YOUTH UNEMPLOYMENT

TABLE 4

Unemployment Elasticity Coefficients with Respect to Changes
in Real Minimum Wages

Teenage Labor Force
Classification

1. All Teens

2. White Teens

3. Non-White Teens

(3)

Model 1
(1) (2)
Jan '54 Jan '54– Jan '54—
Sept '61 Sept '63 Dec '65

Model 2

127

(1) (2) (3) Jan '54– Jan '54– Jan '54 Sept '61 Sept '63 Dec '65

We feel that the results of our study are unambiguous. Increases in the federal minimum wage cause unemployment among teenagers; the effects tend to persist for considerable periods of time and seem to be strengthening as coverage is increased and as enforcement becomes more rigorous. To the extent that a minimum wage creates unemployment among minority group teenagers, it inevitably will exacerbate the social unrest in Urban Ghettos. With the preceding discussion in mind, the authors suggest that serious consideration be given to reducing, or removing, the federal minimum wage for teenagers.

Reprinted from the Proceedings of the 24th Annual Meeting of the INDUSTRIAL RELATIONS RESEARCH ASSOCIATION, Pages: 38-46

The Lag in Effect of Minimum Wages
on Teenage Unemployment

DOUGLAS K. ADIE
Ohio University

Although the existence of the effect of minimum wages on teenage unemployment has been examined by numerous researchers, only recently has there been an attempt to measure the effect.1 The measurement of the lag in the effect has received even less attention.2 Since the effect of minimum wages on teenage unemployment is an important concern of policy-makers, an investigation of the lag structure is in order.

This paper examines the lag structure of the impact of real federal minimum wages on various categories of teenage unemployment rates using three techniques: namely, an autoregressive model, distributed lags and cross-spectral analysis.

Data

Seven classifications of monthly seasonally adjusted teenage unemployment rates supplied by the Bureau of Labor Statistics are used for the period January 1954-December 1970. The classifications are for the following teenage groups, aged 16-19: (1) all teens, (2) all white teens, (3) all non-white teens, (4) male white teens, (5) female white teens, (6) male non-white teens, and (7) female non-white teens. The symbol U, is used to designate the ith classification of teenage unemployment where i=1 7. The monthly seasonally adjusted unemployment rate for all civilian workers, also obtained from the Bureau of Labor Statistics, is designated U. The statutory minimum wage for covered workers, Wm, contained in Youth Unemployment and Minimum Wages (Bulletin 1657, G.L.S., p. 11) is expressed for computational purposes as a monthly series with changes occurring

* The author is indebted to the Earhart Foundation, the Institute for Humane Studies. and the Fred C. Koch Foundation for financial support. Bill Hurley, Joan Presley and John Raisian had a hand in the computations; Marie Devol typed the paper, while I remain solely responsible for errors.

1A general survey of the literature can be found in Youth Unemployment and Minimum Wages, Bulletin 1657, U.S. Department of Labor, Bureau of Statistics, 1970, pp. 30-33.

'See Douglas K. Adie and Gene L. Chapin, "Teenage Unemployment Effects of Federal Minimum Wages," Proceedings of Industrial Relations Research Association, 23rd Annual Meeting, 1970, pp. 117-27, and Thomas Gale Moore, "The Effects of Minimum Wages on Teenage Unemployment Rates," Journal of Political Economy, Vol. 79, No. 4, July/August, 1971, pp. 897–902, for attempts to measure the effects.

38

39

LABOR ECONOMICS

only in the month the federal minimum wage for covered workers changed. Monthly figures for average hourly earnings for production workers, designated AHE, are contained in Employment and Earnings Statistics (B.L.S.) for the period January 1954-December 1970. The real federal minimum wage X used throughout this paper is defined as:

To derive the model consider a non-linear demand for labor func tion,

where D is the quantity of labor demanded expressed in number of workers, I is an index of business conditions and a, b, and c are parameters.

is a non-linear supply of labor function where S is the number of workers offering their labor services; d and f are parameters.

For downward sloping demand and upward sloping supply functions: c<o and f>o, so c-f <o and

Using U, as a proxy for business conditions, and new parameters, (4) can be approximated by the following expression in log form:

In U1 = a + ẞ In U. + ylnX

с

which is the basic form of the model.

(5)

40

IRRA 24TH ANNUAL WINTER PROCEEDINGS

First Order Autoregressive Model

A first order autoregressive model was used instead of ordinary least squares multiple regressions to measure the unemployment elasticities y1, because of positive serial correlation in the residuals.3 The model used is

Yi,

i,t

InU1t plnU1,t-1 = (1 − p) a1 + ẞ1 (lnUe,t — plnUc,t-1)
+ y1 (InXt-m- plnXt-m-1)

(6)

where i is the classification of teenage labor, t is time and m is the number of months lag. Equation (6) is similar to equation (5), except that it is expressed in partial first difference form with the possibility of different time periods being introduced between the unemployment rates and real minimum wages. The regressions were run using equation (6) with time lags of 0, 8, 16, 24, 32, 40, 48, and 60 months. A value of p was chosen for each equation which brought the Durbin-Watson statistics close to 2.0. The unemployment elasticities calculated by this method are listed in Table 1 with their standard errors in parentheses below them.

The elasticities in Table 1 are plotted in diagram 1 to demonstrate more clearly the trend in the elasticities as the lag is lengthened. From diagram 1 it appears as though the unemployment elasticity reaches its maximum around the interval 8-16 months for all classifications of teenagers and then tapers off. It is difficult to pinpoint

The unemployment elasticity for teenage sub-group i is the percentage change in the unemployment rate when the real federal minimum wage changes by 1 percent, and is estimated by 71.

the lags more precisely without calculating more elasticity coefficients.

Of the 56 elasticities listed in Table 1, 31 are significant at the 5% level. Most of the insignificant elasticities occur for 48 and 60 month lags.

For all teens, white teens, non-white teens, female white teens, and male non-white teens the elasticities are still close to 50% of their maximum values after 40 months. The elasticity for white male teens decreases rapidly after 16 months and by 32 months is almost negligible. The elasticity for non-white female teens decreases rapidly after 32 months and is only 14% of its maximum value after 40 months.