Effects of Marriage on Labor Participation

Decreasing marriage trends can be causally associated with the decline of adult male labor participation. This withdrawal of able-bodied workers from productive employment amounts to the removal of one of the key components for domestic production at the macro-economic level.

Percent Males Employed

1. Family Structure and Labor Patterns

Labor participation and the technical capacity of the workforce are the core factors of production that drop during a depression.1) The decline in adult male labor participation2) is associated with non-marriage trends within and across occupation classes3) of the U.S. population. Occupation type and education level4) are generally recognized specifiers5) of human capital.6)

As shown by the charts below, for all occupation classes and through myriad macro-economic changes, married men are consistently more employed than single or cohabiting men. There is a well-defined difference in the rate of not working for men of different marital statuses. This is the “gap” between the upper and lower curves in the left graphs below.

Percent Professional Occupation Class Males Not Working

Percent Skilled Labor Class Males Not Working

Percent Salesmen Not Working

Percent Administrative Support Occupation Males Not Working

Percent Unskilled Labor Class Males Not Working

Percent Service-sector Males Not Working

2. Effects of Marital Status on Labor Participation

Across all occupation classes in the charts above, a consistent and well-defined difference in rates of employment between men of different marital statuses persists (seen in the left graphs). All the while, the married state is consistently losing ground to singlehood and cohabitation (seen in the right graphs). It is the state of marriage that makes the men in these subpopulations be employed in much greater percentages, regardless of other qualities7) found in those men.

Marital state must affect behavior because the population is shifting from one marital state to another as the decades progress.8) If the same quality of man had merely moved from married to unmarried states, he would have brought his propensity to work with him. Had that happened, non-married men would see a relative increase in their group’s propensity to work. This increase did not happen, as shown in the charts above. The men who shifted from married to unmarried states took on a lower propensity to work in the process.

Likewise, if a sub-population of married men with a lower propensity to work were to move into the growing single and cohabiting population groups, the remaining married men should see their average propensity9) to work increase. This is not seen in the charts above, hence no such sub-population exists. The men shifted from married to unmarried states and took on a lower propensity to work through that transition. The abandonment of marriage caused this segment to reduce its labor participation.

3. Other Relevant Factors Affecting Labor Participation

3.1 Feminization of Labor

Women entering the workforce cannot adequately explain the lower level of labor participation by unmarried men (at least after 199010) as this process stabilized by the 1990s (see chart below). After this time male labor participation continued to fall off.11)

Percent Female Employed versus Percent Male Employed

3.2 Globalization of Labor

The effects of globalization of the labor market cannot be an explanation for this lower level of labor participation as researchers have controlled for occupation class,12) and this relative employment difference holds irrespective of occupation class and its susceptibility to global competition.13)

3.3 Favoritism of the Married Over the Single

Relative employment cannot be attributed to management favoritism of married men over singles. Two data are necessary to demonstrate this. First, note that because their human capital is more developed, married men earn more generally.14) Second, note how the gap’s width (between singles and married men, with cohabiting men in between) increases during recessions. Why? Especially during recessions, management acts to hold on to its most valuable labor, even though it is its most expensive labor. This is a general phenomenon: In depressions, lower-skilled labor is let go of first, and in massive quantity.15) It is single men that here are most identified with lower-skilled labor. Relative wage stickiness (i.e. the maintenance of relatively high wages during economic downturns) results as the higher-skilled and higher-paid employees are retained.

Labor patterns demonstrate that markets choose married men because they are, relative to all other groups, the most valuable.16)

The second datum above also shows that non-marriage creates economic insecurity for the whole nation: Compare the latest recession to the recession of the early 1980s. In both recessions each marital status group had similar employment responses, but the latest recession saw a further weakening of the overall workforce because the population that is married is smaller.

4. Non-Marriage Reduces Macro Labor Participation

American labor participation for men has been dropping off since the 1960s. Correspondingly, a “gap” exists between the participation rates of married men and unmarried men. This gap and the population shift towards non-marriage alone17) can immediately account for around half of this decrease in labor participation.18) 19)

5. Risk of a Depression

Entering into marriage affects economic agents’ behavior.20) The abandonment of marriage leads to reduced population21) with its eventual loss in human capital.22) This rejection of marriage is also cause for reduced labor participation. Together, these factors put the United States at risk of economic depression. The continuance of this cultural-demographic drift away from marriage and into household structures that are less productive and less engaged in the economy will exacerbate this risk over time.

1) Timothy Kehoe and Edward Prescott, Great Depressions of the Twentieth Century, Research Department, Federal Reserve Bank of Minneapolis (2007).
A depression sees a fall-off of at least 20 percent from the potential GDP. This fall-off lasts for more than a decade, with most of the fall-off occurring before the first decade of the depression is out.
2) The term labor participation (being employed) is narrower than labor force participation (being employed or currently unemployed), defined by the Bureau of Labor Statistics. Note, men are either employed, unemployed, or “not in labor force.“ As there has been an increase in men with long-term “not in labor force” status, the grouping of unemployed with those not in the labor force is analytically more tractable.
3) Typical occupations within each occupation class include:
Professional occupations: accountants, chemists, professors, doctors, editors, engineers, lawyers, nurses, pharmacists, social workers, teachers, managers
Skilled labor occupation: technicians, carpenters, craftsman, mechanics and repairmen, plumbers, metal workers
Salesmen: all types except retail
Administrative support occupations: clerical secretaries, typists, bookkeepers, phone operators, office boys
Unskilled labor occupations: drivers, deliverymen, furnacement, assembly line workers, day laborers
Service-sector occupations: wait staff, housekeepers, retail salesmen.
4) Education level is analogous to that of occupation type; however, occupation type gives a simpler view into labor dynamics. This overview, therefore, focuses on occupation. For a deeper explanation on education, see Henry Potrykus and Patrick Fagan, “Non-Marriage Reduces U.S. Labor Participation: The Abandonment of Marriage Puts America at Risk of a Depression,” The Marriage and Religion Research Institute (August 2012), available at http://marri.us/research/research-papers/non-marriage-reduces-u-s-labor-participation/.
5) Gary Becker, Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education, Third Ed. (Chicago: University of Chicago Press, 1994).
6) Human capital comprises the skills, capacities, and know-how contained in the human person and valued by the market for economic production.
7) This refers to human capital qualities not measured by occupation class, tenure (age after entry into the workforce—Potrykus and Fagan, Decline of Economic Growth: Human Capital and Population Change), education level, or marital status.
8) This shift occurs both within occupation classes and generally across all working males. Thus, the decrease in employment is not from a population shift across occupation classes; in fact, the contrary happens: The Current Population Survey shows men generally moving away from occupation types that employ them at lower levels.
9) This should hold for both relative and absolute propensities to work.
10) And then researchers may ex post carry out this conclusion back through the 1960s, 1970s, and 1980s. This is justified because of the robust near-constancy of the “employment gap” between marital status classes. Otherwise, there would be a temporary interaction between single male labor and women entering the workforce over the 1960s, 1970s, and 1980s. This interaction then must disappear, to be promptly replaced with the normal marital status effect persisting across the 1990s and 2000s. (Those decades experience the continued, gradual on-take of non-marital household arrangements).
11) The argument that cannot hold would go something like this: Single and cohabiting men find themselves disproportionately in occupations where increased competition from women entering the workforce saps jobs, and their employment prospects suffer in consequence.
12) The argument that cannot hold would go something like this: Single and cohabiting men find themselves disproportionately in occupations where increased competition from labor globalization saps jobs, and their employment prospects suffer in consequence. This argument of course ignores the dynamism of the labor market and the possibility for workers to retrain themselves and find work more in line with their stability preferences. It can be readily conjectured that it is these stability preferences that in part differentiate the marital status classes with respect to their propensity to work.
13) This is a stronger result than that given by merely focusing on occupation classes not susceptible to ‘global labor arbitrage,’ i.e. their labor not having a high degree of relocatability in the act of economic production.
14) The proof of this is Henry Potrykus and Patrick Fagan, The Divorce Revolution Perpetually Reduces U.S. Economic Growth, available at http://marri.us/research/research-papers/the-divorce-revolution-perpetually-reduces-u-s-economic-growth/, techreport (MARRI, 2012). This is perfectly consonant with the greater amount of time employed during which men accrue human capital.
15) Timothy Kehoe and Edward Prescott, Great Depressions of the Twentieth Century, Research Department, Federal Reserve Bank of Minneapolis (2007).
16) Again, informally, firms would not so consistently maintain the much higher-priced labor when budgets are tightest—in recessions and depressions—unless this labor (here the married men of the charts above) had greater value. The quantification of these statements is beyond the scope of this entry.
17) That is, without taking into account any multiplier effect, i.e. a weakening of the overall economy from this pull-back in production.
18) The “Percent Male Employed” Chart above shows roughly a 10 percent drop in participation over the 1960s to the 2010s. Across education classes there is roughly a 15 percent gap at 2010. Also across education classes, there is a shift from 90 percent of adult men aged 25 to 54 being married to roughly 60 percent (or 50 percent, depending on the class) still choosing marriage as one traces over the 1960s to the 2010s. This shift represents a fractional change of about 3/10 of the population losing 15 percent of its employment rate. (Proper accounting for population dynamics can make this number more accurate.) 15% x 0.3 is a 5 percent drop in employment rate for the entire workforce.
19) Women entering the workforce and consequent shifts in the competitive labor market and (household) work preferences may account for part of the overall drop-off in labor participation, though the ‘floor’ levels of the rate of not working (levels during times of economic strength) in the charts above only shift a percentage point or two over the decades. This sociological phenomenon may be coupled with other labor market and sociological dynamics; for example, global labor arbitrage, a labor concentration in the “construction” sector and subsequent collapse of that sector, and welfare benefits may interact with women entering the workforce and the abandonment of marriage.
20) As shown in Potrykus and Fagan, Non-Marriage Reduces U.S. Labor Participation: The Abandonment of Marriage Puts America at Risk of a Depression and Potrykus and Fagan, Decline of Economic Growth: Human Capital and Population Change.
21) Henry Potrykus and Patrick Fagan, Marriage, Contraception & The Future of Western Peoples, available at http://marri.us/research/research-papers/marriage-contraception-the-future-of-western-peoples/, techreport (MARRI, 2011) contains a stark representation of this.
22) Potrykus and Fagan, Decline of Economic Growth: Human Capital and Population Change