• Email
  • Single Page
  • Print

Reagan and the Real America

The United States, in sum, is moving toward a structure of employment ever more dominated by jobs that are badly paid, unchanging, and unproductive. This most industrial of societies, this bourgeois El Dorado has in fact gone further toward a service and retail economy than any other of the largest industrial countries. Two “activities” in international industrial classifications, “wholesale and retail trade including restaurants and hotels” and “community, social, and personal services,” together correspond approximately to American trade and services activities (although they also include public employees and wholesale trade). The United States has by far the highest proportion of its employed population in these activities of any of the five major industrial countries: 51 percent in 1978, compared to 44 percent in Japan and 37 percent in Germany. In the race toward services13 the United States—how shameful, how distant from Republican hopes—had surpassed even Sweden, which still had only 48 percent of its employment in trade and community services.

It is in this morose environment that Mr. Reagan’s industrial Geist must blaze into life. Is such resurrection likely? There are, it seems to me, two successive great dangers ahead. The first is that the policy will not work, that the spirit will fizzle. The second is that the doomed pursuit of the Republican resurrection may lead to a disintegration of the employment-intensive industries, and thus to tremendously increased levels of unemployment, particularly among the poorest and least protected workers. Eventually, as we will see, this may bring about a new “industrial” boom. But it would be one that in virtually no way resembles Mr. Reagan’s essential heavy industry. And the long term in question would be extended, and vicious beyond any part of the postwar economic experience.

In order to explore the first danger, we need to look more closely at the economic consequences of the changing employment structure. Here we see that the movement is from high to low productivity sectors of the economy. For retail trade and services have the lowest productivity, as measured by conventional economic indicators, in the entire nonagricultural economy. In both output per hours worked, and output per person employed, the two sectors are well below the national average.

(It is important to be clear that this low productivity is not a consequence of the proportion of women workers in retail and services. The industrial sector with the highest productivity—finance, insurance, and real estate—is also among the most feminized in its labor force. The sector where productivity is growing fastest, communication, also has a higher proportion of women workers than the national average. Within manufacturing, electric and electronic equipment—the source of much innovation for the entire economy—has a higher than average proportion of women workers. Nor is there a strong correlation between output per hour worked and hourly earnings. Workers in finance are among the lowest paid in the economy.)

We might think of the private American economy as being composed of two sectors, sector I and sector II, where sector II consists of the fast-growing retail and service industries, and sector I of all other industries. Employment in II, as we have seen, accounts for 43 percent of total employment in 1979, up from 38 percent in 1969, and its advantage has been accelerating. But it accounts for less than 26 percent of total output, and this proportion has scarcely changed since 1969.14

Output per employed person in sector II was $9,853 in 1979 (measured in 1972 dollars); in sector I it was $21,433. The consequences of the shift are yet more dire because productivity in II is not only low but also grows slowly, or not at all. Output per employee in II was lower in 1979 than in 1973; in sector I it increased 1 percent per year over the same period.

The shift to sector II is likely, thus, to have affected overall rates of growth of output per employee. Consider, for example, a different kind of economic growth: H-growth, or high-industrial growth. Suppose that, under conditions of H-growth, all of the 10.8 million new private jobs created in 1973-1979 had gone to workers in I. And suppose also that output per worker had still changed at its actual rate in the two sectors. By 1979H, the total product of the economy would have been $1,302 billion, instead of $1,223 billion. Output per worker would have been $17,622 instead of $16,552. Instead of essentially no growth in output per employed person (0.03 percent per year), there would have been an increase of 1 percent per year.

We might consider a more likely alternative—N or normal growth, for an outcome in which the increase in employment from 1973-1979 was divided between sectors I and II in the same “normal” proportion as was the increase in employment from 1949 to 1969. Even under N-growth, output per worker in 1979N would be significantly higher, at $16,781, than the actual outcome. Real growth in output per employed person, over the period between 1973 and 1979N, would have been a modest but respectable 0.261 percent per year.15 These simple calculations are not, of course, comparable with studies of the effects of “structural shift” on productivity defined as output per hour worked. But they do suggest some economic consequences of the movement into sector II employment.16

The shift in employment is thus likely to be implicated in one of the most lamented problems of the recent American economy, namely the slow rate of growth of productivity. Output per hours worked in the private non-farm economy—to revert to the most widely used measure—which increased 2.2 percent per year from 1950 to 1973, increased only 0.5 percent per year from 1973 to 1979; the level of productivity was actually lower in 1979 than in 1978, and is lower in 1980 than in 1979.17

The tribulation of a slowing rate of productivity growth is not new. Mr. Reagan’s Republican predecessors were similarly preoccupied: compare Mr. Nixon’s “comprehensive national crusade” to increase productivity of 1972 with Mr. Reagan’s marginally less felicitous “great national crusade to make America great again.”18 But the United States is peculiarly afflicted in comparison with its international competitors. From 1973 to 1977, it had the lowest rate of growth of productivity of any of the five large industrial powers—lower than even Britain’s.19 Productivity growth rates have been decelerating in all the largest industrial countries; only in the United States has the level of productivity actually been falling.

The structural changes in employment and output also make it less likely that the strictly industrial part of the economy can bring about a new boom. For even within the sector we have called I, “industry”—defined as manufacturing, construction, and mining—is far from dominant. We might describe industry, in this sense, as sector IA, and a sector IB as being composed of all other I-sector activities: transportation, communication, utilities, wholesale trade and finance, insurance and real estate—what could be described as “industrial services.”

Sector IA, on this definition, accounts for only 48 percent of I-sector output (35 percent of total output in I and II) and 63 percent of I-sector employment. Sector IB has by far the highest output per employee of the three sectors: $30,400 in 1979. In sector IA output per employee is relatively low, at $16,300. Over the entire postwar period, productivity in this sector increased at about the same rate as productivity in IB. (Within IA, however, output per employee in manufacturing has increased 1.3 percent per year from 1973 to 1979, while output per employee in mining and construction fell substantially.)

Here again, international comparisons are interesting. Of the five major industrial countries the United States has the lowest proportion of its civilian employment (public and private) in the equivalent of sector IA; 30 percent in 1978, compared to 44 percent in West Germany, for example, or 38 percent in Britain. Once more, the United States is less industrial, in this sense, than Sweden. The positions are the same if we look at manufacturing alone: the United States has less than 23 percent of its total civilian employment in manufacturing, compared to 35 percent for Germany, 27 percent for France, and 25 percent for Sweden.20

How welcoming is this environment for Mr. Reagan’s economic policy, founded as it seems to be in the hopes of industry, of sector IA? The industrial sector is easily surpassed, in employment, by sector II; it is surpassed in total output and in output per employee in sector IB. Only the most prodigious expansion of industrial production, therefore, would suffice to invigorate the entire economy.

Consider, for example, the consequences of a more dynamic industrial performance in the 1970s. Let us double the 1973-1979 rate of growth of output per employee in sector IA—0.615 percent per year—keeping all else constant, and call this hypothetical dynamic economy D-growth. In this outcome, 1979D output per employee in IA would be $16,913 instead of the actual $16,275. But the total growth rate of output per employee would have increased only to 0.261 percent per year; just the same, in fact, as under our earlier hypothetical outcome (1979N) where employment in sector II grew at no more than its “normal” historic rate.

Or imagine the transformation that would have been required in IA in 1973-1979 to shift the private economy not to a paltry 0.2 percent annual growth rate in output per employed person but to its historic rate, for 1947-1973, of 2.055 percent per year. Such a shift, with conditions in the other two sectors held unchanged, would have required an increase in IA output per person employed to $22,297 in 1979S (outcome S for superdynamic). This would imply an annual growth rate of over 6 percent from 1973-1979S: more than twice the historic growth rate of output per employee in this sector, and ten times its actual growth rate.

A “supply side” enthusiast might answer, to these considerations, that the recent emphasis on traditional industry is no more than political rhetoric; that the Republicans once in office will care not at all whether their economic recovery comes about in sector IB, sector IA, or sector LXXXIV. Supplysiders—like their opponents and mirror-images, the “demand-siders” or Keynesians—are unconcerned with the real industries and workers and technologies that constitute an economy. They prefer to live among the abstractions of economic theory and economic politics, whether of “supply” and “expectations” or of “effective demand.”

But the primacy of traditional industry, of IA, is not easily ignored. If the economy were to expand in sector IB, employment would be unlikely to increase fast. And the characteristic behavior of IB corporations—immense semi-public utilities, bus companies and truckers, financial corporations, AT&T—is hardly that of the sensitive, even skittish entrepreneurs, the investors in new factories and equipment to whom the Republican policy is addressed. Yet if growth of output and employment continues to expand in sector II, then the present problems of inflation and low productivity growth will be compounded.

  1. 13

    OECD, Labour Force Statistics, 1967-1978 (Paris, 1980). By “services” we understand, for the United States, activities within the US industrial classification “services”: hotels; personal services; business and repair services; amusements; health, educational, and other professional services. This definition is, of course, much narrower than one which includes under “services” all activities which do not produce goods.

  2. 14

    Here and in what follows employment figures are taken from Employment and Earnings, as above. Output figures for the private nonagricultural economy, in constant 1972 dollars, are from the Department of Commerce’s series “Gross National Product by Industry in Constant Dollars,” Survey of Current Business, May 1980 and July 1979, Table 6.2; and The National Income and Product Accounts of the United States, 1929-1974, Statistical Tables (Department of Commerce, Washington, D.C., 1977), Table 6.2. The national income and product accounts are in the process of being revised; published figures are used here. It is worth noting that the measurement of productivity in service activities poses very serious problems, and that the derivation of price deflators for different industries also requires stalwart assumptions.

  3. 15

    Our illustrations, and the ones that follow, are not of course intended to be realistic. Similar calculations using total hours worked instead of number of employees, although no more realistic, would yield broadly comparable results. Thus output per hour worked in a hypothetical 1978N, where hours worked increased from 1973 to 1978 in the same “normal” sectoral proportion as in 1949-1969, would yield an increase in output per hour worked of 1.278 percent per year from 1973 to 1978N, as compared to the actual 1.120 percent. Figures for hours worked are from Survey of Current Business, July 1979, Table 6.11, and The National Income and Product Accounts, op. cit., Table 6.11.

  4. 16

    Several well-known studies minimize “shift” effects. Thus Edward Denison (“Explanations of Declining Productivity Growth,” in Survey of Current Business, August 1979) suggests that the shift in hours worked to services had “trivial” consequences in 1966-1976; he concludes that most of the decline in productivity in 1973-1976 must be attributed to a negative “residual” to be loosely identified, presumably, with the regress of knowledge. An OECD study of seven countries, using more recent data, also concludes that sectoral shifts in employment had little effect on growth of output per employed person. (“Sectoral Shifts and Productivity Growth,” in OECD, Economic Outlook, Paris, July 1979, pp. 28-35.) But this study groups services into a compendious sector of “other activities” which also includes such high productivity and high productivity growth industries as communications and finance, insurance and real estate. For a more convincing view, see Lester Thurow, “The Productivity Problem,” in Technology Review, November/December 1980.

  5. 17

    US Department of Labor, Monthly Labor Review, August 1980, Tables 31 and 33.

  6. 18

    The New York Times, July 16, 1980.

  7. 19

    OECD, Economic Outlook, July 1979, p. 29.

  8. 20

    OECD, Labour Force Statistics, 1967-1978 (Paris, 1980).

  • Email
  • Single Page
  • Print