actually got ahead of their larger rivals. But even when they do not succeed in taking business from the large firms the small firms provide incentives to the large ones by trying to increase their market shares. It is important to note, however, that fringe competitors contribute to progress in that way only when they are strong competitors. They will not be effective unless their scale of operations is large enough for reasonable efficiency and unless their financial resources are adequate to permit them to take some risks.

Subject to that qualification then small firms play an important competitive role even in industries dominated by much bigger ones. In addition, of course, there are many industries in which most of the output is produced by small firms who provide competitive incentives for one another.

RESEARCH AND INNOVATION

Our economy's capacity to produce goods and services at any one moment depends on the amount of resources and the technical knowledge available and the efficiency with which those resources and techniques are utilized. But the rate of increase in our capacity to produce depends on the rate at which we accumulate resources and on the rate of development and application of new techniques of production. The rate at which new techniques are developed and applied depends on many factors but it can be influenced to an important degree by the size distribution of business firms.

In this section we shall consider the influence of firm size on the ability of firms to undertake research activities and to apply the results in practice.

It appears to be the case that large firms are in a distinctly better position to support systematic research activities than small firms. That is true for a number of reasons.

The strongest of these reasons is the effect of size on the pooling of risks. There are many calls on the financial resources of firms to justify the allocation of additional funds to research, and the management must suppose that the prospective returns from additional research will be greater than those available from other types of investment. As the proportion of resources devoted to research increases, the return required to justify additional research expenditure will increase. Even if there were no risks, small firms would not find it profitable to spend a much larger proportion of investable resources on research than large firms. If a large firm and a small firm both spend the same proportion of resources on research, the large firm can support more projects than the small firm. Since results of research are very uncertain, the small firm with a small absolute research budget takes a greater risk about the outcome of research than a large firm with a large budget. When a sufficient number of projects can be undertaken some are bound to succeed and pay off enough to compensate for the failures. When only a few projects are underway there is a chance that the firm will get nothing for its efforts. Of course, some firms with small budgets will be successful and obtain a very high return on their investment, since the success is not affected by failures in other projects. But most firms have an aversion to making "long shot" investments. As a result, small firms are likely to hold down their research budgets because investment in research is too risky.

23855-58--14

In addition, most small firms are in a riskier position than large firms on other grounds. In general they have lower profits and less access to outside funds. Moreover, they are more dependent on demand in particular localities or for a small number of products than large firms operating over a wider range of locations and products. The general risk position of small firms makes them still less able to gamble on research.

A second factor working against small firms is the existence of economies of scale in research itself. Laboratories require complex equipment which will not be fully utilized unless large numbers of research workers share its use. In addition, there is an interaction between research workers in different fields or on different projects which cannot be exploited in a small laboratory. This is not to suggest, of course, that much useful work cannot be done in small laboratories, but there are some reasons for thinking that large laboratories have an advantage over small ones.

These disadvantages can be overcome to a considerable extent by contracting out research to universities or to firms specializing in research work.

A third advantage for large firms arises from the diffuse nature of the benefits of research. The outcome of scientific research is, in the nature of the case, unpredictable. Investigations aimed at the improvement of a particular product or process may produce results with applications to quite different areas. A large firm with a diversified line of products can view that situation with equanimity. It has a good chance of finding an application for whatever turns up. A small one can use only a small part of the knowledge gained from research.

This disadvantage is, of course, of much greater importance in basic research than in enginering work with a narrow focus.

The advantages of large firms in research activities are reflected in the results of the 1953-54 study of industrial research and development activities. The survey showed that the percentage of companies conducting research and development (RD) programs rose steadily with company size. Only 8 percent of manufacturing firms with less than 100 employees had them while 94 percent of those with over 5,000 employees had RD programs. Manufacturing companies with over 5.000 employees paid for 66 percent of RD work but employed only 40 percent of the workers in manufacturing. In contrast firms with less than 500 employees accounted for only 14 percent of the RD cost thought they employed 35 percent of all workers in manufacturing. Government financed research is even more highly concentrated among the large firms.)

It is clear enough that large firms have a distinct advantage in carrying out industrial research and development. But that is not the whole story. Industry does relatively little basic research. To a large extent industrial research exploits scientific principles developed elsewhere. The existing stock of scientific knowledge at any

ing an enormous reservoir of potential industrial

le research organization works on a limited Ses-the chosen ones depending on the inter

Belence and Engineering in American Industry (U. S.

ests and vision of the firm's management and research directors. Certain possibilities wil be ruled out by a particular firm as impractical, outside the range of the firm's interests, as having insufficient potential markets. In pursuing a given line of research and development, any one individual or any small group is likely to make very serious errors in judgment as to the practicality or market potential of a given research proposal. Large research organizations may be more efficient than small ones. But from another point of view, a good deal may be lost by having a heavy concentration of research and development activity in the hands of a relatively small number of firms. (Ốf course, it is possible for a large firm to decentralize its research and development activities and allow a great deal of independence to research directors, and even to encourage internal competition, but that is not always done.)

The research activities of small and medium-size firms may make an important contribution to our total resources by insuring that the development possibilities overlooked or neglected by larger organizations are tried out and exploited.

Finally it is important to note that much progress in industry takes place without formal research. Many useful developments in technology are worked out by production personnel with relatively little formal research expenditure. Similarly new applications of existing knowledge may sometimes be made without much formal research. In some cases foreign patents may be applied in this country with relatively little further work. Finally minor modifications in products may make them cheaper and open out new market possibilities.

In all these cases the very large firms have relatively little advantage over small or medium-size firms provided that resources are adequate for the necessary investment.

To sum up then, small firms can and do make significant contributions to industrial research activities even though large firms account for a very high proportion of industrial research and development activities.

Their contribution takes the form of (1) exploiting some possibilities neglected by larger research organizations and (2) of a large variety of technological developments in production methods and product design which are carried out without the use of a formal research organization.

Finally, it should be noted that in recent years a number of relatively small firms have appeared where principal activity is research and testing for the account of others.

SMALL BUSINESS AND PRICE STABILITY

We have already seen that small business has a significant influence on the relations among prices, costs, and profits. Now we must turn our attention from the level of prices in relation to costs to the movement of prices and costs over time. The effects of small business on price movements are significant and complex. In some ways small business tends to intensify the inflationary tendencies in our economy, in others to weaken them.

In analyzing the influence of small business on price stability, it will be useful to distinguish between the role played by small busi

ness in industries in which relatively small firms predominate (e. g., textiles and apparel) and in industries in which most of the output is produced by a relatively small number of large firms.

We can compare the movements of prices in industries with many small firms with price movements in highly concentrated industries in which small fringe competitors remain a factor.

In addition it is necessary to draw a distinction between price movements induced by changes in aggregate demand (relative to industrial capacity and labor supply) and those induced by increases in wage costs resulting from trade union action and other forces affecting costs. Such a distinction is somewhat artificial. The extent of the influence of trade unions on wages is influenced by the level of unemployment and capacity utilization in the economy. We cannot ordinarily say whether any particular price increase is due to "cost push" or "demand pull"; in most cases we have to suppose that actual price increases are due to the interaction of both factors. Nonetheless, the two factors are distinct from one another, and different sectors of the economy react differently to the movements of demand and the pressures on wages exerted by trade unions.

Let us first compare the effect of changes in demand on price movements in industries with many small firms with the effect of demand changes on prices in highly concentrated industries. It seems fairly safe to say that prices of goods produced in industries in which many small firms compete in the same market, e. g., textiles and apparel, are more volatile than the prices of goods in more highly concentrated industries.

At any one moment the producers in a competitive industry are producing as much as is profitable, in view of their costs at the existing price. If demand remains unchanged, any producer can sell more by increasing output and shading his price a relatively small amount. When an increase in demand occurs, price will be marked up because no additional output is available at the existing price. Price will have to rise until demand and supply are once again in balance (either through contraction of sales as a result of the price increase or because the higher price induces firms to increase output by working overtime or using obsolete plant).

When demand declines, prices will move downward, but initially, most firms will find that they are selling less than they are willing to supply. Further shading of prices has to continue until supply and demand are once more balanced by increases in sales or reduction in output.

When there are many small firms there is no reason for any one to refrain from price cutting as long as there is excess supply, since competitors may do so anyway; if they are not, the firm cutting the price would take a little business from each of a number of firms without influencing the action of any of them.

In such industries prices will move upward and downward with short-run changes in demand. The range of movement will be suffi cient to keep supply and demand in balance. Over longer periods of course, supply can be increased or decreased by the construction of additional plant or the abandonment of old plants. The adjustment of capacity places a limit on the extent to which price can deviate

from average cost in efficiently operated new plants, except for short periods.10

Prices of the products of highly concentrated industries are much less responsive to short-run movements in demand than prices in highly competitive industries. When demand declines managements tend to avoid open price cutting because each firm expects that others would retaliate if it cut prices. On the other side, when demand increases, large firms often refrain from raising prices even though they could sell more than they can produce at the existing price.

It is easy to see why large firms tend not to cut prices when sales are low. Unless sales are very responsive to price, it will be unprofitable for any firm to seek more volume by cutting prices when its rivals are fairly sure to retaliate. In order to insure price stability when sales are declining it is necessary to refrain from raising them when sales are rising. In addition, large firms are usually sensitive about their share of the market and are therefore unwilling to risk the loss of customer goodwill by frequent price changes. At least some large firms are sensitive to customer relations and political considerations. Finally, if prices are raised during periods of high demand, unions may attempt to capture some of the increased profits in the form of a wage increase which cannot be reversed if prices should fall.

When there are only a few large firms in an industry, the relation between prices and direct operating costs tends to be very unresponsive to changes in the relation between demand and capacity. However, when an industry is fairly heavily concentrated, but contains a certain number of small firms which together account for an appreciable proportion of output, the picture may be somewhat different. When demand declines, the smaller firms may lead the way in cutting prices (either openly or more commonly through various types of unannounced discounts and other concessions to buyers). They tend to do so partly because each small firm can gain a substantial percentage increase in volume by taking a small percentage of sales from the larger ones if they can cut prices for a time without retaliation. The small firms may lead prices down, eventually forcing the large ones to follow them. To the extent that prices decline when demand declines in the industry, they will also tend to rise when demand increases. In general, then, it can be said that as the proportion of an industry's output produced by small firms increases, the tendency for prices to vary with demand will also increase. That proposition is, of course, only a broad and loose generalization. Price movements are influenced by many other aspects of industrial structure besides the size distribution of the firms in the industry.

If demand moves upward and downward (relative to capacity) prices (in relation to direct costs) will tend to move upward and downward as well. The variation in both directions will tend to be larger in industries in which a large share of output is produced by small firms than in those in which most of the output is produced by a few large ones.

10 It should be noted that in localized industries like trade and service and some branches of manufacturing only a few firms may compete in any one market even though most of the firms are small.