Part of Real World Divorce: web edition | Kindle edition
The chapter explores the impact of the U.S.'s unusual family law system on the U.S. economy. The impact can be divided into four categories:
Spending is the most obvious area and the easiest to measure. Most American states have made getting custody of a child more lucrative than going to college and getting a job. At the same time we've made assignment of custody uncertain and subject to expensive litigation. We've instituted no-fault divorce in which plaintiffs seeking divorces are guaranteed to win every lawsuit that is filed, but retained the litigation structure from the old days when the defendant had a chance to block a divorce from going through.
The family law system also influences economic behavior by litigants and potential litigants. The second factor that we look at is the tendency of spouses in an intact marriage to behave differently depending on the prevailing divorce law. As discussed below, the low-income partner may not feel the need to work if he or she is accumulating a claim on the higher-income partner's earners.
Once in divorce litigation, the incentives for both parties to avoid profitable economic activity are substantial. Lawyers told us that they routinely advise plaintiffs not to get jobs so as to avoid impairing their alimony and child support entitlement. This leads to the third category of impact.
Finally there is the effect on children. The government provides financial incentives for Americans to divorce each other and/or rear children in single-parent homes. Do these children earn less as adults? Consume more in social services?
This analysis tries to compare the size of the economy with the American system of family law and litigation and the simplest possible European-style system in which (a) divorce can be obtained via an administrative process, (b) no alimony is available, (c) collecting child support is not profitable, (d) a custody presumption forestalls most litigation regarding children, and (e) legal fees in the event of litigation are limited.
[We are grateful to economist Alena Malyshkina for assistance with this chapter.]
"The USA has the highest proportion [among 16 countries] of children, as much as 50 percent, with any experience of living outside a two-parent family when they turn 15. … in many Western and Eastern European countries it is more common to find that around a fourth or a third of all children have an experience of that kind, at some time during childhood. … The USA stands out as an extreme case... "
-- "Children's experience of family disruption and family formation: Evidence from 16 FFS countries" (Andersson 2002, Demographic Research 7:7)
It seems safe to assume that divorce and single parenthood are more common in the U.S. than in other countries at least partially because divorce and single parenthood are more profitable here than in other countries. However, we don't have a good way to tease out the influence of financial incentives from cultural factors. Therefore the analysis below assumes that the rate of American parents splitting up, or never getting together in the first place, would be roughly the same, even if the profit potential were reduced to European levels. Thus there is no analysis of the costs of running two households rather than one.
Texas gathers unusually good statistics regarding its court system and we can use them to approximate total American spending on family court litigation.
According to U.S. Census American Community Survey data, 2009, Texas has a rate of divorced men (10 percent) that is close to the national average (9.2 percent). Lawsuits under family law are about 15 percent of all litigation in Texas. The Annual Statistical Report for the Texas Judiciary, Fiscal Year 2013, reports that Texas had 1.17 million child support cases and 386,000 other types of family law cases, about 1.56 million cases combined. This represents about 15 percent of the total of 10.1 million cases "disposed of" by the Texas courts in 2013.
The federal government's Bureau of Economic Analysis calculates that for 2012 the "gross output" of the Legal Services sector of the economy was $294 billion. Using the Texas percentage, and assuming that divorce, custody, and child support cases generate the same amount of legal fees as the average case, and that family law has a similar mix of litigation and other work as do other parts of the law, the total fees paid by American consumers to fight with each other in these areas is about $44 billion.
This estimate is consistent with an often-quoted statistic that divorce is a "$50 billion industry" because mental health professionals, accountants, and other expert witnesses also get paid when a person sues his or her spouse. It is also not too far from the "$30 billion in attorney fees alone" statistic cited in an April 10, 2013 Forbes article ("Can This Y-Combinator Startup's Technology Keep Couples Out Of Divorce Court?").
How does the Texas/BEA-based estimate of $44 billion compare to other sectors of the economy? The BEA says Americans paid $46 billion in 2012 for child care services. The $44 billion is more than the $39 billion that the BEA says we spent on "[private] elementary and secondary schools," which educate more than 5 million K-12 students. The $44 billion in fees is more than the $39 billion spent on "spectator sports". Parents spent more suing their co-parents (and defending those lawsuits) than Americans spent on books (also $39 billion), dairy cows and milk ($38 billion), bread and other bakery products ($36 billion), building aircraft engines and parts ($36 billion), veterinary care ($30 billion), manufacturing heavy trucks ($30 billion), or drilling oil and gas wells ($29 billion).
We asked Professor Allen what the effect of this spending was for macroeconomic growth. "You're creating transfers," he responded, "rather than new wealth. So [the money spent on fees] is a deadweight loss to the economy." Does that mean any time someone is paying a lawyer the economy is shrinking? "No," said Allen. "If you're in a complicated world you may need lawyers to settle disputes." Thus a principal difference between divorce in the U.S. compared to divorce in a Civil law jurisdiction is that we've turned the break-up of a two-income middle class couple with children into a "complicated world" that requires the kind of legal attention (and fees) as a contract lawsuit between two large companies.
Certainly whatever American parents are currently spending on divorce litigation is money that they could have spent on their children instead.
What would a European-style system cost? The Centers for Disease Control gives the number of divorces annually in the U.S. as roughly 850,000. If the average cost under an administrative process is $2000, e.g., to pay an accountant, that would be $1.7 billion per year. If we assume that, as in Denmark and Sweden, a handful of those cases would end up in litigation, we can round the alternative system cost up to $2 billion. Americans are thus spending an additional $42 billion per year in fees and costs.
Another cost of litigation is the time spent by litigants that they could have spent working (see the Litigation chapter for a calculation of this for one case). If we assume that every case involves two adults who spend, compared to an administrative process, at least 150 additional hours meeting with lawyers, researching the law in self-help books and on the Internet, being deposed, waiting in courtrooms, etc., that's 255 million hours. If we use the U.S. average wage of $23/hour (Bureau of Labor Statistics) and factor in the labor force participation rate of 63 percent, that's an additional $3.7 billion. (How time-consuming is an administrative divorce? A Danish father: "having been divorced for 11.5 years I have filled out a total of two forms: the divorce papers and an agreement to take care of the kids even if we were not living together.")
The government spends over $6 billion on running state offices of child support enforcement. If child support were not profitable it would be a smaller percentage of loser parents' income and be easier to collect. Let's assume then that these offices of child support enforcement reverted back to their 1979 cost of $200 million (about $665 million in 2016 dollars), adjusted to about $1 billion to reflect population growth of roughly 50 percent since 1979, and therefore that an additional cost of $5 billion is spent chasing down the big American numbers.
Lawyers and child support enforcement office employees need a courtroom in which to argue. California is home to 12 percent of the U.S. population and spends $3.8 billion running its courts (2016-17 data). If we assume the California's costs are typical and use the Texas percentage to allocate costs to divorce, custody, and child support litigation, that's roughly $4.75 billion spent to run state family courts. If we assume that administrators handling divorces would cost $100 per hour and spend an average of 9 hours per divorce, that's a total cost of $765 million, a savings of $4 billion compared to running our litigation-based system.
Nobody goes to prison for non-payment in countries where child support is limited to half of average actual child-related expenses. So we can consider all of the cost of incarcerating defeated parents to be related to the American system. No authoritative nationwide data is available, but a reasonable estimate (source) seems to be 50,000 people in jail or prison on any given day. This seems consistent with data from Michigan in the History chapter (1 in 7 loser parents jailed). The average annual cost of incarceration was $31,286 per inmate in 2012 (Vera Institute). Adjusted for inflation that's roughly $33,000 in 2016. Multiplied by 50,000 and the total is $1.65 billion. People imprisoned for not paying child support tend to have low incomes, so let's assume that they are earning an average of $20,000 per year. That translates into a shrinkage of the economy by $1 billion.
The best estimates that we could find suggest that 20-50 percent of the workload of state child abuse investigators is related to false denunciations by custody and child support plaintiffs seeking a personal financial benefit (see the Domestic Violence chapter). Certainly in Massachusetts it seemed to be that each child of divorce or unmarried parents generated an average of close to 1 investigation (0 for some children, but 1 at each stage of litigation, including post-divorce custody modification litigation, for others). These investigations often resulted in a government worker having to show up at the courthouse, wait for a lawsuit to be called, and then testify. However, we don't know of a good way to break these costs out of the total budget for social services agencies, so we're considering them at $0.
Total litigation, judicial, enforcement, and imprisonment costs therefore, above and beyond an administrative system, are roughly $57 billion. However, we don't include this in our analysis of a change in GDP because we assume that if consumers weren't spending money on divorce litigation, they would, at least in the long run, spend it on something else. For example, instead of spending $75,000 in divorce legal fees, the separating spouses might spend it on solar panels for the roofs of their new respective houses. However, we can consider a return on investment from that alternative spending. The legal fees will never yield a cash return in the future, but the solar panels should.
The legal environment hasn't changed much since 1990, so we can assume that we've spent, in today's dollars, roughly $1.5 trillion on divorce litigation since 1990. Children were not as profitable prior to the introduction of the child support guidelines and therefore custody litigation was not as intense. Let's assume that the costs were only half as high between 1970 when the no-fault revolution started and 1990 when children became reliable cash sources for adults. That's another $560 billion spent or more than $2 trillion total. What if $2 trillion had been invested in something productive instead, such as the solar panel alternative above? Let's assume a 2 percent annual real rate of return. That would translate to an economy $40 billion larger today.
[We are not accounting for the costs of disruption to the legal industry because law in the U.S. changes gradually and state-by-state. The custody litigator whose business dried up in 2015 when Nevada changed to 50/50 shared parenting can move across the border and take the Bar exam in winner-take-all California.]
"Divorce laws and the economic behavior of married couples," by Alessandra Voena, a University of Chicago economist, is a 2016 paper in Microeconomic Insights. Voena found that no-fault-but-take-the-money-anyway divorce laws reduced married women's labor force participation rate by 5 percent:
Household survey data from the United States shows that the introduction of unilateral divorce in states that imposed an equal division of property is associated with higher household savings and lower female employment rates among couples that are already married.
During the 1970s and 1980s, divorce laws were rewritten around the United States. Until then, mutual consent—the consent of both spouses—was often a requirement and upon divorce, property was assigned to the spouse who held the formal title to it; usually, this was the husband.
Then, profound state-level reforms brought about the so-called “unilateral divorce revolution.” Most couples now entered a legal system in which either spouse could obtain a divorce without the consent of the other and also keep a fraction of the marital assets, often close to fifty percent.
… forty percent of married couples and about one-third of all people over their lifetimes are divorced.
… a property division regime change that favors one spouse can improve her position inside the marriage, particularly if she can obtain divorce without the other partner’s consent.
… in such states [providing no-fault divorce plus 50/50 property division], women who were already married became less likely to work, by approximately 5 percentage points. By analyzing additional time use surveys between 1965 and 1993, I find that the decrease in the labor supply of women was associated with an increase in the amount of leisure time they enjoyed.
In states with equal division of property, the law favors women at the time of divorce. When the equal division of property grants them more resources in the event of divorce than they are receiving in the marriage, unilateral divorce means that they can use the threat of divorce in their favor while remaining married, thereby increasing their leisure.
There were approximately 65 million married women in the U.S. in 2011 (Census). Thus in a German-style system where a spouse's earnings cannot be tapped during or following a divorce, roughly an additional 3.25 million women would be in the workforce, earning money and paying taxes, if not for the combination of no-fault divorce and 50/50 property division. Median women's earnings in the U.S. were $38,740 per year in 2016 (Bureau of Labor Statistics). The Social Security Administration gives a ratio of 65 percent for median-to-mean wages and therefore the mean wage would be $59,600. The women who have withdrawn from the workforce in response to family law incentives would have earned approximately $194 billion.
Voena's findings are consistent with statistics from other sources. Women who graduated from America's top universities, and who there have greater access to high-earning husbands, exhibit a low labor force participation rate: "A recent study by Joni Hersch, professor at Vanderbilt Law School, makes that case. She looks at female graduates of our top universities – those presumably who have the best shot at shattering the glass ceiling – and finds that once they have children, they are more likely to quit their jobs than are women who graduated from less selective schools. … Perhaps most astonishing is that only 35 percent of women who have earned MBAs after getting a bachelor’s degree from a top school are working full time, compared to 66 percent from second-tier schools." ("Why Women Are Leaving the Workforce in Record Numbers," Fiscal Times, April 17, 2013)
Voena worked with data from a time period in which it was uncommon for a wife to out-earn a husband and in which judges were reluctant to apply alimony and property division laws in a gender-neutral fashion. In most states, a present-day husband of a higher-earning wife would face the same incentives as the wives in Voena's sample. How many men are potentially affected? "In 38 percent of heterosexual American marriages, the woman outearns her husband." (fivethirtyeight.com, February 5, 2015) If men respond to financial incentives in the same manner as women, quite a few have withdrawn from the labor force in response to the U.S. family law environment. Attorneys told us quite a few anecdotes about men who scaled back or quit their jobs once they had been married long enough to establish a financial claim against their higher-earning wives. However, without a paper like Voena's, but about men, we can't put a precise number on this behavior and therefore are accounting for the effect at $0.
For this element, and the ones below, we assume that when people withdraw from the workforce the economy shrinks by however much they were earning. In fact, the shrinkage might be a little less if wages for remaining workers rise in response to the smaller workforce.
Attorneys told us that they advised all of their clients to reduce their earnings during divorce litigation. Higher earnings hurt plaintiffs by reducing alimony and child support entitlements that might be sticky for 20 years or more. HIgher earnings hurt defendants by increasing the likely initial orders to pay alimony and child support. Lawyers in every American jurisdiction reported that judges were reluctant to reduce alimony and child support awards in the event that a defendant later suffered a reduction in income and thus it was critical to reduce income during the initial phase of litigation.
The typical American divorce seems to take roughly two years (see our state-by-state chapter). Using 850,000 as the number of divorces annually (CDC statistics cited above), there would be roughly 3.4 million Americans embroiled in the process at any given moment. Custody and child support litigation involve at least an equal number of people, bringing the total to roughly 7 million. Using the BLS labor force participation rate (above) of 63 percent, 4.4 million of those would ordinarily have jobs.
Let's assume that family court litigants who lose jobs take longer to find new ones, that some litigants withdraw from the workforce altogether, that some reject promotions or new jobs that would require more hours, and that some litigants refused overtime offered. If this reduces total earnings by 10 percent, that's equivalent to a reduction in earnings of 440,000 workers. The BLS reports median weekly earnings for all Americans of $827 in 2016. Adjusting for the median/mean differential this would be an average of $66,160 per year. That's 29 billion subtracted from the economy.
How big are the alimony flows? A May 15, 2014 U.S. Treasury Inspector General for Tax Administration report said that "In Tax Year 2010, 567,887 taxpayers claimed alimony deductions totaling more than $10 billion" and further noted that recipients of alimony reported at least $2.3 billion less in alimony income than payors reported paying. This result is that the Treasury lost tax revenue on $2.3 billion and general taxpayers have to make up the balance to keep the government funded. What other U.S. industry is comparable to collecting alimony? The U.S. Bureau of Economic Analysis says that manufacturing "Apparel and leather and allied products" is roughly a $10.5 billion industry. In other words, the total payments to all clothing factories in the U.S. is roughly equal to the total alimony paid (depending on whom you think is telling the truth to the IRS). We will assume for our analysis that 550,000 Americans in 2017 are either paying or receiving alimony.
Both plaintiffs and defendants have disincentives to earn money under the typical U.S. state's alimony system. The successful alimony plaintiff who later gets a job or a pay raise will have a lower "need" and is at risk of having payments cut during the next visit to the courthouse. The alimony lawsuit loser who works harder and gets a pay raise will have a greater "ability to pay" and may expect to lose one-third to one-half of that pay increase to the plaintiff. For both parties the alimony formula functions as a higher income tax rate and discourages work by reducing the percentage of income that can be spent.
Alimony defendants tend to have higher-than-average incomes and a fair amount of flexibility regarding how many hours they devote to work. Anecdotally, attorneys told us that alimony plaintiffs wait until the day that the alimony stops before taking a job. Alimony defendants told us that they were so filled with hatred for their plaintiffs that they acted against their own financial interest in refraining from income-producing activities that would have exposed them to a larger alimony claim. Certainly it seems safe to assume that the $10 billion in transfers ordered discourage at least $10 billion in work by plaintiffs and defendants. (This could happen if a defendant paying $50,000 per year declines a promotion requiring 10 extra hours per week of work and paying $25,000 per year addition while the plaintiff rejects a $25,000-per-year job offer in favor of continuing to receive alimony at the same rate.)
The U.S. child support system functions in much the same way as alimony, discouraging both plaintiffs and defendants from earnings. Let's look at the impact on the economy compared to a system in which both parents share only the average actual spending by parents on children. First we need to know roughly how many people are affected by these incentives.
How many Americans are profiting from child support? There is no official or comprehensive data source and courthouse records are often sealed. The U.S. Census Bureau conducts Current Population Surveys (CPS) that sample about 57,000 households at random and include data on over 139,000 people. From a survey in April 2012 they estimated that approximately 500,000 Americans are receiving more than $10,000 per year in child support and therefore potentially earning a profit over the USDA-estimated cost of child-rearing. Using UCLA Professor William Comanor's calculation of actual spending by parents (see the Methodology chapter), we would use a threshold of about $4300 per year. The Census Bureau estimated that approximately 2 million Americans are receiving more than that amount.
Are the Census estimates right? We downloaded the March 2014 data, more than 139,000 records, and did our own analysis. Considering that the U.S. Treasury says that 567,887 reported on their IRS 1040 returns that they were paying alimony, the Census CPS data under-counted alimony recipients by approximately 50 percent. Apparently people are not forthcoming when asked "Are you living off alimony?" Timothy Grall, a U.S. Census Bureau expert in this area said in an email "Historically, we've also noticed differences between what survey respondents stated they receive in child support versus what the payers report, versus what administrative records show."
Census 2014 data yield an estimate of 7.25 million Americans who are actually willing to say "I am receiving child support," inconsistent with the Federal Office of Child Support Enforcement FY2015 data indicating that officials were collecting money on behalf of 16 million children and that 55 percent of parents entitled to child support "have just one eligible child." Attorneys interviewed reported that the most lucrative child support orders were against higher-income parents and, because the payors did not fall behind, the scale is much larger than the 16 million children and $32 billion in cashflow that went through the FY2015 government collection system. If we adjust the Census data for the roughly 2X factor by which alimony is underreported, we find that 15.5 million American adults are receiving child support.
Thus when trying to determine how many Americans profit from collecting child support, we can choose from the following numbers:
All of the above numbers represent a low percentage of the total population, but that's partly because the total population includes 5-year-olds and 95-year-olds. Census 2014 data, corrected by the alimony under-reporting factor, indicate that approximately 17 percent of American women aged 30-40 collect child support. The corresponding number for men is less than 1 percent. There is a fair amount of state-by-state variation. For example, in the 50/50 jurisdiction of Arizona, only 4 percent of women aged 30-40 collect child support. It is 15 percent in Maryland and Florida; 18 percent in Massachusetts; 21 percent in Utah and Texas; 27 percent in New Hampshire; and nearly 30 percent in Tennessee. Using the same age range of 30-40, men can be found collecting child support in Delaware (4 percent) and Hawaii and Michigan (3 percent). In every other state it is 2 percent or below.
How significant are the above numbers of 500,000 to 4 million? Since the child support guideline system went into effect circa 1990, the labor force participation rate for Americans age 25-54 has fallen by 2 percentage points (Federal Reserve Bank of St. Louis). The total U.S. population in this age range is roughly 100 million. Therefore the number of Americans who earn a profit from child support is in the same range as the number of "prime age" Americans who have withdrawn from the workforce.
Other ways to look at these numbers in terms of workers, all based on Bureau of Labor Statistics data:
"Parental Responses to Child Support Obligations: Causal Evidence from Administrative Data" (Rossin-Slater and Wust; December 8, 2014 American Economics Association Conference) found significant behavioral changes in response to the modest-by-American-standards Danish child support orders, limited to between $1,330 and $5,964 per year per child (updated to 2017 exchange rates; see our International chapter). Only the very poorest American defendants are ordered to pay less than $1,330 per year and even they might be motivated to change their behavior so as not to be ordered to pay more. Thus we can say that all 15.5 million Americans who pay and all 15.5 million Americans who receive child support cash are likely to change their behavior, at least to some extent.
Using a comprehensive data set for all Danish adults, the authors found that for every 1,000DKK ($141 at 2017 exchange rates) additional that a father is supposed to pay under the Danish formula, the following behavioral changes are observed:
Why are fathers ordered to pay child support less likely to be in the labor force?
labor supply decline reflects transitions into disability insurance and discretionary early retirement programs … the decline in paternal labor force participation implies that,
at least for some fathers, child support obligations play the role of income taxes, with the substitution effect dominating the income effect.
What about mothers? In Denmark their own income does not affect their entitlement to child support revenue. Yet the economists expect them to work fewer hours:
For a mother, an increase in the obligation is a positive income shock that is independent of her own earned income. As such, we may expect an increase in maternal demand for leisure and therefore a reduction in her labor supply.
It is tough to get good data on the average child support order nationwide. More than 90 percent of Americans ordered to pay child support, however, are men and their median earnings were $47,372 per year in 2016 (Bureau of Labor Statistics). Adjusting for the Social Security Administration's correction factor, this would be an average of $72,880 per year. However, child support defendants tend to have lower-than-average incomes, even before getting sued, so we'll just use the median number as the average for defendants.
State child support formulae are not wildly different at this income level so we can take the simplest ones, e.g., Wisconsin, and multiple by 17 percent for one child and 25 percent for two. The Federal Office of Child Support Enforcement says that just over half of orders are for just one child, so we'll average these two rates. The median child support payor therefore is ordered to pay $9,948 per year. If we take a floor of $2,150 per year as the amount that a parent would pay using the "half of actual spending" rule and then assume that American men behave like the Danes, the result would be an 11 percent reduction in labor force participation. Given the difficulty of getting a child support order reduced to reflect a new, lower income, we can cut this in half to 5.5 percent. The labor force participation rate of men age 25-54, which would be typical for a child support lawsuit defendant, was 88 percent in 2016 (whitehouse.gov), a decline of almost exactly 5.5 percentage points since 1990 when the current child support system was introduced. The 5.5 percentage point decrease in labor force participation takes roughly 825,000 men out of the workforce, shrinking the economy by $39 billion.
See also "Declining Employment among Young Black Less-Educated Men: The Role of Incarceration and Child Support" (Holzer, et al. 2004; Urban Institute):
[child support] orders constitute a large tax on the earnings of low-income noncustodial fathers. Child support orders for low-income noncustodial fathers are in the range of 20–35 percent of income (Pirog et al. 1998). When combined with payroll taxes and phase-out ranges for food stamp benefits, the marginal tax rates on these men are often as high as 60–80 percent (Primus 2002).
Anecdotally we heard stories about men refusing promotions that would have required more hours of work. The average annual pay raise is roughly 3 percent. Let's assume that child support lawsuit defendants, facing a higher tax rate than other Americans, work less energetically and are less desirable to employers, e.g., due to taking days off to appear in court or pick up children at court-ordered times. We therefore expect their raises to be only 1.5 percent per year for approximately 16 years (considering the age of child at which the typical plaintiff sues and also the 18-23-year period over which child support is payable in the U.S.). The worker with 3 percent raises will be earning 1.6X after 16 years. The worker with 1.5 percent raises, just 1.27X. We prepared a spreadsheet using the following assumptions
During the period of child support payments, the defendant earns 88 percent of average. Despite raises returning to normal, the defendant earns only 79 percent of average after the child has aged out of the system. This is due to the fact that, though the growth rate is the same, the child support defendant's base income at age 41 is lower.
For the 15.5 million current child support defendants (see above), the reduction in earnings at the median (not average; see above) wage would be approximately $88 billion. Let's assume that there are a roughly equal number whose days of paying court-ordered child support are over and whose annual raises have resumed at average rates, but from a lower base. That takes another $154 billion out of the economy.
What about on the plaintiff side? Some U.S. states are like Denmark, in which the plaintiff's own earnings aren't relevant to the child support entitlement. In that situation the economists note, above, that "we may expect an increase in maternal demand for leisure and therefore a reduction in her labor supply." Some states, however, provide additional disincentives for child support recipients to work via an "income shares" model in which the recipient's entitlement goes down slightly as wage income goes up. Let's assume that the average recipient earns $1,000 per year less than if only a basic child support of $2,150 per year were received. That takes roughly $15 billion out of the economy.
As noted in the Child Support Litigation without a Marriage chapter, a foreign visitor who becomes pregnant in the U.S. can have a taxpayer-funded state office of child support enforcement obtain a court order against the American father of the foreign-resident child. Depending on the state and the father's income, the result for each child might be $20,000 to $200,000 per year leaving the U.S. economy and being spent in a foreign country.
There is no authoritative source of data regarding the number of foreign child support plaintiffs using the U.S. court system and state offices of child support enforce. Thus we are considering the shrinkage from these exports of cash at $0.
(There are presumably at least some U.S. residents who receive child support from foreigners living abroad under foreign guidelines. This would tend to boost the U.S. economy. However, as foreign child support guidelines are typically only 1/10th as lucrative as American guidelines, this would not offset the shrinkage from Americans paying foreigners.)
As noted in the Child Support Litigation without a Marriage chapter, in jurisdictions where child support is lucrative and/or unlimited, attorneys report handling the sale of abortions at a discount to the net present value of likely child support payments. Receiving $250,000 or $500,000 for an abortion should affect the recipient's labor force participation in much the same way as winning a lottery. I.e., the American who successfully sells an abortion will be less motivated to work. As these transactions are generally handled private and U.S. Census surveys don't contain questions regarding paying or receiving money for abortions, we are factoring this effect at $0.
Anecdotally we heard about a lot of behavioral changes by new romantic partners of divorce, custody, and child support litigants. For example, the new partner in a household where alimony or child support checks were rolling in would work less because all of the basic expenses were being covered by by the recipient's old romantic or sexual partner(s). The new partner in a household where alimony or child checks were being written would also work less out of fear that a plaintiff would be able to tap into the new partner's income as well, e.g., via a modification motion on the grounds that the defendant's "ability to pay" had gone up.
We currently don't have enough data to estimate these effects and therefore are considering them at $0.
As discussed in the Children, Mothers and Fathers chapter and in the Litigation chapter, the typical U.S. state's litigation-intensive primary/secondary parent divorce system inflicts a lot of financial and psychological damage on children.
Anecdotally a high percentage of children of litigated divorce end up in psychotherapy, but we don't have a good source of data for the aggregate cost so we consider it at $0.
The easiest way to estimate the harm in dollar terms is by combining studies. From "The Unexpected Legacy of Divorce: Report of a 25-Year Study" (Wallerstein and Lewis; Psychoanalytic Psychology 21:3 2004):
Only 57% of the divorce group achieved their bachelor's degree as compared with 90% in the comparison group. … Unhappy, [those who did attend college] settled for fields of study that were not their first choice, at lower ranked institutions than their parents had attended. It was at this time that one young person, echoing the emotions of many others, commented bitterly, "I paid for my parents' divorce."
In other words, there was a roughly 36 percent drop in the chance that a child would attain a bachelor's degree. Then we look at Malin Bergstrom's study in which children in 50/50 parenting arrangements had almost as high levels of well-being as children of intact families.
Let's assume that in an administrative divorce or custody system in which children were not lucrative enough to fight over, that (a) children's college funds wouldn't have been expended to pay lawyers, and (b) at least 40 percent of children (the Swedish number) of separated parents would live a 50/50 shared parenting environment. Let's assume that this reduces the penalty by half and therefore that children of separated parents were 18 percent less likely than children of intact families to attain a bachelor's degree. Census data shows that in the overall U.S. population of 25-44-year-olds, roughly 36 percent have a bachelor's degree. This includes children of divorced and never-married parents, so the likelihood that a child of an intact family will attain a bachelor's should be at least 40 percent. A child who'd been subject to the current U.S. family law system therefore would be only 26 percent likely to attain a bachelor's while a child of a system without litigation or profitable child support would be 33 percent likely. Census 2016 data show that roughly two-thirds of American children live with "two parents." However, the core survey question(S RRP) used does not distinguish between a biological parent and a stepparent. As noted in the introduction, Pew Research look at the related Census American Community Survey data and found that "Fewer than half (46%) of U.S. kids younger than 18 years of age are living in a home with two married heterosexual parents in their first marriage. This is a marked change from 1960, when 73% of children fit this description, and 1980, when 61% did." Given the trend, it seems safe to assume that roughly 25 percent of current workers have been subjects of the U.S. family law system and that about 33 percent of future workers will be.
There were 152 million Americans with jobs and therefore roughly 38 million who were affected by our litigation-oriented winner-take-all process (states mandating shared parenting is a relatively recent phenomenon). We would expect 10 million of these workers to hold a bachelor's degree while 12.5 million would have earned a degree if they'd been only half as damaged by the fallout from one parent deciding to separate/sue. That's 2.5 million additional workers with bachelor's degrees. The Bureau of Labor Statistics, March 15, 2016, shows that median earnings are $17,628 higher for workers with a bachelor's compared to an associate's degree. Compared to a system in which only half as much damage was done to children's college prospects, we have subtracted $44 billion per year from the U.S. economy.
Gross Domestic Product (GDP) is a relatively easy number to gather, but it is an imprecise measure of how well off people are. For example, the salaries of people paid to break windows and then re-glaze them, leaving a house exactly as useful as it was before, get lumped into GDP the same as the salaries of carpenters who build nice new houses. GDP includes revenue from selling OxyContin to addicts, the salaries of people working in rehab centers, and the costs of running prisons to hold convicted drug dealers. An otherwise identical society that didn't have any opioid addicts and didn't incur these costs might have a lower GDP, but nobody would say that they were worse off.
The U.S. family law system results in a portion of GDP being spent on things that don't contribute to "prosperity" or "material well-being." For example, lawyers told us that the $50+ billion in legal fees and other direct costs, discussed above, turn co-parents into permanent enemies, rendering children worse off. While some plaintiffs were delighted with the results that they achieved in family court, most litigants on both sides were unhappy with the process and the results, partly due to the fact that most of the assets being fought over were consumed by fees.
Additional misallocations come from alimony plaintiffs and defendants both having incentives to spend as much money as possible. Consider the successful Massachusetts alimony plaintiff who put some of her profits toward tearing down and rebuilding (in place) a 7,000-square-foot house. She might have preferred to use that money to buy a beach house instead, but then would have exposed herself to an argument that her "need" wasn't sufficient to justify the alimony that she was getting. Similarly, the between-jobs Florida litigant who was counseled to "rent the most expensive apartment in Tampa" to help establish a higher baseline need for alimony. What if he would have preferred a more modest apartment and a series of vacations?
The child support system distorts spending in similar ways to alimony. Plaintiffs will buy expensive outfits for children during litigation and then, once they think the judge won't be looking, it is Costco and Target for next year's wardrobe. Larger distortions occur because, in states where child support amounts depend on a child's schedule, the winner parent will often seek to keep the child(ren) away from the loser parent. They child will be parked in daycare or an after-school program, typically at the loser parent's expense, during periods of time when the loser parent is available and both child and loser parent would be happier together. The cost of daycare goes into GDP, but the parent and child would have been happier if the money had been spent on something else, given to charity, or burnt.
Winner-take-all tends to result in distorted use of real estate, one of the most expensive items in the American economy. "When I'm called back in 5 or 10 years later on a modification, Mom is living with one part-time kid in a 5,000-square-foot house," said an attorney. "Dad is in an apartment with the new wife, their kids, and the object of the child support lawsuit part-time. Of course if I'm plaintiff-side I argue that Dad's lifestyle is more luxurious and Mom needs additional child support to keep up." An important argument for continued cashflow is that the 2-year-old child needs the stability of keeping the same house for 16 more years. Thus nobody ever stands up in court to say "It would make more sense for the 5,000-square-foot house to be exchanged for two 2,500-square-foot houses."
The winner-take-all custody system that is prevalent in most U.S. states often results in a misallocation of children. Attorneys told us that plaintiffs who were able to get millions of dollars via property division, e.g., because the defendant's money was earned during the marriage in a 50/50 state, tended to be uninterested winning the primary parent war. "Plaintiffs go to court to get cash and if they have to take the kids to get the cash they'll do it," was how one attorney phrased it. Oftentimes a motivation for the divorce was a desire to have sex with new and/or younger partners and enjoy adult activities that can be funded with cash. Those goals were impeded by having kids around more than 50 percent of the time, but in winner-take-all jurisdictions the kids were the only way to get the house and enduring cashflow. Plaintiff, defendant, and children would all have been better with a 50/50 schedule but a court couldn't award that and also the sought-after house and cash.
We don't know of a good data set for these misallocations and we aren't sure that they reduce headline GDP even if they reduce American prosperity. Thus we account for them at $0.
In 2016 one of us wrote up a weblog entry:
A 48-year-old friend with an engineering degree was recently laid off, along with 39 co-workers. Let’s look at the incentives that the government gives him.
He could sue his former employer for age discrimination (EEOC says that the law “forbids age discrimination against people who are age 40 or older”). However, the employer seems to have recognized this, for he reports that “I just signed a release in exchange for six months pay and six months health insurance.”
He could look for another job immediately, but points out that “They pay you $22,000 to not work,” and this entitlement (unemployment benefits) will be lost if he gets a job.
His wife has a high-paying job. If he returns to work and she sues him for divorce, she is on track to win the kids (97 percent of women become the winner parent in Massachusetts, which disfavors 50/50 arrangements), profitable child support, the house, and possibly alimony (if his new job pays better than hers). Suppose that instead he decides that 48 is his retirement age and does not obtain a new job. In the event of a divorce lawsuit, because he has been at home with the kids he improves his chances of getting primary or at least shared custody. This plus his lack of income reduces his exposure to a child support lawsuit. He may retain the house (he needs somewhere to live and doesn’t have any money). He can tap the wife for alimony.
Here's how it adds up and compares to the $18.675 trillion U.S. GDP:
Item | Amount ($B) |
difference in number of divorces and children in single-parent households compared to Europe | 0 |
return on $2 trillion in legal fees, etc. | 40 |
cost of child abuse investigations instigated by custody and child support plaintiffs | 0 |
reduction in work by married women from no-fault and 50/50 | 194 |
reduction in work by married men (lower income partners) from no-fault and 50/50 | 0 |
reduction in work by adults during litigation | 29 |
reduction in work, post-trial, due to alimony | 10 |
withdrawals from labor force, e.g., to disability or retirement, following child support order | 39 |
reduction in earnings by child support defendants due to lower annual raises | 154 |
reduction in work due to imprisonment for failure to pay child support | 1 |
reduction in work due to receiving child support | 15 |
exports in cash to foreigners receiving U.S. child support | 0 |
reduction in work by Americans who have sold abortions | 0 |
reduction in work by stepparents, new partners, etc. due to alimony and child support | 0 |
reduction in earnings by children of family court lawsuits due to lack of bachelor's degrees | 44 |
psychological damage to children from litigation | 0 |
reduction in prosperity, but not GDP, due to misallocation of resources | 0 |
Total | 526 |
U.S. GDP 2016 | 18675 |
Shrinkage due to family law system | 2.82% |
The economy would be roughly 3 percent larger if we were to replace our litigation-oriented winner-take-all (in most states) divorce, custody, and child support system with a system in which (a) divorce were available via an administrative process, (b) children did not have a cash value worth fighting for (child support limited to likely child-related expenses), and/or (c) there was a 50/50 shared parenting presumption.
What does this mean in concrete terms? If we gave up on our litigation- and profit-oriented family law system we could be just as wealthy, and live just as comfortably, while taking an extra 1.5 weeks per year of vacation. Alternatively, if the additional $526 billion in GDP were absorbed by the federal government as additional tax revenue, the federal budget deficit for FY2016 would be eliminated.
How do people with advanced degrees respond to guaranteed cash? "Why Do We Tenure? Analysis of a Long Standing Risk-Based Explanation" (Brogaard, et al., 2016; SSRN):
Using a sample of all academics that pass through top 50 economics and finance departments between 1996 and 2014, … We attribute the decline in the quantity and quality of research output post-tenure to a reduction in effort incentives, … we consider another publicly observable event around which there is likely to be a significant change in incentives: the promotion to Full Professor. Full Professorship is often associated with a substantial increase in pay, as well as rights to contribute to university decision-making. … We find that the number of publications per year rises until two years prior to a professor attaining the rank of Full Professor, and falls steadily over the next 12 years. The same pattern is apparent for the number of home runs [heavily cited and therefore important papers] per year. The percentage decrease in each variable is nearly identical to the post-tenure period, and doubly significant.
"Taxation and Labor Supply of Married Women: The Tax Reform Act of 1986 as a Natural Experiment" (Eissa 1995; NBER) covers women who could rely on a man's income for at least some living expenses. The paper describes married women, but the same reliance is possible for a woman collecting alimony or child support. The author calculated an elasticity of supply of roughly 0.7, i.e., if the after-tax wage went up 10 percent the amount of labor supplied would go up by 7 percent. Thus a plaintiff receiving alimony of half the difference between her earnings and her defendant's would suffer a loss of 50 cents for each additional dollar earned. Assuming a 40 percent federal and state combined income tax rate, Her net pay for an additional $1 of gross pay would be just 30 cents (60% times 50 cents; she would effectively be in the 80-percent tax bracket) rather than the 60 cents she would keep in the no-alimony case. Her expected response to this 50-cent reduction in net pay would be to reduce her working hours by 35 percent compared to if she did not have the alimony flowing. This kind of scaling back from full-time to part-time work is consistent with anecdotal reports from attorneys.
The most comprehensive study of how Americans respond to financial incentives and whether this can produce an overall contraction of our economy is The Redistribution Recession: How Labor Market Distortions Contracted the Economy (2012; Oxford University Press) by Casey Mulligan, a professor of economics at University of Chicago (see also this review). The book looks at the expansion of government-provided means-tested welfare programs that began in 2009. In other words, benefits that Americans could get that were conditioned on not earning too much. These included at least the following: unemployment benefits, food stamps, mortgage forgiveness, debt forgiveness (other than mortgage), and health insurance subsidies. Professor Mulligan summarizes previous research:
Decades of empirical economic research show that the reward to working, as determined by the safety net and other factors, affects how many people work and how many hours they work. To name a small fraction of the many studies: Hoynes and Schanzenbach (2012) show how potential participants stopped working or reduced their work hours when the food stamp program was introduced. Studies of unemployment insurance (see Appendix 4.2) find that program rules have a statistically significant effect on how many people are employed, and how long unemployment lasts. Yelowitz’s research (2000) shows how a number of single mothers found employment exactly when, and where, state-level Medicaid reforms increased their reward from working. Gruber and Wise (1999) and collaborators show how the safety net for the elderly results in less employment among elderly people. Autor and Duggan (2006) and the Congressional Budget Office (2010) explain how the number of disabled people who switch from work to employment-tested disability subsidies depends on the amount of the subsidy relative to the earnings from work. Murphy and Topel (1997) show how poor wage growth among less-skilled men helps explain their declining employment rates during the 1970s and 1980s. Jacob and Ludwig (2012) show that means-tested housing assistance reduces labor force participation and earnings among able-bodied working-age adults.
The percentage of Americans who work declined from 66 percent in 2009 to around 63 percent in 2013, where it has stayed. This cannot be attributed to an aging population because labor force participation rate increased over the same period in Singapore, whose citizens are older, on average. The reduction corresponds to 7 million Americans withdrawing from the world of work. Mulligan shows that it was those Americans who qualified for various forms of means-tested government welfare who were most likely to quit working. This was because for every hour of waged labor that they worked there was a reduction in welfare benefits received. As with people receiving alimony or child support under a formula that includes the recipient's income, this reduction functions like a higher income tax rate: workers get to keep and spend a smaller fraction of their gross earnings. Some people who qualified for mortgage forgiveness ended up with a tax rate higher than 100 percent, i.e., they would have a lower spending power for every additional hour of waged labor due to the resultant reduction in benefits.
Our attachment to litigation and winner-take-all as the solution to family break-up brings down our GDP by a substantial, though not crippling, amount. Misallocation of resources as a result of the family law system's incentives mean that, compared to other countries, we are not as well off as our headline GDP would suggest.