Evolution of the married wage premium

The married pay premium is the term used to depict the phenomenon whereby married work forces, on norm, earn more than single work forces. In a 1996 paper ( Loh, 1996 ) begins by saying that “virtually all cross-sectional pay surveies find that presently married work forces typically earn a higher pay than their non-married counterparts” . Further, Cornwell & A ; Rupert ( Cornwell and Rupert, 1997 ) province that “Married work forces earn more than single work forces. This fact is impregnable and is robust across informations sets and over time” .

Even after commanding for a big figure of factors, including differences in degrees of instruction, experience and other features, the estimated derived function is big, runing from 10-50 % depending on the survey.

There are three chief theories that attempt to explicate the phenomenon. The first Centres on married work forces being more productive, due to household specialization. The 2nd theory is the choice hypothesis, which reverses the causality. It purports that work forces with a higher earning potency are more attractive in the matrimony market and are therefore more likely to get married. The 3rd is that married work forces possess an unobservable familial feature that at the same time increases the likeliness of them gaining more ( due to some employer prejudice ) , and makes them more attractive on the matrimony market.

There are other accounts for the pay premium which have been investigated. For case, Reed and Harford ( Reed and Harford, 1989 ) suggest that married work forces work under inauspicious working conditions, and therefore the premium they receive is a “compensating differential” . Research workers in other Fieldss suggest two other grounds. One is that married work forces are better rewarded because they conform to a societal outlook that work forces should be married and back up their households. In this position, non-married work forces and married, working adult females differ from the societal “ norm ” and are therefore penalised comparative to married work forces. Second, married womans are resources for married work forces to work in fostering their callings. Specifically, married womans contribute to their hubbies ‘ success by supplying direct and indirect support, for illustration, advice and emotional assistance, or constructing societal dealingss that improve the hubbies ‘ standing with higher-ups. Unmarried work forces, or work forces with working married womans, do non bask such “spousal contributions” and are therefore penalised ( Pfeffer and Ross, 1982 ) .

Empirical Statisticss

There are voluminous Numberss of empirical surveies which include matrimonial position as a silent person variable to explicate the presence of an associated pay premium. The consequences of these surveies tend towards a general consensus that matrimony has a important positive consequence on net incomes. However, there is some dissension as to the size of the derived function.

In a survey of the differences in male net incomes in 12 different states, in the Luxembourg Income survey ( Schoeni, 1990 ) , it was found that married work forces had higher rewards in all 12 states, commanding for a figure of factors. The estimated coefficient of the matrimony variable ranged from 3 % to 31 % . In reiterating and polishing his analysis in 1995, Schoeni extended his survey to 14 states, and though he found that married work forces had higher rewards in all 14, the scope of the estimated coefficient altered to 0 % – 33 % . ( Schoeni, 1995 )

The historic tendencies of the matrimony premium for American workers is examined by Loh ( Loh, 1996 ) and the estimations of the size of the coefficient ranges from 11 % in 1959 to 23 % in 1969. See table 1.

Table 1

Consequence of matrimonial position on hourly rewards for males 1940-80






Married Males






Divorced or separated












The first widely cited empirical survey of the matrimony pay premium was ( Hill, 1979 ) . Hill used informations from the 1976 Panel Study of Income Dynamics to prove for the consequence of matrimony on rewards and found that married work forces have higher rewards that widowed, divorced or separated work forces, who in bend, have higher rewards than never-married work forces. In consecutive arrested developments that the matrimony silent person remained important for work forces, despite adding really elaborate regressors to her analysis, including controls for human capital, work history, wellness position, business, industry and figure of kids. The estimated coefficients lay between 0.26 and 0.31 for white work forces, and 0.22 and 0.27 for black work forces. ( Hill, table 3, p589 )

Insert One More Survey

Theory 1: Family Specialization

The theory that household specialization makes married work forces more productive than individual work forces, has its footing in the work of Gary Becker ‘s theory of the household ( Becker, 1973 ) . Specialization allows one partner to concentrate on market production and the other on place production, enabling the former to get more market-specific human capital and thereby, higher rewards. Becker ‘s original theory on the formation of human capital attack recognises a direct relationship “between clip spent in a peculiar economic activity and the inducement to put in human capital particular to that activity” ( Becker, 1975 ) . He besides says that under changeless or increasing returns to scale, differences in comparative advantage among otherwise indistinguishable persons create different inducements to put in specific human capital and in allotment of clip as good. Hence he explains the gender division of labor within a family by concentrating on initial differences in comparative advantage between work forces and adult females in market and family work. Consequently, hubbies specialise in market work and married womans in family work because the chance cost of making so in each instance is lowest.

Attempts to prove this hypothesis have used steps of the married woman ‘s employment as a placeholder for specialization. The principle is that if specialization leads to higher rewards for work forces because their market productiveness is enhanced, so married work forces whose married womans do non work in the market, or work fewer hours, should hold higher rewards than single work forces or work forces with married womans in employment. However Loh ( Loh, 1996 ) counter-intuitively found that work forces whose married womans worked longer in the labor market received a higher matrimony pay premium than those whose married womans did n’t work at all.

On the other manus, Gray ( Gray, 1997 ) found the expected opposite relationship. The matrimony pay premium was 1.8 % lower for each extra 10 hours per hebdomad worked by the married woman in the labor market. The matrimony pay premium was 17 % for hubbies whose married womans specialised in family production, compared with 9 % for hubbies whose married womans spent 40 hours or more in the labor market.

These conflicting consequences suggest that wives engagement in the labor market is a weak index of the effects of family specialization on the matrimony pay premium. Hersch ( Hersch and Stratton, 2000 ) argue that this is because of viing income and permutation effects i.e. married work forces whose married womans are employed may pass less clip on housekeeping than married work forces whose married womans do n’t work because household income is greater, or they may pass more clip because the value of the partners clip may be greater. Therefore the net consequence of this depends on the magnitude of each constituent. The usage in the two old surveies of bridal employment as a placeholder for specialization assumes that the permutation consequence dominates the income consequence. However ( South and Spitze, 1994 ) find that the clip allocated to home production of hubbies whose married womans are employed, is non well different from that of those whose married womans are non employed.

2: Choice Hypothesis

The choice hypothesis reverses the way of the causality of the specialization theory. It suggests that work forces with a higher net incomes potency are valued as more economically attractive in the matrimony market and are hence more likely to get married than work forces with a lower net incomes possible. This is related to an thought mentioned briefly by Hill ( 1979 ) that work forces who are higher earners are viewed as more attractive couples.

Harmonizing to Nakosteen and Zimmer, “The net incomes premium mirrors the choice pattern.” ( Nakosteen and Zimmer, 1997 ) Married work forces are hence more productive, non because they become more productive through matrimony, but instead because they were more productive before matrimony. Nakosteen and Zimmer ( 1997 ) argue that “ [ … ] the matrimony premium is a statistical artifact originating from a screening procedure in which high pay males tend to be selected into marriage.” However, when choice effects are accounted for in their arrested development, the matrimony premium remains big, but becomes statistically undistinguished.

To command for choice prejudice, Ginther and Zavodny exploit a potentially exogenic cause of matrimony among some work forces. They use the day of the months of work forces ‘s first matrimony and their kids ‘s births to analyze whether the return to marriage differs between work forces who have a kid shortly after matrimony and other married work forces. They assume that constructs outside of marriage followed by matrimony, or as they term it, “shotgun weddings” , make matrimony a random event, thereby purging out choice effects from the arrested development. ( Ginther and Zavodny, 2001 )

Their consequences suggest the differences in matrimonial pay premium associated with work forces conventionally married and those undertaking shotgun nuptialss to be little, comparative to the size of the matrimony premium. The consequences do propose that the matrimony premium takes a different signifier across the two type of matrimony, with conventional matrimony participants gaining a higher immediate pay premium, but somewhat smaller returns to marriage overtime, as compared with males who conceive a kid out of marriage. This indicates the choice prejudice to be little or negligible.

Further, in a paper proving both the productiveness and choice hypotheses, ( Chun and Lee, 2001 ) find no grounds to back up the choice theory. They find that immeasurable gaining potency is non positively correlated with unobservable traits valued in the matrimony procedure and province that “the choice of high ability work forces into matrimony is improbable to be the cause of the married pay premium” .

3: Unobservable Genetic Characteristics/ Employer Bias

This theory is based around the impression that there is some employer bias accounting for the married pay premium. It is argued that employers use matrimonial position as a signal as to the character of the person and it is held that married persons possess qualities that are desirable to both married womans and employers, such as motive, trueness, finding and dependableness.

In a recent survey of the matrimony pay premium, Bardasi and Taylor used panel informations methods in their research which allowed for correlativities between unobservables, matrimony and rewards. They argue that failure to make so would hold biased the coefficients of involvement and some of the returns ascribed to marriage may really hold been returns to some unseen qualities correlated with matrimony. ( Bardasi and Taylor, 2008 )

Their cross-sectional analysis generated a pay premium for married work forces of 17 % , consistent with many of the old findings. However this falls by one tierce when they explicitly control for intra-household specialization. Specifically, each family job for which the married woman is responsible additions the hubbies pay by 2 % . However, utilizing panel informations they find that the matrimony pay premium falls, bespeaking that more than half of the ascertained premium in cross-sectional analysis is caused by unseen “individual specific heterogeneity” .

A old survey ( Cornwell and Rupert, 1997 ) find that the estimated effects of matrimony are virtually zero one time unobservable single specific effects are accounted for. Additionally, with their informations they find that to the extent that there is a addition from matrimony, it is “purely an intercept shift” . This means that the matrimony pay premium does non vary with old ages married, which is grounds against the theory of matrimony heightening productiveness through specialization, since it would be expected that capital accretion would be cumulative over a period of old ages and hence for the matrimony premium to increase with clip.


The size of the married pay premium has been reasonably consistent over clip ( Goldin, 1990 ) . However, ( Blackburn and Korenman, 1994 ) find that it has decreased by 10 per centum points over the clip period 1967-1988.

They draw on Census informations for their arrested development analysis, and utilize the figure of old ages married as a placeholder for human-capital accretion within matrimony and their findings suggest that a autumn in the mean figure of old ages married among males has contributed to the diminution in the derived function.

In a paper trying to place the cause of the autumn in work forces ‘s return to marriage, the two chief theories offering accounts of why married work forces earn more than single work forces, are explored. ( Gray, 1997 ) He examines whether the recent additions in adult females ‘s labour force engagement could be the cause of worsening matrimony premiums. For illustration, in the US in 1960, the married adult females ‘s engagement in the labour force was 30.5 % ; by 1980, this rate had increased to 56.5 % . Therefore, if the matrimony pay premium originates from family specialization, so the recent rises in adult females ‘s labour force engagement could do it more hard or less worthwhile for hubbies to roll up human capital and hence falling matrimonial pay premiums.

However, it is possible that the increased labor market engagement of adult females is at least partially explained by, for illustration, lower birthrate rates or other alterations that lower place production demands on adult females ‘s clip ( i.e. labour salvaging devices ) , without altering the demands on hubby ‘s clip. ( Blackburn and Korenman, 1994 )

Alternatively, if the choice hypothesis were to keep, so the increasing figure of delayed or forgone marriages- due to the worsening stigma attached to being unmarried- could supply the account. In this state of affairs, the pool of high ability work forces available on the matrimony market would shrivel, coercing adult females to be less selective. The consequence of this would be decreased mean ability of work forces relative to that of single work forces, and therefore a diminution in the matrimony pay premium.

The diminution in the return to matrimony has been linked to marriage being a weakened establishment, given the altering construction of household units, and changing societal norms gnawing the gender division of labor linked with cohabitation. ( Waite and Gallagher, 2000 )

In response to this, Cohen ( Cohen, 2002 ) uses information from population studies to demo that the diminution in the matrimony pay premium may be overstated when informations for live togethering twosomes is included in the arrested development. The principle behind this is that gender based division of labor for live togethering twosomes is really equal compared with married twosomes. ( South and Spitze, 1994 ) so they therefore receive a smaller pay premium, as there is small, if no specialization. Since estimations of the matrimonial pay premium reflects net incomes inequality between work forces in different household state of affairss, and specifically compares married work forces with some mention group. Therefore the composing of the mention group is really of import and Cohen topographic points live togethering work forces into the mention group.

He finds that commanding for cohabitation reduces the rate of diminution of the matrimony pay premium by 35 % . This consequence poses a challenge to the productiveness theory which would propose that as the figure of hours worked by adult females additions so the diminution in work forces ‘s return to marriage should fall as there is less household specialization. However, adult females ‘s on the job hours increased by 39 % over the period studied, proposing that the division of labor within the matrimony establishment is non greatly changed.


While the being of a matrimony pay premium is good documented for work forces, probe into adult females ‘s income derived functions by matrimonial position has received much less attending. However, there is some grounds to propose the being of a punishment for being married, or conversely, a premium for being individual or cohabiting.

This is shown in research by Madalozza, who finds through empirical observation that married adult females have lower wage than non-married or live togethering adult females. She estimates a statistically important spread in rewards between married and live togethering adult females in the scope of 49-53 % , and when comparing married and individual adult females, the differential additions. ( Madalozzo, 2008 )

BARDASI, E. & A ; TAYLOR, M. ( 2008 ) Marriage and rewards: A trial of the specialisation hypothesis. Economica, 75, 569-591.

BECKER, G. S. ( 1973 ) A Theory of Marriage: Part I. The Journal of Political Economy, 81, 813-846.

BECKER, G. S. ( 1975 ) Human capital. 2nd Ed.

BLACKBURN, M. & A ; KORENMAN, S. ( 1994 ) The worsening marital-status net incomes derived function. Journal of Population Economics, 7, 247-270.

CHUN, H. & A ; LEE, L. ( 2001 ) Why do married work forces earn more: Productiveness or matrimony choice? Economic Inquiry, 39, 307-319.

COHEN, P. N. ( 2002 ) Cohabitation and the worsening matrimony premium for work forces. Work and Occupations, 29, 346-363.

CORNWELL, C. & A ; RUPERT, P. ( 1997 ) Unobservable single effects, matrimony and the net incomes of immature work forces. Economic Inquiry, 35, 285-294.

GINTHER, D. K. & A ; ZAVODNY, M. ( 2001 ) Is the male matrimony premium due to selection? The consequence of shotgun nuptialss on the return to marriage. Journal of Population Economics, 14, 313-328.

GOLDIN, C. ( 1990 ) Understanding the Gender Gap: An Economic History of American Women

New York, Oxford University Press.

GRAY, J. S. ( 1997 ) The autumn in work forces ‘s return to marriage: Declining productiveness effects or altering choice? Journal of Human Resources, 32, 481-504.

HERSCH, J. & A ; STRATTON, L. S. ( 2000 ) Household specialisation and the male matrimony pay premium. Industrial and Labor Relations Review, 54, 78-94.

HILL, M. S. ( 1979 ) THE WAGE EFFECTS OF MARITAL STATUS AND CHILDREN. Journal of Human Resources, 14, 579-594.

LOH, E. S. ( 1996 ) Productivity differences and the matrimony pay premium for white males. Journal of Human Resources, 31, 568-589.

MADALOZZO, R. ( 2008 ) An Analysis of Income Differentials by Marital Status. Estudos Economicos, 38, 267-292.

NAKOSTEEN, R. A. & A ; ZIMMER, M. A. ( 1997 ) Work force, Money, and Marriage: Are High Earners More Prone Than Low Earners to Marry? Social Science Quarterly, 78, 66-82.

PFEFFER, J. & A ; ROSS, J. ( 1982 ) The effects of matrimony and a on the job married woman on occupational and pay attainment. Administrative Science Quarterly, 27, 66-80.

REED, W. R. & A ; HARFORD, K. ( 1989 ) The matrimony premium and compensating pay derived functions. Journal of Population Economics, 2, 237-265.

SCHOENI, R. ( 1990 ) The Earnings Effects of Marital Status: Consequences for Twelve Countries.

SCHOENI, R. F. ( 1995 ) Marital position and net incomes in developed states. Journal of Population Economics, 8, 351-359.

SOUTH, S. J. & A ; SPITZE, G. ( 1994 ) Housework in Marital and Nonmarital Households. American Sociological Review, 59, 327-347.

WAITE, L. J. & A ; GALLAGHER, M. ( 2000 ) The Case for Marriage: Why Married Peoples Are Happier, Healthier, and Better off Financially.