to leave the labor force when they become unemployed since they are secondary earners in the family. Generally, married men of working age stay in the labor force unless they are unable to work.

Occupations. The occupations in which Negro and white women tend to work reflect to some extent the lower educational attainment and lack of training of the Negro women as well as discriminatory hiring practices. Half the Negro married women were in service occupations in March 1967, about evenly divided between private household workers and other service workers. Among the white wives, 16 percent were in service occupations, with the proportion in private household work minimal. A higher proportion of white than Negro wives were in professional occupations. Relatively twice as many white as Negro wives were employed in clerical occupations. A significant portent for the future for both white and Negro women is the higher proportion of single women in clerical occupations relative to the married, indicating that, as they marry, the occupational distribution of wives will be affected accordingly.

The Shape of the Future

Women's labor force participation is expected to continue to increase in coming decades. The female labor force has been projected for 1980 at about 32 million, more than half again as large as it was in 1960. Much of the growth is expected to

be among married women, through some rise in labor force rates as well as population increase. Their economic role in the family has already achieved importance, despite the low level of their earnings relative to men's. Women should benefit from the provisions of Title VII of the Civil Rights Act of 1964, which, among other provisions, prohibits discrimination because of sex in hiring, upgrading, and all other areas of employment. However, whether women enter the upper reaches of professional and other occupations to a greater degree than now depends on factors such as the acceptance of women in such work by society, the opportunity and willingness to take work of increased responsibility, the desire for requisite education and training, commensurate pay as a return for the career investment, and adequate facilities for child day care. Experience in some countries suggests that all of these are possible of achievement in the United States.

The network of demographic and socioeconomic factors which influence women's labor force participation makes it difficult to predict the kind of life pattern which American women may be in the process of shaping for their future counterparts. The American experience may continue in an economic and cultural climate in which labor force participation for married women is, to a large extent, a matter of choice, based on individual aspirations, values, and preferences. Or it may culminate in an economically, culturally, or publicly imposed compulsion that women, like men, must work, as has already happened in some countries.

Air Transport: Trends in Output Per Employee

JOSEPH E. DRAGONETTE

AND CHESTER MYSLICKI*

OUTPUT PER EMPLOYEE INDEXES for the air transportation industry have increased at a rate of almost 8 percent a year since 1947, the highest rate for any major industry. As a result of this rate of gain, productivity levels in 1966 were five times greater than in 1947. By contrast, output per worker in the total private economy during the same period increased at an annual average rate of only 2.8 percent. Despite the high productivity gains, employment has advanced steadily, since demand for air transportation services has grown even more rapidly. (See chart.) The growth rate was not maintained consistently during the postwar period and reflected generally very high increases (averaging 11 percent a year) for the late 1940's and early 1950's, a slackening of the rate to about 4 percent from 1955 to 1961, and a return to the 11-percent rate since 1961.1

The 1947-55 period, despite a slow start in 1947, evidenced extraordinarily large increases in output through 1955, averaging almost 17 percent a year. The industry was expanding rapidly into new service areas: coach service was introduced, local service carriers were certificated, and allcargo carriers were established. Since productivity gains were fairly high, except in 1952, changes in employment ranged widely from actual decreases at the beginning of the period to a 12-percent increase in 1952. For the entire period, however, as output more than tripled, employment rose by only one-third.

The significantly lower rates of productivity growth during 1955-61 were associated with vary

ing patterns of output and employment changes. Between 1955-57, annual output increases continued at the high 1947-55 rate, but employment gains of about 11 percent a year nearly matched output increases. During 1957-61, output gains fell off sharply to less than half of those in the 1947-57 period, while employment increases averaged slightly more than 3 percent a year. The most revolutionary technological change in 'the air transportation industry, which contributed to subsequent increases in productivity, was introduced during this period of lowest productivity gains. New turbojets had three times the capacity and were twice the cost of the DC-7's introduced in 1953.

Between 1961-66, output increased at a rate of over 17 percent a year with some acceleration evident. As productivity averaged about 11 percent a year, employment has been accelerating fairly rapidly from a 1.6-percent increase in 1962 to 9.4 percent in 1966.

Group Measures

Ordinarily, productivity reports for individual industries are limited to analysis of total output and input measures as data for individual companies or establishments are considered confidential. For the air transportation industry, however, basic source data for each airline are published. In order to provide background material output per employee measures for individual airlines have been calculated to provide measures of dispersions and to build up totals for carrier groups.

*Of the Office of Productivity, Technology, and Growth, Bureau of Labor Statistics.

1 BLS indexes of output per man-hour and per employee for the private economy and major sectors use the concept of net output, while the industry measures are based on final gross output. Lack of man-hour data in air transportation restricts calculations. Productivity indexes for the air-transportation industry are developed in Indexes of Output Per Employee, Air Transportation Industry, 1947-64 (BLS Report 308, 1966). The report includes a technical description of the procedures. The indexes cover SIC industry 4511: Air Transportation, Certified Carriers.

a Lack of comparable data for individual airlines prior to 1957 limited analysis to the post-1956 period.

3 Indexes for individual airlines are derived from the same data and use the same methods as the industry indexes. We have not identified the individual companies, primarily because the individual measures do not have the same degree of reliability as the overall measures. Our estimating techniques, which can provide relatively unbiased results for the industry, may be less valid for individual airlines. This is especially true for estimates of employment, strike adjustments, quality changes, and the omission of items such as excess baggage and subsidies.

285-796 O-68-2

The measures for individual airlines ranged from increases of over 20 percent a year to an average decline of 1 percent. These extreme rates were registered by small airlines. The extreme skewness of the size distribution-five of the airlines constitute about 65 percent of total employment-suggests the need for groupings that are more homogeneous in terms of services and size.

The average rate of productivity increase for the air transportation industry is a function of the changes in productivity for individual airlines and the effects of shifts in the relative importance of companies with different productivity levels. For the 1957-66 period, a test of shift effect was made by comparing the usual productivity measure with a measure calculated by combining individual rates for airlines with 1957 employment weights. The effect of shifts was found to be negligible, as the rate of gain from the measure

excluding the effects of shifts differed by only 0.1 percentage point from the industry rate of 8.2 percent.

The Civil Aeronautics Board has classified the certificated airlines currently operating into eight groups or subgroups. (See table 1 for listings of the airlines.) These groupings, based on company data, often reflect a considerable mix in type of operations and thus place some limitations on analysis. For example, some of the domestic airlines also engage in international operations; these operations are included in the output and input measures of the domestic group and are excluded

from the data for the international territorial group. An analysis of some of these groups is presented below.

Big Four System. This group covers the four largest domestic trunk systems and contains over

50 percent of the employment. Output per employee (table 2) rose 6.0 percent from 1957 to 1966 which is significantly below the 8.2 percent average for the entire industry and thus even much lower than the remaining sectors of the industry. The deviation seems to be due almost entirely to the extremely low rate of growth for the 1957-61 years. The individual airlines rate of productivity increase varied substantially from 3.1 to 8.2 with the coefficient of variation exceeding 30 percent.* Output increases for the Big Four averaged 10.8 percent, also below the industry average of 12.4 percent. Output increases lagged behind the industry rate largely because the Big Four system had below average participation rates in the most favorable growth areas, such as international travel, freight, helicopter service, and local feeder lines. The output range for individual airlines ex

The coefficient of variation measures the extent to which indicated changes for the individual airlines are clustered about a mean percentage change.

2 Average employment for 1966 reflects time lost due to strikes.

tended from 6.6 to 13.3 percent. Employment increases of 4.5 percent a year exceeded the industry average but individual company rates ranged from 2.1 to 7.5 percent.

Other Domestic Trunks. These cover seven airlines of generally moderate size with fairly widespread area operations. Employment accounts for about 20 percent of total employment. Output per employee increased at the rate of 10.7 percent. Five of the airlines exceeded the industry average of 8.2 percent, one nearly matched it, and one fell slightly below. The variability within this group was less than that for the Big Four with coefficient of variation of 26 percent. Output also expanded (16.5 percent) at a rate above the industry average of 12.4 percent, with an individual high of 26.4 percent and a low of 10.7 percent. The employment change of 5.2 percent was accompanied by extremely wide patterns of change for individual airlines, extending from 1.4 percent to 11.3 percent.

Local Service Carriers. Thirteen relatively small companies are included. They tend to operate in limited regions only; they make up almost 9 percent of total employment. While output increases exceeded 18 percent per year, output per employee rose only 8.1 percent or just under the industry rate. The small scale operations of the local service carriers made it impractical to purchase the new large-size jets that the bigger airlines could afford, thus limiting output-per-employee increases.

The associated employment rate change of 9.5 percent was more than twice the industry increase and boosted the relative employment share from about 6 percent in 1957 to about 9 percent in 1966. The absolute range for average annual output, employment, and productivity rates of individual airlines within this group tended to exceed ranges for other groups simply because of the larger number of airlines. However, the coefficient of variation, which is a superior measure of dispersion, was generally lower for the local service group for both output and employment rates, and did not depart significantly from the measures calculated for dispersion of productivity rates.

put, and employment, reflecting lack of homogeneity within this grouping.

All-Cargo Carriers. This consists of three small airlines with about 2 percent of total employment. The output per employee rate (14.4 percent) was the highest of the five major groups and the output increase (16.6 percent), the second highest. Employment at 1.9 percent a year rose the least.

Technology and Outlook

Technological changes have been closely associated with a series of extensive changes in new types of equipment. The introduction of the turbojet in 1958 and its rapid spread is the latest and most revolutionary of four major reequipment programs beginning in 1945. The full effect of these changes tends to lag behind the introduction period because time is required for adoption and effective utilization throughout the industry. Improvements in traffic handling facilities, many of an evolutionary character, have also been important. Thus, installation of automatic reservation systems, improved cargo facilities, better airports, improved air traffic control systems, and maintenance facilities were vital factors.

In terms of purchase of new equipment, the immediate outlook is for the acquisition of shortand medium-range passenger craft and convertible passenger-cargo planes. Another round of new technology, involving both subsonic and supersonic transports and vertical or short takeoff and landing craft, is possible by 1975. Continued progress in the traffic handling facilities is necessary for maintenance of efficient service.

Output is expected to increase at a high rate with aircargo being the likely area for greatest expansion. Freight and express revenue ton-miles were only 11 percent of total revenue ton-miles in 1947 but had expanded to 24.5 percent in 1966. Industry estimates are that passenger traffic will likely grow from 7 to 10 percent annually through 1970 while aircargo may increase at 20 percent a year. The rising demand for air transportation services may more than offset expected productivity increases and thus call for continued increases in employment.