• The School-to-Work Transition for Men
• The School-to-Work Transition for Women
• The Static Perspective of the School-to-Work Transition

Given our definition of school leaving (and realizing that there is some increase in schooling attainment from the initial SLG), in this section we present a conventional static picture of the employment activities of young men and women by SLG as they age. A similar, but not identical, description could be generated from a cross-sectional survey such as the Current Population Survey.[12] In the discussion that follows, we first analyze the patterns for young men in total, then disaggregated by race or ethnicity. We then examine those patterns for women.

We note at the outset that the static analysis presented in this section does not fully exploit the longitudinal nature of the NLS-Y data which allows us to measure time spent in various states as young adults progress through their early labor market careers. In the dynamic analysis that follows in Section 4, we use the NLS-Y longitudinal employment histories (for a given individual) to examine patterns of job holding and job duration over the early work career.

The School-to-Work Transition for Men

One picture of the school-to-work transition can be developed from the distribution of activity status of cohorts of young men at differing ages. Our analysis differentiates four activities defined hierarchically (i.e., those who might be included in more than one activity category are included in the earlier category):

1.  Working full time (35 or more hours per week) 


2.  Attending school (and not working full time) 


3.  Working part time (and not attending school) 


4.  Neither working nor attending school.   

With this classification, young men working 35 or more hours per week are classified as full-time workers even if they attend school. (This classification is consistent with our definition of SLGs, but not with the alternative definition applied in Appendix B.) The activity status is determined as of the day the person turns the given age, rather than as an average over the entire year the person was a given age. An individual is only included in the sample once he/she leaves school and we can assign an SLG.[13]

Table 3.1 presents the distribution of the sample of young men at each age across these four activities for four of our five SLGs (excluding the BA+ group). The experiences of the high school dropout SLG are shown in panel a of Table 3.1. For this group, the data in the table are consistent with Osterman's (1980) view of "hanging out." At age 21, a quarter (26.7 percent) of males in the high school dropout group (31.9 percent according to the alternative SLG definition of Appendix B) are neither working full time nor in school, and through age 29 the figure barely drops below 20 percent (22.0 percent according to the alternative definition).

Panel b of Table 3.1 indicates there is some "hanging out" among high school graduates as well. Through age 21, more than 20 percent are neither in school nor working full time. As with the high school dropouts, this fraction comes down only slowly during the early twenties. Not until their 27th birthday does the percentage of high school graduates neither working full time nor in school drop below 10 percent (i.e., working part time or not working and not in school). However, the rate is still much lower than for high school dropouts at the same age, a group in which 19.3 percent is still neither working full time nor in school.

For the higher SLGs, it becomes relevant to ask whether we want to compare people by their chronological age or by their time since school leaving. While the rows in Table 3.1 define a specific age, each successive row represents approximately one additional year after school leaving. Comparing across the panels of Table 3.1, it is clear that the transition to full-time work becomes smoother as the amount of education acquired before school leaving increases. Beginning with the college graduate SLG, in approximately the second year after graduation (i.e., at the 23rd birthday), less than 8 percent is neither employed nor in school, and only an additional 6 percent is employed part time. The some college SLG fares only slightly worse. At age 20 (approximately the same point since school leaving), the fraction not working or in school is under 8 percent and their part-time employment rate is under 10 percent. The high school graduates fare much worse.

At age 19, their nonactivity rate (neither in school nor working) is 15.3 percent and their part-time employment rate is 13.4 percent; and the high school dropout SLG is worse still at age 18, with a nonactivity rate of 21.3 percent and a part-time work rate of 10.0 percent. Thus, if we view full-time work as the normative state, the rates in the second year after school leaving for the four SLGs from the college graduates down to the high school dropouts are 80.8, 64.6, 62.1, and 48.0 percent.

As discussed earlier, an alternative perspective is possible. These high rates of non-full-time work are consistent with the leisure-equalization hypothesis discussed in Section 2. Figures 3.1 and 3.2 reinforce this alternative perspective. Figure 3.1 plots the fraction in each SLG neither working nor in school at each age, while Figure 3.2 plots the proportion in each SLG working part time at each age.Because higher SLGs leave school later, the line for each higher SLG starts farther to the right.

Figure 3.1--Percentage of Men in School-Leaving Group Neither Working Nor In School

Figure 3.2--Percentage of Men in School-Leaving Group Working Part Time

According to Figure 3.1, after age 25, once the nonactivity rate for college graduates drops sharply following their first two years out of school, the striking feature is the similarity of the nonactivity rate across all but the high school dropout group. The high school dropouts clearly stand out, with considerably higher rates of nonactivity over the entire period examined. The rates for the high school graduate and some college groups fall steadily as they age and their college-graduating peers leave school.

Figure 3.2 presents similar results for the fraction of each SLG working part time at each age. Again, except for high school dropouts, the other three SLGs start out with a considerable amount of part-time work. Such work is clearly transitional. Within a few years of school leaving (about three years for high school graduates, two for those with some college, and one for college graduates), the rates of part-time work fall sharply. Once again, however, the higher SLGs have a smoother transition, but the gaps among SLGs close considerably over time.

From this static analysis of activity status at each age, it appears that the outlier group is not the high school graduates but the high school dropouts. For dropouts even at age 23, the nonactivity rate is 18.4 percent, and their full-time employment rate is only 71.3 percent. At age 25, the nonactivity rate only drops to 14.2 percent, while the full-time employment rate attains only 77.9 percent--about 10 percentage points below the rates for the high school graduate and college graduate SLGs at the same age.[14]

Differences in the Patterns for Men, by Race and Ethnicity

While the figures for all men reveal that there is some heterogeneity in their work and schooling behaviors across SLGs, we can also examine whether there are differences in these patterns within SLGs by race and ethnicity. For this analysis, we define three groups of men in the NLS-Y: black non-Hispanics, Hispanics, and white non-Hispanics and others (primarily Asians). We refer to these groups as blacks, Hispanics, and whites, respectively. Although the NLS-Y oversampled minority men, there are sometimes too few black and Hispanic men in each age and SLG cell to make valid comparisons. For this reason, we are unable to make any comparisons across race/ethnicity groups within the college SLG, and for some of the younger and older ages in the other SLGs.

Table 3.2 shows the fraction of men at each age within the four activity categories by race/ethnicity, first for the dropout group, followed by high school graduates and those with some college. The basic pattern for the SLG as a whole is evident within the three race/ethnic groups. In particular, rates of full-time work are always lower for high school dropouts than for high school graduates, while the reverse is true for the rates of nonactivity. The differences between dropouts and graduates are of a similar magnitude for all three groups. For example, at age 26, 82 percent of black high school graduates are working full time, while only 69 percent of black dropouts fall into this category. The corresponding figures for Hispanics show a similar gap, from 82 percent to 72 percent. Likewise, among whites, 89 percent of graduates is in full-time work compared with 80 percent of dropouts.

Although the patterns are similar within the race/ethnic groups shown in Table 3.2, there are striking differences among the three groups in the distribution of men within each SLG across the four activities. The sharpest contrasts occur in the rates of nonactivity. At nearly every age, for example, black high school dropouts are more likely to be out of work and out of school than their Hispanic and white counterparts. Hispanics rank consistently in the middle, with the next highest level of nonactivity among the dropouts; the whites have the lowest fraction of dropouts out of work and out of school at any given age. For a given age, black dropouts are twice as likely as white dropouts to be nonactive. This same ranking (blacks being more nonactive than whites) is evident for male high school graduates, and the differences are again large. The pattern is repeated among those with some college for virtually all ages where a comparison can be made.

Differences exist within the SLGs, as well, in the fraction engaged in full-time work at any age. Here again, white men stand out as having consistently the highest fraction of dropouts classified as full-time workers. The pattern is broken only at younger ages for high school graduates and those with some college. While black dropouts and high school graduates typically have lower rates of full-time work compared with Hispanics, the two groups show very similar fractions in this status among those with some college. In fact, white dropouts look like black high school graduates in their rate of full-time work at each age.

The results in Table 3.2 are consistent with other data documenting significant differences in labor force participation rates and unemployment rates among white, black, and Hispanic men (Freeman and Wise, 1982; Wachter and Kim, 1982).

	N				Working full time				In school, notworking full time				Working part time, not in school				Not working, not in school
Age	B	H	W		B	H	W		B	H	W		B	H	W		B	H	W
a. High school dropouts
18	201	177	442		43.3	47.5	49.0		15.6	11.0	22.8		7.0	16.9	9.8		34.2	24.5	18.5
19	293	226	512		41.6	53.5	59.2		10.4	8.2	14.6		12.8	9.5	8.7		35.1	28.7	17.6
20	311	232	527		52.7	61.3	65.3		4.7	7.6	10.3		10.1	6.7	6.2		32.5	24.4	18.2
21	315	230	524		57.1	62.0	69.7		3.6	4.7	7.6		10.3	5.2	5.2		28.9	28.1	17.6
22	312	229	515		60.7	62.0	74.3		4.1	4.0	6.3		7.5	10.3	5.6		27.7	23.7	13.9
23	307	223	509		59.2	69.1	74.6		2.0	3.7	6.1		8.4	3.7	4.7		30.4	23.5	14.7
24	298	216	500		60.6	77.5	75.6		0.6	1.1	4.2		8.0	3.2	7.3		30.9	18.2	12.9
25	294	210	488		63.2	72.8	82.1		0.7	0.6	2.8		6.2	8.1	5.4		29.9	18.4	9.7
26	263	194	441		68.7	71.9	80.3		0.8	1.4	2.0		6.1	10.5	4.4		24.4	16.2	13.3
27	207	133	321		68.9	75.8	82.9		0.0	0.6	1.5		2.8	5.9	5.3		28.3	17.7	10.3
28	145	--	179		68.1	--	77.1		1.4	--	3.6		2.2	--	3.7		28.2	--	15.6
b. High school graduates
18	118	--	277		45.7	--	59.4		2.5	--	4.8		18.5	--	20.2		33.3	--	15.6
19	281	132	612		50.6	65.2	63.7		8.9	6.8	9.3		12.8	11.8	13.6		27.6	16.3	13.4
20	355	149	669		60.6	71.7	67.4		8.0	7.3	9.8		10.3	6.8	12.9		21.1	14.2	9.8
21	358	152	667		69.6	73.2	71.4		7.5	8.8	8.4		8.8	6.0	9.4		14.1	11.9	10.8
22	356	151	658		71.1	73.9	77.7		4.0	8.2	6.9		5.6	4.1	7.2		19.3	13.9	8.2
23	353	149	655		74.0	78.4	82.1		5.1	4.8	5.0		6.0	4.9	5.8		15.0	11.9	7.0
24	347	146	650		78.7	84.1	85.2		1.2	3.9	2.6		6.1	2.8	6.0		14.1	9.2	6.2
25	337	144	642		80.9	85.5	89.2		0.7	2.0	2.2		6.2	3.7	3.9		12.2	8.9	4.7
26	317	134	584		82.3	81.7	88.7		0.5	5.0	1.5		4.2	4.7	4.5		12.9	8.6	5.4
27	246	115	456		81.1	77.5	90.9		1.4	4.0	2.1		6.3	9.9	4.8		11.3	8.6	2.2
28	185	--	334		80.4	--	92.4		1.4	--	1.0		7.1	--	2.4		11.2	--	4.2
29	118	--	207		85.8	--	90.1		0.6	--	4.4		2.0	--	1.7		11.5	--	3.8
c. Some college
20	100	--	214		51.9	--	67.1		23.4	--	18.2		11.0	--	9.0		13.7	--	5.6
21	138	--	307		52.5	--	64.4		24.1	--	24.6		7.3	--	4.0		16.2	--	7.0
22	167	103	350		59.5	63.0	62.3		17.1	11.7	27.5		9.9	12.9	7.0		13.6	12.4	3.2
23	178	107	371		65.7	65.5	66.1		12.6	16.1	18.3		10.0	8.9	8.6		11.8	9.5	6.9
24	185	111	379		74.5	72.5	77.0		8.9	13.9	10.7		9.0	5.2	5.1		7.6	8.4	7.2
25	181	107	380		77.1	74.8	81.7		6.2	15.4	8.2		7.3	3.2	7.1		9.5	6.6	3.1
26	169	101	364		77.1	75.4	84.4		3.5	8.6	7.7		6.1	7.1	5.1		13.2	8.9	2.8
27	137	--	298		77.8	--	86.5		6.1	--	4.7		5.9	--	5.1		10.2	--	3.8
28	111	--	246		85.9	--	90.2		4.4	--	3.0		4.0	--	4.2		5.7	--	2.6
NOTES: B = black non-Hispanic, H = Hispanic, W = white non-Hispanic and other.     N is the number of individuals in the sample at least through a given age. Results are shown when sample size for a given age-SLG-race/ethnic group combination exceeds 100.

These differences persist for the three school-leaving groups we consider in Table 3.2. Thus, the patterns evident in the aggregate are not simply the result of differences in the educational composition of men in the three race/ethnic groups, but are due as well to underlying differences in early employment experience within different schooling groups. In Section 4, we further explore whether these differences are evident when we view the early labor market period from a dynamic perspective.

The School-to-Work Transition for Women

As we noted in Section 2, we stratified our analysis by gender because we expected that fertility among young women will differentially affect their schooling and early labor force patterns. While we expect more women to be out of the labor force or school as a result of pregnancy and childrearing responsibilities, we do not explicitly attempt to account for this "activity" in either the static analysis presented here or the dynamic analyses that follow in the next section. Here we present a set of tabulations that, while more limited, parallel those just presented for men.

We begin by calculating the fraction of women in each of the four activity statuses at each age, stratified by SLG. Table 3.3 reports results comparable to those we reported for men in Table 3.1. Figures 3.3 and 3.4 (comparable to Figures 3.1 and 3.2 for men) show the fraction of women neither working nor in school and the fraction working part time.

A comparison of the results for women with those seen earlier for men reveals the expected differences. The fraction of women neither working nor in school is higher at every age compared with that of men for the four SLGs shown in Figure 3.3. Instead of the decline with age in rates of nonactivity seen for men, the lines are approximately flat for women in the three lowest SLGs. Among women in the college graduate SLG, there is evidence of a U-shaped pattern: The percentage neither working nor in school declines until about age 24, and the fraction increases thereafter. This pattern is most likely the result of the childbearing that was delayed for these women until they had completed schooling and had begun their work career. Among women, there is less evidence of a convergence in the pattern across the three highest SLGs. The ordering of activity rates of the SLGs and approximate gaps between the rates for SLGs--more educated women have lower nonactivity rates--does not diminish appreciably as the women age.

However, women with a college degree and those with some college look more similar compared with high school graduate women. As before, the high school dropout women offer the sharpest contrast.

The pattern of part-time work among women, as shown in Figure 3.4, is also substantially different from that seen for men. At virtually every age, women in each SLG are more likely than men to be working part time. As with their male counterparts, there are few differences past the early twenties for women in the four SLGs; the fraction of college graduate women working part time is within a few percentage points of the share for high school dropouts. At variance from the men, however, is the little change in the fraction working part time as the women age. From age 20 to 30, roughly 15 percent of women are in this category.

Figure 3.3--Percentage of Women in School-Leaving Group Neither Working Nor in School

Figure 3.4--Percentage of Women in School-Leaving Group Working Part Time

The Static Perspective of the School-to-Work Transition

In this section we have viewed the school-to-work transition as a series of snapshots taken at each age in the early labor market careers of young men and women in the NLS-Y. When these series of snapshots are pieced together, we are left with a picture of the early labor market experiences that is extremely varied. For some--particularly men who leave school without a high school diploma and minority men in the high school graduate SLG--we see a substantial fraction who are not in a job and not in school. While fewer men are in this position by their late twenties compared with their early twenties, the fraction is still high even after more than 10 years of potential labor market experience. This fraction exceeds that experienced by men with more schooling, such as college graduate men, who appear to reach a high rate of full-time employment soon after leaving school. As we might expect, the patterns for women vary substantially from those of their male counterparts, presumably because women are more likely to take time out from schooling and work for childbearing and childrearing, responsibilities that may conflict with schooling and work.

While this static view of the school-to-work transitions reveals some important contrasts among young adults based on gender, school-leaving group, and race/ethnicity, the analysis obscures the underlying dynamics of the early labor market career. Should we take the static view presented in this section as evidence of considerable "churning" and "milling about"? As young adults move between the various states of work and school, how many jobs do they hold and for how long? Is the process of settling down similar regardless of schooling attainment at the time of labor market entry or other demographic characteristics? These are the issues we address in the next section.

[12] It would not be identical because one cannot compute SLGs from the CPS. Instead, one would use current schooling. As we saw in the preceding section, the two concepts are not identical.

[13] As a result, the sample sizes increase in the early years, then decrease in later years because, by the 1990 interview, many of the sample members had not yet reached the older ages and because of panel attrition. Results are shown only if the cell size for a given age-SLG combination exceeds 150. When we disaggregate by race-ethnicity, the minimum cell size is lowered to 100.

[14] The comparison with the some college SLG is not appropriate, since many of them are back in school. Few of the high school dropouts are in school in their mid-twenties.