Knowledge that Transforms

EITTLE is known about the real output, productivity, and price trends in the construction Nevertheless, there exists an undercurrent of feeling that the performance of this giant industry has lagged considerably behind the American economy as a whole. Economists have tended to shy away from this area of research because the available statistics have been viewed as inappropriate or unreliable. Consequently, the field has been pre-empted, more or less, by the popular press which has propagated a most distinct notion. The industry, in the words of one writer, is noted for its incredible inefficiency (particularly home building). As an explanation of . why new homes cost too much, he stated that . home builders are forced to use wasteful and obsolete techniques and materials, do little or no research [and] are hobbled by a maze of outdated building codes.' To the extent that such an opinion is held within the economics profession it usually can be traced to the Department of Commerce's construction composite cost index.2 Since World War II the composite has risen faster than most other well known price indexes and the implication has been drawn that much of the rise has been due to inefficiency. However, the composite is essentially a simple inputcost index 3 and no inference about productivity can be logically deduced from it. To the knowledge of this writer, only a few economists have published any research on real output, productivity, or prices in construction. Most of these have been interested in construction only incidentally, as one sector to be integrated into the whole. Their estimates relating to construction can be summarized as follows: Schultze found that construction prices rose 53.7 per cent over the decade beginning in 1947.4 For the same period, the composite rose 47.5 per cent. Alterman and Jacobs estimated that the average annual change in real product between 1947 and 1955 was 5.6 per cent and 2.5 per cent came from an increase in real product per man-hour.5 This figure is surprisingly large when contrasted with the popular view. Yet it appears from their work that construction productivity lagged considerably behind the average for the economy and even behind total services. Kendrick deflated value added by an output deflator closely akin to the composite to yield an index of real output and estimated that real output per man-hour rose at an average rate of 3.9 per cent between 1948 and 1953,6 but productivity had not advanced between 1909 and 1948. Haber and Levinson, writing in the early fifties, thought that productivity had been increasing by about 1.5 per cent per year in residential construction and two per cent in industrial and commercial building.7 They attributed this low rate to the . economic characteristics of the industry. The most recent estimate is that of Domar, et al.8

O VER the past few decades measures of market, or industry, concentration have acquired respectable, if not completely untarnished, status as useful indicators of monopoly power. The publication in recent years of papers emphasizing the importance to the monopoly problem of changes in market shares of individual firms, however, can in many respects be viewed as challenges to the status of the static measures.' More important, these dynamic measures of market structure hold out the hope that, when analytically incorporated with static measures, they may prove to be of great help in identifying and understanding monopoly power. This paper presents a large body of data on both static and dynamic measures that could provide a basis for increasing this understanding. But first a brief analysis suggests how these two types of measures may be useful when inferring the presence of monopoly power in a particular industry.

DISCUSSIONS of municipal financial difficulties nearly always involve fundamental problems of local public investment requirements. It is curious, therefore, that studies of the determinants of such outlays have been so scant. One probable reason in the United States is that the relative financial responsibility for local investment of municipal, county, state, and federal governments may vary considerably from region to region and even among communities within any given region. This complexity of government structure, combined with a diversity of local accounting procedures, has probably hampered collection of adequate data. In general, studies of municipal expenditures have tended to combine current and capital outlays into a single variable.' As Brazer has pointed out, this procedure obscures much that should be revealed, since the forces which influence investment outlays are different from those affecting current expenditures.2 The present study develops a substantial number of hypotheses, covering four principal groups of causal factors affecting community investment. These center on demographic, housing, commercial, and industrial characteristics, the influences of which are then examined in the light of empirical evidence from East Flanders, a Belgian province which serves as an excellent laboratory for these purposes. Belgium is one of the world's most densely populated and highly industrialized nations. Although its total area is relatively small, the number of municipalities per unit area is very high. East Flanders, for example, covers 1150 square miles and contains 297 municipalities. This phenomenon results in part from the fact that Belgium has no equivalent to the American county; any given location is always within some municipality. All municipalities in Belgium participate in a common system of investment finance, making it possible to obtain highly standardized data for local investment. Consequently, variation in such outlays in Belgium is a more accurate reflection of differences in actual requirements than would be the case in the United States, where differential ability to pay muddies the analytic waters.3 A further element of homogeneity in the present case is the dominance of the clothing and textile sectors in all industrial municipalities. This has permitted fairly straightforward identification of industrial importance on the basis of employment data, without the complications that could arise from qualitative differences in industrial structure. The empirical evidence considered below consists of data obtained in a special municipal investment survey prepared by the author and completed by competent local authorities over a six-month period. The data include all investments in responding municipalities during the five-year period 1956-1960. The 269 respondents account for over 95 per cent of the population of the province.

IN recent years there has been a good deal of discussion concerning the relationships among market structure, research and development, and the rate of technical change. Much of this discussion has focussed on the question of whether large firm size is a necessary condition before firms will engage in research, and whether research and development (R and D) is likely to grow more or less than in proportion to increases in firm size. A further set of questions deals with the relationship between research and the rate of technical change experienced by the firm. Can variation in the latter be explained largely by differences among firms in the size and character of their research programs? Are economies of scale in R and D likely to be present? What is the effect of firm size on the productivity of a research establishment? This paper provides an empirical analysis, concerned with these questions, of the experience of the United States pharmaceutical industry during the period between 1955 and 1960.

T NHE structure of production of an economic system, represented by the matrix of input-output coefficients, has traditionally been held to be determined by technology. Thus the coefficients are sometimes called coefficients. If this is, in fact, the case one should expect to discover a productive structure which is common to all economic systems having a like technology.' For example, one should expect to find certain characteristics of the input-output matrices of all industrialized countries, which have a technological origin. Previous studies have compared the productive structure of different economic systems in purely economic, or even arithmetic terms, so that they may fairly be described as taxonomic. On the other hand, this paper suggests that there are certain fundamental elements which may be found in the productive structure of modern economic systems which are purely technical in character. It is demonstrated that the economic systems of Japan and the United States, although superficially dissimilar, contain almost identical patterns of industries which are strongly interrelated. This pattern, or framework, of productive relations has several interesting properties, which are found to be shared by the pattern of interindustry relations in other economies. The theoretical implications of this discovery are briefly discussed.

IF economists could measure the need for reserves, they might be able to agree on the right way to reform the international monetary system. Most of the economists who propose drastic reform do so because they anticipate a shortage of reserves; some even believe that the shortage is upon us. Those who advocate more gradual change believe that reserves are adequate now and for the next several years; some even believe that reserves are excessive. Unfortunately, there is no way to measure the adequacy of reservesnot even to make historical comparisons. Scitovsky's comments illustrate several of the problems involved in appraising the global stock of reserves:

ENVIIRONMENTAL health services, especially refuse collection and sewage disposal, though relatively small in terms of outlays compared to other urban government services, have played a major role in the rapid urbanization of society. They have helped to eliminate the frightful dangers to health that can threaten densely populated communities and decimate their populations. Economists who are concerned with planning the provision and financing of urban government services might find the analysis of refuse collection useful from a methodological point of view. Their interests tend to focus on issues of scale economies, break-even points, the cost of differential service levels, the determination of user charges, etc. In addition, projections of service needs and costs occupy their attention. Inquiries into these matters require information on government service demand and cost functions which are unusually difficult to determine because of the economist's inability to find readily quantifiable output units and quality characteristics. However, as will be shown below, refuse collection cost analysis, in this respect, poses less difficulty than any other local government service, except perhaps water supply. In much of the world, refuse collection services are mainly provided by governments, although this need not be the only method.' Indubitably, these services do not constitute pure public goods; only to a limited extent do they fulfill either social or merit wants.2 Or, to put it differently, while pure public goods have very many major indirect or neighborhood effects, in the case of refuse collection such effects are relatively minor. Residential refuse collection (and disposal) service involves the regularly scheduled pickup, collection and disposition by a sanitary agency of solid waste materials resulting from the operation and maintenance of households; its main components are garbage, ashes and rubbish.3 It is useful to distinguish between residential refuse collection on the one hand and industrial refuse collection on the other. The two usually are carried out separately and the latter takes place virtually on a daily basis. Unlike residences, commercial and industrial establishments vary greatly in the amount of refuse they produce, and many large industrial firms provide their own collection and sometimes even disposal service. In the following pages, an ideal residential refuse collection cost model will be presented. Because of data difficulties and the marginal importance of some variables, this model will next be simplified and implemented in relation to a case study. Finally some conclusions are pre-

DURING World War II, shipbuilding output per man-hour evidenced a remarkable improvement. Between December 1941 and December 1944, man-hours required to produce a Liberty vessel, an emergency type developed for mass production, fell from an index value of 100 (December 1941 100) to 45, an increase in output per man-hour of 122 per cent, or an average annual increase of about 40 per cent.' Compared to the long-run improvement in man-hour productivity elsewhere in the United States economy the shipbuilding experience is impressive. Over the period 1909 to 1961, annual growth in output per man-hour, averaged over the entire economy, was 2.4 per cent.2 Even during a subperiod of rapid improvement, 1919-1929, and in a sector where improvement was most conspicuous, manufacturing, the average annual rate was only 5.6 per cent. In light of these figures, an explanation of the World War II shipbuilding productivity experience would be desirable, not only for its own sake but because it would shed light on secular forces constantly operating in the United States economy. An explanation is provided in this note with particular emphasis placed on the role of organizational and individual learning resulting from accumulated production experience.3 The extreme and unusual shipbuilding conditions that prevailed during World War II were conducive to extensive learning. New products were adopted, mass production techniques were introduced, labor inexperienced at shipbuilding was recruited, and the direction of production was placed in the hands of eminently successful managers who were, however, inexperienced at shipbuilding. This combination of unique events represents an extreme instance, which will facilitate the isolation of learning or adaptation effects from other effects, if learning is quantitatively important.