was no eviden ce o f dam age w hen a tiny, sm all-tow n bank, unable to pay the costs of autom ating and updating its facilities, unable to p ro vide new custom er services, unable to increase its lending limits and obtain funds at com petitive rates, unable to attract and train top-n otch bankers, agrees — or asks for — a m erger w ith a large institution.1 On the other hand, opponents o f multi-office banking argue that it results in a concentration o f econom ic and political power, to the detriment o f the public. 'Address of Walter J. Charlton to the Illinois Manufacturers Association, reprinted in American Banker (April 29, 1974). As one opponent of multi-office banking put it: T od a y w e are on ce again threatened b y supercon centrations o f econ om ic and political pow er. . . . Such institutions, am ong them the m ulti-office giants of banking, have grow n aw ay from the people, are no longer responsive to the individual. It’s “ the pu blic b e dam ned,” all over again.2 This article examines banking structure in Eighth District states, emphasizing concentration in states and Standard Metropolitan Statistical Areas (SM SA s). The effects o f regulation — especially regulation of branching and holding com pany activity — on bank ing concentration are considered. Then, the effects of concentration on bank performance are analyzed. REGULATION AND BANKING STRUCTURE IN EIGHTH DISTRICT STATES Bank structure can be directly affected b y regula tion of entry, mergers, branching, and acquisitions of banks by multiple bank holding companies. Since state restrictions on entry and merger do not differ significantly in the Eighth District, one w ould expect to see little difference in bank structure among the states due to entry or merger laws. On the other hand, Eighth District state laws concerning branching and multiple bank holding companies differ considerably and may therefore contribute to differences in bank ing structure among the states.3 Less restrictive regu lation of branching and holding companies can affect structure by resulting in more branches or subsidiary banks and few er independent banks. On the other 2Fred T. Brooks, “ Independent Banking: A Hometown Philosophy,” The Independent Banker (November 1973), p . 6. 3The Appendix to this article provides some details of the regulations on entry, merger, branching, and multiple bank holding companies in each of the Eighth District states. Page 11 F E D E R A L R E S E R V E B A N K O F ST. L O U I S hand, it is also possible that branch banks or holding com pa nies can increase the number of banking organizations operating in an area b y establishing de novo branches or banks in an area in which they did not pre viously operate. JULY ENTRY A N D M ERGERS IN EIGHTH DISTRICT STATES (D ecem be r 3 1 , 1 9 6 8 -D e c e m b e r 3 1 , 1 9 7 3 ) C h a n g e in N u m b e r of Banks Illinois and Missouri are unit banking states, as was Arkansas until 1972. Arkansas now allows limited branch banking, as do the remaining states in the Eighth District.4 There is no state-wide branching in the District. Bank M e rg e r s C lo sin g Banks 0 8 8 Illin o is 95 102 0 .0 4 M is s o u r i 19 22 0 .0 9 Limited B ranch B a n k in g States In d ia n a —5 7 12 Kentucky — 3 6 10 1.66 M is s is s ip p i — 4 12 16 1.33 18 19 T en ne ssee 1 1.71 0 .0 5 ♦Arkansas is treated as a unit banking state since it did not allow branching until 1972. Source: Annual Report, Federal Deposit Insurance Corporation, 1969-1972. 1973 data obtained from FDIC. Effect of Branching on Entry and Merger — O p ponents o f branching argue that branching will result in a reduction in the number o f independent banking organizations. This is likely to occur partly because branches will be opened where new banks might be established if branching were prohibited, and partly as a result of bank mergers. Such mergers are less likely to occur in unit banking states because the office of one bank w ould have to be closed or services offered at one office restricted. In recent years the limited branch banking states in the Eighth District have had fewer new banks established and more mergers than the unit banking states. As indicated in Table I, the ratio o f the num ber of mergers to the number of new banks from 1968 to 1973 ranges from zero to 0.09 for the unit banking states and from 0.05 to 1.71 for limited branch bank ing states. The number of banks increased in the three unit banking states and in one limited branch bank ing state, Tennessee, and decreased in the three other limited branch banking states. Effect of Branching on SMSA Concentration — This decreased entry and greater frequency o f mergers has resulted in a greater concentration o f deposits in branch banking states than in unit banking states. H igh concentration is especially likely for areas 4In unit banking states a bank may not have full-service branches, although one or more limited-service facilities may be permitted within a limited distance from a bank’s home office. In limited branch banking states, a bank may have more than one full-service office but may not operate fullservice offices at locations throughout the state. Arkansas is regarded as a unit banking state in the analysis that follows, since it was classified as such for four of the five years under consideration. Page 12 N ew B a n ks Ratio o f N u m b e r o f M e rg e r s to N u m b e r of N ew Banks U nit B a n k in g States A rk a n sa s* Branching Regulation and Structure 1974 smaller than states, if branching is limited to such areas. As the accom panying chart shows, the concen tration of bank deposits in Eighth District SMSAs is greater in limited branch •banking states than in unit banking states. The SMSAs included in the chart are the four largest SMSAs with population greater than 100,000 in each Eighth District state, except for Chi cago which is excluded as atypical. The concentration measure used is the “four-bank concentration ratio” — the percentage of total bank deposits in an area held by the four largest banking organizations. Since SMSAs can be taken as approximations o f market areas f o r m a n y banking serv ice s , the c o n c e n t r a tio n ratios can be interpreted as market concentration ratios.5 The amount o f business in a market also influences concentration; as the amount o f business expands, the concentration of deposits generally declines. This can be seen most easily in a highly simplified example. Suppose there is a size o f bank that is associated with minimum average cost and that entry is not regu lated. Because of com petition among existing firms 6Banking markets are likely to be confined geographically be cause of the costs of visits to a bank. Only banks in a limited area are likely to be relevant alternatives for many customers. Various factors considered in defining SMSAs, such as com muting patterns, suggest that they are integrated economi cally. SMSAs as defined in 1970 are used with one change: parts of SMSAs that are not in the same state as the central city in the SMSA are excluded. For example, only the de posits in banks located in the Missouri portion of the St. Louis SMSA are included in calculating the concentration ratio for St. Louis. One reason for doing this is that banks across a state line are likely to be less relevant alternatives for bank customers. For example, it may be more difficult to et a loan across a state line because costs associated with ling mortgages and repossession in another state would be greater. F E D E R A L R E S E R V E B A N K O F ST. LOUI S JULY 1974 P ercen tage of D e p o sits in S M S A s H eld b y Four L arge st B a n k in g O r g a n iz a tio n s D e n o te s S M S A s in U n it B a n k in g S ta te s N o te : S M S A s a re b a s e d o n 1 9 70 definitions,- o n ly c o u n t ie s in the s a m e sta te a s the central city a re in c lu d e d . F o r e x a m p le , b a s e d o n the 19 70 definition, the St. L o u is S M S A in c lu d e d the city of St. Louis a n d fo u r c o u n t ie s in M is s o u r i, a n d tw o cou nties in Illinois. Fo r this stu d y , the t w o Illin o is c o u n tie s w ere e xclud e d. and entry of new firms, the average size o f firms tends to be that which is associated with minimum average cost. Therefore, as a market expands new firms enter and concentration falls. The SMSAs in the accom pany ing chart are arranged on the basis of population — a rough measure o f market size — with the smallest on the left and the largest on the right. The fourbank concentration ratio tends to be lower in larger SMSAs than in smaller ones. For areas of approxi mately the same size, concentration is higher in the limited branch banking states than in the unit banking states. Multiple Bank Holding Company Regulation and Structure Most states in the Eighth District prohibit the formation of new multiple bank holding companies or the acquisition o f additional banks by any existing holding companies. At present, only Missouri and Tennessee allow formation o f multiple bank holding companies or further acquisitions b y them. Recent legislation in Tennessee prohibits acquisitions b y m ul tiple bank holding companies under certain circum stances. Legislation designed to restrict the size of multiple bank holding companies has also been en acted in Missouri. Both laws reflect concern over increased state concentration of bank deposits. R ecen t Activity — In the last five years, the number o f multiple bank holding companies has increased substantially in Missouri and Tennessee. In Missouri, there were 3 such companies at the end o f 1968 and 24 at the end o f 1973; in Tennessee, there were 3 at the end of 1968 and 9 at the end of 1973. The number o f banks in Missouri controlled b y these holding companies at year-end 1973 was 144, 12 times the number of banks controlled at the end o f 1968. H olding companies in Tennessee now control 48 banks, more than 5 times the number of banks con trolled at the end of 1968. Acquisitions of existing banks account for the majority of the increase in sub sidiary banks, but 7 banks in Missouri and 1 bank in Tennessee were chartered as cle novo subsidiaries. The shares o f bank deposits in Missouri and Ten nessee controlled by multiple bank holding companies Page 13 F EDERAL. R E S E R V E B A N K O F S T LOUIS have increased dramatically in the last five years. These companies controlled 9.4 percent of total Mis souri bank deposits in 1968 and 55 percent in 1973. In Tennessee, they controlled 3.5 percent of total bank deposits in 1968 and 49 percent in 1973. However, examining the effect of multiple bank holding companies on the share of state deposits in this way overstates the increase in the share of state deposits controlled by large organizations. The de posits in Missouri and Tennessee that are controlled by holding companies consist largely of deposits in the com panies’ lead banks. At year-end, deposits in lead banks were 37.1 percent of Missouri bank de posits and 39 percent o f Tennessee bank deposits. These percentages accounted for 67.3 and 79.6 per cent o f the total deposits controlled by multiple bank holding companies in Missouri and Tennessee, re spectively. Thus, while the proportion of state deposits controlled by holding companies has increased dra matically, that increase primarily represents form a tion of holding companies by larger banks, rather than an increase in the concentration o f deposits in large banking organizations. State Concentration — A preferable way of look ing at the effect o f holding com pany acquisitions on concentration is to consider the effect on state concentration ratios, the share o f deposits in a state held by a specified number of the largest banking organizations. This differs from looking at the share o f deposits controlled by all multiple bank holding companies, since the smaller holding companies are not considered and the number of organizations is held constant. If it is assumed that deposits of subsidiary banks grew at the same rate after acquisition as they would have without acquisition, then the increase in con centration is simply the difference between the actual concentration ratio and a calculated ratio which as sumes no banks were acquired after 1968. Thus, it is estimated that multiple bank holding companies in creased the four-bank concentration ratio by 9.4 per centage points in Missouri and 4.4 percentage points in Tennessee (see Table II). The comparable figures for the ten-bank concentration ratios are 12.8 per centage points for Missouri and 10 percentage points for Tennessee. SMSA Concentration - The same procedure can be used to estimate the effect of multiple bank holding companies on the four-bank concentration ratios in SMSAs. No Tennessee holding com pany ow ned more than one bank in any o f the four largest SMSAs by Page 14 JULY 1974 T a b le II C O N C E N T R A T IO N O F DEPOSITS IN LARGE B A N K IN G O R G A N IZ A T IO N S (December 31, 1973) Perce ntage of D e p o sits H e ld b y Fou r La rgest B a n k in g O r g a n iz a t io n s in the State A ctu a l C o n ce n tra tio n Ratio N o -a c q u is it io n C on ce ntra tio n R a tio 1 In cre a se in C o n c e n tra tio n - M is s o u r i 3 1 .9 % 2 2 .5 % 9 .4 % T en ne ssee 3 6 .5 32.1 4 .4 Perce ntage o f D e p o sits H eld b y Ten La rgest B a n k in g O r g a n iz a t io n s in the State A ctual C on ce ntra tio n Ratio N o -a c q u is itio n C on ce ntra tio n R a tio 1 Increa se in C o n c entratio n 2 M is s o u r i 4 6 .4 % 3 3 .6 % 12 . 8 % Tenne ssee 6 2 .7 5 2 .7 10 *The no-acquisition concentration ratio gives the percentage of deposits held by the largest banking organizations, assuming no holding company acquisitions were made after December 31, 1968. 2The increase in concentration due to acquisition by holding com panies between December 31. 1968 and December 31, 1973 is the difference between the actual and the no-acquisition concentration ratios. the end o f 1973. If it is true that affiliation with a holding com pany neither increases nor decreases a bank’s growth, then in Tennessee multiple bank hold ing companies have not affected SMSA concentration. On the other hand, holding companies have acquired two or more banks each in some Missouri SMSAs. In St. Louis, the concentration in the four largest bank ing organizations is 3.7 percentage points higher than the 44.7 percent without such acquisitions; in Kansas City, concentration is 5.3 percentage points higher than the 52.4 percent; and in Springfield, it is 4 per centage points higher than the 82 percent without such acquisitions. Undoubtedly, a contributing cause to this difference betw een Missouri and Tennessee is the prohibition of branch banking in Missouri and permission of county-wide branching in Tennessee. INTERPRETING HIGHER CONCENTRATION The foregoing discussion has shown that less re striction of branching and multiple bank holding companies is associated with higher concentration, but has given no basis for evaluating the significance o f higher concentration. This is at the heart o f the controversies about branch banking and multiple bank holding companies. An increase in a concentration ratio merely says that the percentage of deposits held by the largest banks has increased. It is the expected F E D E R A L R E S E R V E B A N K O F ST. L OUI S JULY 1974 effects o f this increase on the performance o f the banking organizations that are of interest, not the increase itself. late and prepare their cases too scantily and hastily to b e fu lly effective w here their interests conflict w ith those o f large com panies.8 Increases in concentration o f banking resources in states have a com pletely different interpretation than increases in SMSAs. Concentration in a state is an aggregate concentration measure; concentration in an SMSA is a measure o f market concentration. The essential aspects o f an aggregate concentration meas ure are that it is measured for a political entity, that it includes several markets, and that it emphasizes the relative sizes attained by some firms operating in several markets. Market concentration is measured, to the extent possible, for a geographic market — “the area within which the price of a com m odity tends to uniformity, allowance being made for transportation costs.”0 Mutual Forbearance — The hypothesis that bank holding companies operating in several markets can agree on, and therefore affect, prices in those markets can be illustrated with a simple example. Suppose there are ten banking markets in a state and two banks in each market. Initially these banks are in dependently owned, but subsequently each o f two holding companies in the state acquires one bank in each market. The hypothesis — sometimes called “mutual forbearance” — is that the tw o holding com panies w ould be more likely to reach an agreement to raise prices in the ten markets than w ould the two banks in each market separately.9 Thus, the prices paid for services by bank customers w ould be ex pected to rise on average. State Concentration The effects some observers have attributed to an increase in banking concentration at the state level are on the prices charged and influence over the state government’s legislative process.7 It is argued that the effect on market prices occurs through the ability o f large organizations operating in several markets to agree among themselves and to intimidate smaller firms. The effect o f large organizations on the legislative process occurs because: Large com panies start w ith certain initial political disadvantages because they are in the spotlight, b e cause there is som e suspicion o f their pow er, and becau se small com panies are m ore numerous. H o w ever, the large com pany can often ov ercom e its handicap and obtain a d ecid ed advantage b y political expenditures. T h e cam paign contributions o f large com panies and the occasional case o f direct or in direct bribery are p robably the least significant sources o f the large com pa n y’s political pow er. M ore im por tant, the large com pany spends w hatever m on ey is n eeded to argue effectively on beh alf o f its interest w here a political issue affects it. . . . W h ile some smaller business interests m ake a com parable show ing through associations set up for the purpose, the experience o f a W ashington official is that small com panies generally find ou t w hat is happening too 6George J. Stigler, The Theory of Price (N ew York: The MacMillan Company, 1966), p. 85. 7These are the effects briefly mentioned by Samuel H. Talley, “ The Impact of Holding Company Acquisitions on Aggregate Concentration in Banking,” Staff Economic Studies, no. 80, Board of Governors of the Federal Reserve System, pp. 1-2. The problem is analytically the same as that discussed by Corwin D. Edwards, “ Conglomerate Bigness as a Source of Power,” Business Concentration and Price Policy (Princeton: Princeton University Press, 1955), pp. 331-52, and in an ac companying “ Comment” by George W . Stocking, pp. 352-59. But are there really forces leading to mutual for bearance? W ou ld the tw o holding companies in this example be more likely to agree than the twenty individual banks? The arguments supporting this hypothesis are that the benefits from such an agreement w ould be greater, and the costs o f deviating from it — chiseling — would also be greater than if all the banks were individually owned. One agreement, instead o f ten, could apply to all markets; therefore, the benefits w ould be greater from an agreement. If one o f tw o banks in a market cheated on the agreement, that is, low ered prices for at least some customers, then the decrease in prices w ould occur in only one market. But if one o f the two holding companies cheated in a market, then the “price war” could spread to all markets. Thus, the costs o f deviating from an agreement w ould also be greater. The mutual forbearance hypothesis is not, however, substantiated by any empirical tests for banking or other sectors o f the U.S. econom y. Furthermore, the arguments for it are less than com pletely convincing. The inherent desirability o f one agreement is dubious. The primary defect o f one overall agreement is the existence o f different demand and cost conditions in different markets; under these conditions, ten separate 8Edwards, “ Conglomerate Bigness,” pp. 346-47. 9An increase in state concentration of deposits does not, how ever, necessarily reflect an increase in the probability of mu tual forbearance. Such an increase may result from deposit growth in only one bank or acquisitions in markets where no other state-wide banking organizations operate. In these cases, banking organizations would not be racing each other in more markets than before. Page 15 F E D E R A L R E S E R V E B A N K O F ST. L O U I S agreements could be superior from the holding com panies’ point of view. Also, it has in no way been established that an appropriate response to chiseling in one market w ould be a price war in all ten markets. This w ould only ensure that the decreased “profits” in one market w ould spread to ten markets. Predatory Pricing — The ability of large organiza tions operating in several markets to discipline smaller ones operating in one market is the other alleged way that large organizations could increase prices paid by customers. In other words, if a small firm should be so bold as to chisel, it is argued that the larger organiza tion w ould respond by cutting its prices in order to decrease the “profits” o f the small chiseler.10 This practice has often been called “predatory pricing” or “cutthroat com petition.” It may be designed either “to teach the chiseler a lesson,” to drive him into a merger, or to force him out of business. The large firm w ould have no differential advantages when it cuts prices unless it incurs costs greater than its revenue. Both the larger and the smaller firms w ould simply make less “profit” than they w ould otherwise. The large firm could only have a differential advantage when it is incurring losses. The advantage attributed to large firms when they cut prices below their cost and their competitors’ costs is superior access to capital — a lower price paid for capital. They could finance the losses at a lower cost than smaller firms; therefore, they w ould have a differential advantage in a price war. This low er cost of capital to large firms purportedly exists because the price war could be financed b y funds generated in other markets where the larger firms are receiving “m onopoly profits.” But the large firm would not really have a low er cost of capital. The cost of anything is the highest-valued alternative foregone. And as long as one of its alternatives w ould be to loan to a small firm (with a cost of capital the same as the small firm it is trying to intimidate), then this alternative would be superior to financing a price war. An additional argument against the likelihood of predatory pricing is summarized by George J. Stigler in a fictional discussion in which a potential victim of predatory pricing b y John D. Rockefeller tells a lender: 10This decrease need not involve a price cut below marginal cost, as has been argued by B. S. Yamey, “ Predatory Price Cutting: Notes and Comments,” Journal of Law and E co nomics (April 1972), pp. 129-42. He also provides a sum mary and evaluation of the literature generated by the classic paper on this subject, John S. McGee,, “ Predatory Price Cutting: The Standard Oil (N . J.) Case,” Journal of Law and Economics (O ctober 1958), pp. 137-69. Page 16 JULY 1974 “ T h ere is a threat o f a three-m onth price war, dur ing w hich I w ill lose $10,000, w h ich unfortunately I do not possess. If you lend m e the $10,000, I can survive the price w ar — and on ce I show your certi fied ch eck to R ockefeller the price w ar w ill proba bly never b e em barked upon. E ven if the p rice w ar should occu r, w e w ill earn m ore b y co-operation after w ard than the $10,000 loss, or R ockefeller w ou ld never em bark u pon the strategy.” 11 And indeed, Rockefeller did buy out his rivals rather than try to drive them out o f business b y such tactics. It is noteworthy that one cannot conceive that bank regulators w ould allow a price war. As a matter of fact, much of bank regulation is designed to suppress price competition, replacing it with other forms o f competition. As Ray M. Gidney, a recent Comptroller of the Currency, said: I think the im portant thing w e should h op e for is a degree o f enlightenm ent on the part o f p e o p le that run these banks so that they go out and give service. That is w here w e w ant com petition, com petition in giving service.12 Concentration o f Political P ow er — For large firms to have a disproportionate effect on legislation, it w ould be necessary for them to have a differential advantage in contributing to political campaigns, d e livering votes, or providing information to legislators. These are the three means by which politicians’ votes in legislatures can be influenced.13 At least in banking, there is no presumption what soever that large organizations have any differential superiority in any o f these activities. Campaign con tributions by corporations are illegal, and therefore will not generally be made. The wealth position of small banks’ stockholders generally w ould be more substantially affected by banking legislation than the wealth position of stockholders o f large banking organizations. Therefore, the small number o f stock holders o f numerous small banks are more likely to make contributions and vote on the basis of legislators’ votes on banking legislation than are the numerous stockholders of a few holding companies. It is also unlikely that large holding companies have any supe riority over small banks in supplying information to 11George J. Stigler, The Organization of Industry wood: Richard D. Irwin, Inc., 1968), p. 116. (H om e 12U. S. Congress, Senate, Regulation of Bank Mergers, 86th Cong., 1st sess., 1959, Committee on Banking and Currency, p. 31. For an economic analysis of why producers prefer to substitute nonprice competition for price competition, see Stigler, The Organization of Industry, pp. 23-28. ,:iSee Albert Breton, The Economic Theory of Representative Government (Chicago: Aldine Publishing Co., 1974), pp. 74-98. F E D E R A L R E S E R V E B A N K O F ST. L O U I S legislators. Bankers, through state and independent bankers associations, are well organized, whether they are large or small. Thus, just as increased state concentration does not necessarily have any effects through mutual forbear ance or predatory pricing, there is no substantial effect o f increased state concentration on the legis lative process. Market Concentration There are two hypotheses related to increased mar ket concentration. One hypothesis is that increased concentration is associated with low er prices charged to customers due to the efficiency o f large-scale op erations. A ccording to the second hypothesis, in creased concentration facilitates agreement — explicit or tacit collusion — about prices in the market and is thus associated with higher prices charged to custom ers.14 Those w ho are in favor o f relatively few re strictions on branching and acquisitions b y multiple bank holding companies assert the first hypothesis when presenting their views. Those w ho are opposed to branching and holding com pany acquisitions of banks refer to the second hypothesis when interpret ing the higher concentration which results from branching and multiple bank holding companies. E fficiency — The hypothesis relating efficiency and concentration is an attempt to answer an important question: H ow does higher concentration develop? Higher concentration is synonomous with higher mar ket shares o f the larger firms. The arguments advanced in the discussion o f state concentration imply that high market concentration does not result from predatory pricing b y larger firms. A firm’s larger share o f a market means that a larger percentage of the business in the market is going to that firm. This larger share o f the business must re sult because customers prefer that firm to others, per haps because o f a low er price or a higher-quality product. Thus, it can be argued that concentration is higher because some firms are relatively more efficient. This argument is weaker for banking than for other industries because o f entry and branching regulations. Implicit in the argument is the assumption that the most efficient firms in servicing customers are those 14The use of the term “ collusion” in this paper is not to be confused with the legal definition. Collusion as used in this paper includes legal collusion and a great deal more — any explicit or implicit agreement — not just explicit agree ments. Those setting the prices need not even intend to agree; they need only act as if they agree. JULY 1974 that are presently operating and that expansion is largely determined by relative efficiency. But entry regulation is designed precisely to protect existing banks from the com petition of new banks; this is also a consideration in branching applications. Also, bank ing regulators determine to some extent w ho receives charters and which banks grow through branching. The interests o f regulators and bank customers are not necessarily coincident. Existing empirical evidence does not falsify the hypothesis that increased concentration results from the relative efficiency o f larger banks.15 Collusion — The increased concentration, according to the second hypothesis, facilitates collusion among banks. Just as a single firm operating in a market maximizes the wealth of all the owners of the firm, an agreement betw een firms operating in a market can maximize the wealth of all the owners of the firms. There is, o f course, a problem of distributing the gains from the agreement, a problem which does not arise for a monopoly. Each of the firms in a market can increase the wealth of its owners if it cheats on the agreement while all other firms adhere to it. The chiseling can take various forms. It may be secret price cutting, changes in credit terms, or changes in the quality of the product in some other way. W hether the firm is cutting its pecuniary prices or making the nonpecuniary terms o f its sales more attractive to buyers, it will desire to keep its actions secret. If the other firms 15Much of the literature on the relationship between bank’s sizes and efficiency — so-called “ economies of scale” — is reviewed and analyzed in Jack M. Guttentag and Edward S. Herman, “ Banking Structure and Performance,” The Bulletin of New York University Graduate School of Business Administration (February 1967), pp. 105-25, 169-96. Two recent studies are: Frederick W . Bell and Neil B. Murphy, “ Costs in Commercial Banking,” Federal Beserve Bank of Boston Research Report No. 41; and Lionel Kalish III and R. Alton Gilbert, “ An Analysis of Efficiency of Scale and Organizational Form in Commercial Banking,” Journal of Industrial Economics (July 1973), pp. 293-307. The empirical evidence generally indicates that banks of less than about $10 million deposit size are relatively high-cost banks and that other banks probably have about the same average cost. There are, however, unresolved conceptual and measure ment problems, which apply to all studies of economies of scale in banking, that imply the empirical evidence must be interpreted carefully. Two of these problems are the inability to account adequately for changes in the types of deposits received and loans made as sizes of banks expand and the individual nature of banks’ cost functions. For elab oration and analysis of the latter problem, see Milton Fried man’s “ Comment,” pp. 230-38, on Caleb A. Smith, “ Survey of the Empirical Evidence on Economies of Scale,” pp. 21330, in Business Concentration and Price Policy, and Harold Demsetz. “ Industrv Structure. Market Rivalry, and Public Policy,” Journal of Law and Economics (April 1973), pp. 1-9. Interpreted in the light of Friedman’s and Demsetz’s analyses, the empirical evidence is consistent with larger banks being larger because they are more efficient. Page 17 F E D E R A L R E S E R V E B A N K O F ST. LOUI S find out, they will take account o f the price cutter’s actions and set a low er price or improve the nonpecuniary terms o f their sales. This antagonism between the interests o f the firms with an agreement is the basis o f the relationship between market concentration and the prices charged customers in markets. The smaller the number o f firms in an agreement and the larger their relative market shares, the easier effective collusion is likely to b e.16 Another enemy o f collusion, second to existing firms’ mutual antagonism, is entry of new firms into the market. Those not presently in the market are not part o f the agreement and, seeing the returns being received by firms in the market, are likely to be in duced to enter. If they are subsequently made part o f the agreement, the returns to those originally co l luding are diluted. If they are not made part o f the agreement, then the effect on the collusive prices is the same as that o f a discovered chiseler — a lower price. Entry and the possibility of it can attenuate the relationship betw een concentration and prices. But regulation o f entry in banking effectively protects many markets from entry and in all cases reduces the probability o f entry. If increased collusion follow s from increased con centration, one w ould expect to observe higher loan rates and higher rates o f return for banks. The exist ing empirical evidence does not falsify this hypothesis either.17 16Thomas R. Saving, “ Concentration Ratios and the Degree of Monopoly,” International Economic Review (February 1970), pp. 139-46, briefly presents the theory behind the concentration ratio as a concentration measure. George J. Stigler “ A Theory of Oligopoly,” Journal of Political Econ omy (February 1964), pp. 44-61, analyzes the mutual antagonism and relates enforcement costs to the Herfindahl index of concentration. Concentration ratios and Herfindahl indexes are highly correlated. 17Guttentag and Herman, “ Banking Structure and Perform ance,” pp. 80-104, also review the evidence accumulated before 1967 on the relationship between concentration, prices, and “profits.” Additional evidence is provided by Donald P. lacobs, Business Loan Costs and Bank Market Structure: An Empirical Estimate of Their Relations, O c casional Paper 15, National Bureau of Economic Research, and Arnold Heggestad, “ Market Structure, Risk, and Profit ability in the Banking Industry,” in Proceedings of a con ference on Bank Structure and Competition, Federal Reserve Bank of Chicago, 1972. Taken as a whole, the empirical evi dence is consistent with a statement that higher concentra- Page 18 JULY 1974 CONCLUSION The controversies over branching and multiple bank holding com pany expansion are concerned with the effect o f these multi-office organizations on con cen tration. In larger SMSAs, the Eighth District states with limited branch banking tend to have higher concentration o f deposits in the four largest banking organizations than the unit banking states. State con centration has increased as a result o f the rapid ex pansion o f multiple bank holding companies in Mis souri and Tennessee, but the effect on SMSA con centration has been somewhat less. These findings are o f little interest until the impli cations of increased concentration are assessed. It has been shown that there are no convincing arguments and no empirical evidence that state concentration has any significance. On the other hand, evidence has been advanced w hich supports both the hypothesis that increased market concentration results from the efficiency of large organizations and the hypothesis that increased concentration facilitates collusion among the organizations. The relationship between efficiency and concentration, b y itself, implies that banks’ customers gain as a result o f higher concentra tion; but the relationship betw een collusion and con centration, b y itself, implies that banks’ customers lose as a result o f higher concentration. Less restrictions on branching and multiple bank holding companies is associated with higher concentration. Therefore, there are both potential benefits and costs for banks’ cus tomers from such lessened restrictions. Since it is so often thought to be implied, it is nec essary to point out that this amalgam o f collusion and efficiency does not imply that omniscient regulators could weigh the costs and benefits of higher concen tration and determine a more “optimal” banking struc ture than that which w ould develop by permitting free market forces to operate. Such a weighing process is impossible. The evidence regarding both costs and benefits is imprecise and h obbled by the use o f avail able accounting data to measure econom ic concepts. It is unlikely that the factors that result in imprecision will soon be overcom e and even more unlikely that present accounting data can be replaced b y more relevant data. tion results in higher rates on loans and higher rates of return for banks. APPENDIX: SELECTED BANK REGULATIONS IN EIGHTH DISTRICT STATES Regulation of banking can directly affect banking structure through limitations on entry, mergers, branch ing, and acquisitions of banks by multiple bank holding companies. All of these activities are subject to regula tion at both the Federal and state levels.1 Entry Before a bank may begin operations, Federal or state banking officials must grant a charter to the bank’s or ganizers. The division of authority for examining charter applications, as well as applications for other actions dis cussed in this Appendix, is shown in Table A-I. One purpose of entry regulation is to prevent a new bank from opening where it might fail or its opening might cause an existing bank to fail. Therefore, the Fed eral banking officials consider the future earnings pros pects of the bank and the effect of entry on the “ sound ness” and earnings prospects of existing banks. They also consider the effect of the new bank on the “convenience and needs of the community to be served” — that is, the extent to which the new bank might provide more serv ices or services at lower cost than are currently available. The state laws, which are similar to the Federal laws, are outlined in Table A-II. The concern of bank chartering agencies that existing banks not be hurt by new entry and that new banks meet a “needs” criterion has resulted in fewer banks be ing chartered than would have been in the absence of such regulation.2 One effect of this relative difficulty of entry has been to limit the development of competition in banking. T a b le A -I Division of Authority Am ong Bank Regulatory Agencies Agency C h a r t e r in g * B ra n c h in g C om p troller of the C urren cy N a tio n a l b a n k s N a t io n a l b a n k s R e sulting b a n k a n a tio n a l b a n k Federal Reserve B o ard State m em ber b a n k s State m em ber b a n k s R esulting b a n k a state m em ber bank Federal D ep osit In surance C o rp o ra tio n N o n m e m b e r in su red state b a n k N o n m e m b e r in su re d state b a n k R esulting b a n k a n onm em ber in su re d b a n k State B a n k in g A g e n c ie s A ll state b a n k s A ll state b a n k s R esulting b a n k a state b a n k M e rg e rs ^Charters for national banks are issued by the Comptroller o f the Currency, and state bank charters are issued by state banking agencies. The Federal Reserve Board reviews charter applications when acting on applications o f state banks for membership in the Federal Reserve System. The FDIC reviews charter applications when acting on applications of state nonmember banks for deposit insurance. JFor a more detailed discussion of Federal regulation, see Gerald C. Fischer, American Banking Structure (N ew York: Columbia University Press, 1968). Tables at the end of this Appendix provide some of the details of state laws on entry, mergers, branching, and multiple bank holding companies. -For a discussion of entry into banking, see Sam Peltzman, “ Entry in Commercial Banking,” Journal of Law and Eco nomics (O ctober 1965), pp. 11-50. Page 19 FED ERAL Page 20 T a b le A - ll A rk a n sa s 2. C ap ita l Requirem ents 5 or m ore q u a lifie d n a tu ra l p e r so n s, m a jo rity A r k a n s a s re si Requisite cap ital sub scrib ed in g o o d faith, the minimum d e p e n d in g on com m unity size. dents, w ith confid en ce of 3. B a n k ’s Prospects 4. C o n v e n ie n ce & N e e d s 5. A d v e rs e Effect on O th e r B a n ks There exists a p ub lic necessity o f the b u sin e ss o f the com BANK 1. Pe rso n s A p p ly in g RESERVE State Banking Laws on Entry m unity. com m unity. OF 5 or m ore in co rp o ra to rs, re si M e e ts m inimum ca p ita l re q u ire ments, the m inim um d e p e n d in g Future e a r n in g s fa v o ra b le . C o n v e n ie n ce a n d n eeds o f a rea so u g h t to be served will be p r o moted. on com m unity size. L O U IS dents o f Illin o is, w ith g e n e ra l ch aracter such th a t “ r e a s o n a b le ST. Illin o is p ro m ise of successful o p e ra tio n e x ist s ." In d ia n a Ke ntucky M is s is s ip p i M e ets m inim um capital re q u ire A p p r o v a l o f D ep artm en t of A p p r o v a l b y D ep artm en t of A p p r o v a l b y D ep artm en t of Fin a n c ia l Institutio ns “ in its Fin a n c ia l In stitutio ns “ in its Fin a n c ia l In stitutio ns “ in its residents. ments, the m inim um d e p e n d in g on com m unity size. d isc re tio n ’’. d isc re tio n ". d isc re tio n ” . 5 or m ore p e rso n s fo r b a n k ; 7 o r m ore fo r trust; 7 o r m ore for M e e ts m inim um ca p ita l re q u ire m ents, the m inim um d e p e n d in g R e a so n a b le a ssu ra n c e of s u f Public co nve nie nce a n d a d v a n ficient volum e of b u sin e ss for ta g e w ill be prom oted. b a n k & trust; 1 3 o r m ore fo r b a n k , trust, a n d real estate title. on city size. success. 5 o r m ore p e rso n s o f full a g e a n d o f g o o d m oral a n d b u sin e ss C a p ita l stock a n d surplus re R e a so n a b le prospects o f grow th D eterm ine w h ethe r the p ublic qu ire d , the m inimum d e p e n d in g n ece ssity req u ire s that the p ro character. on com m unity size. of the a re a a n d its fin a n cia l re sources, ex p e cta tion s o f profit p o se d new b a n k sh o u ld be (n o t b ra nch b a n k s ) a n d e f a b le op e ra tio n s. chartered a n d a llo w e d to fect on them. 1 0 o r m ore n a tu ra l p e rso n s of law fu l a g e , m ajority In d ia n a Record of e a r n in g s a n d co n d ition o f e x istin g unit b a n k s operate. M is s o u r i 5 or m ore p e rso n s w ith c h a r a c ter, re sp o n sib ility , a n d g e n e ra l fitness of p e rso n such as to com Requisite capital sub scrib ed in P ro b a b le volum e o f b u sin e ss in C o n v e n ie n ce a n d need o f com P ro b a b le volum e of b u sin e ss g o o d faith a n d p a id in cash, the m inimum d e p e n d in g on lo ca lity sufficient to in su re a n d m a inta in so lv e n c y of the new m un ity ju stify a n d w a rra n t o p e n in g o f new b a nk. m and confidence. com m unity size. bank. in lo ca lity sufficient to m a in tain so lv e n c y o f new b a n k a n d e x istin g b a n k s, w ithout e n d a n g e r in g sa fe ty of a n y b a nk. T en ne ssee A d e q u a te capital a n d surplus, the m inim um d e p e n d in g o n C o n d itio n s in com m unity afford re a so n a b le p ro m ise o f successful p ro p o s e d officers a n d directors h a v e sufficient exp e rie n ce , a b ility , a n d s t a n d in g to a s su re r e a s o n a b le p ro m ise o f successful com m unity size. op e ra tio n. P ublic need a n d a d v a n t a g e s prom oted. N e e d in com m u nity fo r new b a n k , c o n sid e rin g a d e q u a c y of e x istin g b a n k s. JU LY 5 o r m ore in co rp o ra to rs, m a jority re sid e n ts o f T e n ne ssee ; o p e ra tio n . 1974 F E D E R A L R E S E R V E B A N K O F ST. L O U I S JULY 1974 T a b le A - ll l State I nking Laws on Mergers A rk a n sa s Illin o is In d ia n a Kentucky M is s is s ip p i M is s o u r i T e n n e sse e M e rg e rs a llo w e d with sufficient d is closure. R esulting b a n k meets re q u ire m ents for fo rm in g new b a n k a n d agre e m e n t fa ir to all concerned. A g re e m e n t of sh a re h o ld e rs a n d b o a rd s of the m e rg in g b a n k s; p rior a p p ro v a l b y D ep artm ent of Fin ancial Institutio ns “ in its d isc re tio n ." D ocum ents s u b mitted to stock h o ld e rs; institu tions located in sam e city or county. M a jo rit y of b o a rd o f directors; terms o f co n d itio n s la w fu lly a g re e d upon. C o n se n t of ho ld e rs o f twothirds o f capital stock g e n e ra lly ; d isclo su re o f a greem ents. Sa m e county; meets re q u ire m ents fo r fo rm in g n ew b a n k ; a g r e e ment fa ir; not c o n tra ry to p ub lic interest. Mergers Prior approval by a Federal banking agency is required for all mergers resulting in an insured bank. If the result ing bank is to be a state bank, approval of state banking officials is also necessary. The provisions of state laws on mergers in Eighth District states are outlined in Table A-III. In deciding whether to approve or deny a merger ap plication, Federal banking officials are required by the Bank Merger Act to consider the effects of the merger on competition, the future prospects of existing and pro posed institutions, and the convenience and needs of the community to be served. No merger which would result in a monopoly or would tend to create a monopoly can be approved. If a merger would result in a “substantial lessening of competition,”3 it can be approved only if the probable effects of the merger in meeting the conveni ence and needs of the community to be served clearly outweigh the anticompetitive effect. Branching All banks in a state, whether national or state-chartered, are subject to state laws concerning the locations of branches, the number of branches allowed, the services that may be offered, and the capital required for opening a branch. There are both unit banking and limited branch bank ing states in the Eighth District. The least restrictive state law (Mississippi) permits branch banking within 100 miles of a bank’s home office. The provisions of the present state laws are contained in Table A-IV, p. 22. ■'This term is not defined in legislation but refers to a less ex treme reduction in competition than would result from crea tion of a monopoly. Multiple Bank Holding Companies F ederal — Prior to 1956, the bank acquisitions of mul tiple bank holding companies were virtually free of Federal regulation. The Bank Holding Company Act of 1956 brought this activity under Federal control by re quiring prior approval by the Federal Reserve Board of any action resulting in ownership of 5 percent or more of a bank’s stock by a company owning two or more banks. In determining whether or not to approve an acquisition, the Board was required to consider banking and com petitive factors, but the competitive factors were given less emphasis than the so-called banking factors. In addi tion, acquisition of a bank outside a holding company’s home state was prohibited, unless the acquired bank’s state law specifically allowed such acquisitions. Amendments to the Bank Holding Company Act which were passed in 1966 and 1970 shifted the emphasis from banking factors to competitive factors and brought onebank holding companies under the Board’s supervision. The 1966 amendments provided that the Board apply the same tests in considering acquisitions by bank holding companies that Federal officials apply in bank merger cases. The 1970 Amendments brought companies owning one bank under the purview of the Board. State — Formation of multiple bank holding companies is prohibited in the majority of the Eighth District states. This has been the case for some time in Illinois, Indiana, Kentucky, and Mississippi. In 1971, an Arkansas law was passed which prohibits companies from becoming multiple bank holding com panies or existing holding companies from making any further acquisitions; prior to that time, there had been no restrictions on such companies in Arkansas. Until this year Missouri and Tennessee had no restrictions on multiple bank holding companies; both states now have laws which limit the size of holding companies. The Tennessee Bank Structure Act of 1974 places additional restrictions on holding company acquisitions. Details of state laws are provided in Table A-V, p. 23. Page 21 Table A-IV State Banking Laws on Detached Offices W h ic h B an ks N u m b e r of D etach ed O ffices Location Pow ers A rk a n sa s A n y b ank. N o limit on n um ber o f fullservice branches. W ith in co rp o rate limits of h om e-office city o r tow n if gre a te r tha n 3 0 0 feet from a n y other b a n k ; in a n y city o r tow n in hom e-office coun ty with p o p u la tio n at le ast 2 5 0 a n d n o hom e office o f a n o th e r e x istin g b a n k ; in a n y p la n n e d com m u nity d e velop m e nt in hom e-office county w ith p o p u la tio n at least 2 5 0 ; n o th in g a llo w s b ra n c h e s o u tsid e hom eoffice county. A ll la w fu l b a n k in g activities a s fu lly a s in the m ain office. Illin o is A n y bank. O n e Facility. N o t m ore tha n 1 5 0 0 feet from hom e office a n d g e n e ra lly m ore than 6 0 0 feet from other b a n k s ' prem ises. Receive d e p o sits, cash a n d is sue checks, d ra fts a n d m on ey o rd ers, c h a n g e m o n e y a n d receive p a ym e n ts on e x istin g debts. In d ia n a (a ) (a ) O n e b ra nch for every $ 2 0 0 , 0 0 0 of ca p ita l a n d surplus. (a ) A n y w h e re in county. (a ) (b ) B a n k s in counties with p o p u la tio n less than 5 0 0 . 0 0 0 or h a v in g 3 or m ore seco n d -c lass cities. (c) B a n k s in counties with p o p u la tio n gre a te r than 5 0 0 . 0 0 0 a n d less than 3 se co n d -c la ss cities. (b ) N o limit. (b ) (c) N o limit. (b ) A n y city o r town in hom e-office coun ty if no b a n k located in the city or town. (c) A n y city o r town in county. ( a ) A n y b a n k with capital a n d su rp lu s not less than $ 1 0 0 , 0 0 0 with p rincip al o f fice in an a re a with p o p u la tion less than 8 ,0 0 0 . ( a ) O n e b ra nch for each $ 1 0 0 , 0 0 0 o f ca p ita l a n d surplus. (b ) A n y b a n k with cap ital a n d su rp lu s not less than $ 2 0 0 , 0 0 0 w ith p rin c ip a l o f fice in a city with p o p u la tio n of 8 , 0 0 0 o r m ore a n d less than 2 0 ,0 0 0 . (c) A n y b a n k in city with p o p u la tio n o f 2 0 , 0 0 0 or more. (b ) O n e b ra nch for each $ 2 0 0 , 0 0 0 o f ca p ita l a n d surp lus. Kentucky M is s is s ip p i A n y bank. T en ne ssee Sam e as ( a ) . (a ) S a m e p ow e rs a s p rin c ipa l office. (b ) Sam e as ( a ) . (c) O n e branch fo r each $ 2 5 0 , 0 0 0 of ca p ita l a n d surplus. (c) (c) Sam e as ( a ) . (a ) W ith in hom e-office city if p o p u la tio n at least 1 0 , 0 0 0 a n d w ithin hom e office c o u n t y a n d a d ja ce n tco u n tie s, but not in a n y city o r town with p o p u la tio n less than 3 5 0 0 a n d o n e or m ore e x ist in g b a n k s or b ra nch b a n k s. (b ) W it h in ra d iu s o f 1 0 0 miles from hom e office, but not in a n y city o r tow n with p o p u la tio n less than 3 1 0 0 a n d o n e o r m ore e x istin g b anks. (a ) (a ) W ith in limits o f city, tow n, o r v illa g e o r u n in co r p ora ted a re a in w hich hom e office is located a n d hom eoffice coun ty (b u t not w ithin 4 0 0 feet of a n o th e r b a n k g e n e r a l ly ) . (b ) In a tow n with p o p u la tion o f 1 5 5 0 o r less, w hich d o e s not h a ve b a n k in g s e rv ices a n d is not m ore tha n 10 m iles from the b a n k 's m ain office. ( a ) Checks p a id , d e p o sits received, d e p o sits w ith d ra w n , c h a n g e m ade, e x c h a n g e m ade, b a n k m on ey o rd ers issu ed , sa fe d e p o sit b o x e s m a in ta in e d a n d rented a n d lo a n p a ym e n ts received. ( b ) S a m e a s (a ) A n y bank. ( a ) B ranch offices — limit. (b ) S a m e a s (a ) (b ) B ranch b a n k s — m a x i mum of 1 5; ca p ita l re q u ire ments of $ 1 0 0 , 0 0 0 p lus m inim um cap ital req u ire d for unit b a n k at location. (a ) A n y b ank. (a ) T w o facilities. (b ) O n e facility. (b ) B a n k s in th ird -class counties w ith p o p u la tio n of 3 5 , 0 0 0 o r less. Sam e as ( a ) . ( a ) C o u n ty -w id e b ra n c h in g except if hom e office o f a n other b a n k in tow n o r city (b u t not b ra n c h in g b a n k 's hom e-office tow n ) o r if hom e office o f a n o th e r b a n k in u n in co rp o ra te d a re a w ith in o n e mile. ( b ) S a m e a s (a ) (a ) M is s o u r i (c) (P ow e rs of b a n k .) no Sam e as (a ) (b ) ( Pow ers o f b a n k . ) S a m e a s ( a ). ( Pow ers of b a n k .) N o statu to ry limit on the W ith in hom e-office county, n u m b e r o f b ra nches. 1L oan s m ay n ow be m ad e at these facilities, p u rsu an t to M is s o u r i C o m m issio n e r of Finance R ulin g N o . 15 of Ju ne 2 7 , 1 9 7 4 . A n y bank. Page 22 FEDERAL T a b le A - V State Banking Laws on Multiple Bank Holding Com panies IL L IN O IS IN D IA N A KENTUCKY M I S S IS S IP P I C o m p a n y o w n in g o r C o m p a n y w ith 2 5 % A n y c o rp o ra tio n with c o n tro llin g 2 5 % or co ntrollin g 1 5 % or o w n e rsh ip . m ore o f s h a re s of m ore than o n e b ank. m ore of sh are s of m ore than o n e bank, a n y object, p u rp o se o r p o w e r of directly M IS S O U R I In co rp o ra tes d efini tion from Federal B a n k H o ld in g C o m p a n y Act o f 1 9 5 6 , a s o r in d irectly o r g a n iz in g, o w n in g , o r o p co ntrollin g election of m ajority o f directors TEN NESSEE am en d ed . of m ore than one OF e ra tin g b a n k s in g ro u p s o r chains. BANK C o m p a n y o w n in g o r RESERVE ARKAN SAS 1. Definition (M B H C s) b a n k , h o ld in g 1 5 % ST. o r m ore o f v otin g L O U IS sh are s o f m ore than o n e b a n k in trust for s h a re h o ld e rs' benefit. 2. R estrictions on ( a ) Be c o m in g a n MBHC U n la w fu l U nlaw fu l U n la w fu l N o p e rson (in d iv id ual o r c o m p a n y ) m a y U n la w fu l o w n o r control m ore tha n o n e -h a lf of the c a p ita l stock in tw o o r m ore b a n k s. (b ) A c q u isitio n s U n la w fu l fo r an U n la w fu l fo r a n y U n la w fu l fo r M B H C N o p e rson co n trollin g (Im p licitly H o ld in g co m p a n ie s N o restrictions o n a c M B H C to acq u ire a s sets o f a b a n k . M B H C to take actions to a cq u ire grea ter u n la w fu l.) tha n 5 % control o f a n y b a n k ; a ls o u n w h o se b a n k s u b s id ia rie s h a v e 1 3 p e r q u irin g b a n k s which resultin g in o w n in g or co ntrollin g greater m ore tha n o n e -h a lf of the c a p ita l stock in one b ank m ay a c five y e a rs o r more, than 5 % o f voting la w fu l fo r s u b s id ia r y to a cq u ire a sse ts of cent o r m ore o f total com m ercial b a n k d e posits in the state sh are s o f a n y bank. q u ire a n y stock in a n o th e r b a n k . (a fte r the b a n k s d e a n y b a nk. duct fo re ig n d e p o sits, certificates o f d e p osit of m ore than $ 1 0 0 , 0 0 0 , a n d outof-sta te c o rre sp o n d ent d e p o sits from their d e p o sits) m ay h a v e o p e ra te d for except that h o ld in g co m p a n ie s w h o se su b sid ia rie s ha ve 1 6 .5 percent o r m ore o f in d iv id u a l, p e r so n a l, a n d corp orate d e m a n d a n d s a v in g s d e p o sits m a y not a c q u ire a n y a d d itio n a l b a n k s. not a cq u ire a d d i tion a l b a n k s. (c) D e novo banks U n la w fu l (Im p licitly (Im p licitly u nlaw ful.) u n la w fu l.) (Im p licitly u n la w fu l.) P roh ib ite d until 1 9 8 0 , except in the fo u r la rge st counties. B H C m ergers U n la w fu l (Im p licitly u nlaw fu l.) O th e r O n l y c lo se ly related activities. (Im p licitly u n la w fu l.) 1974 Page 23 (e ) U n la w fu l JU LY (d )