Half Changed World

RebelDad asked today if anyone could find the Census data that journalists are using to say that there were 147,000 SAHDs in 2004, up from 98,000 in 2003.  Of course, I took that as a challenge, and dug it up.    It’s this table, cell I7.

However, the part of this table that caught my attention was rows 27-38, which have income data for different types of married couple families with children under 15.  This is the first hard data I’ve seen on the subject.  I have seen lots of conjectures, including Stephanie Coontz’s statement (in Marriage) that the only two segments of the population in which male breadwinner families predominate are the bottom 25 percent of the income distribution and the top 5 percent, and Nathan Newman’s provocative suggestion that SAHMs are "luxury goods."

So what do the data say? First, that married two-parent families are overall fairly well off — over 40% have incomes over $75,000 a year, and only 7.3% are poor.  Second, at the level of detail the Census provides, such families with SAHMs are generally worse off have lower cash incomes than average — only about 31% have incomes over $75,000, and 12.2% are poor.

The income categories most likely to have a SAHM are those with annual family incomes between $10,000 and $25,000.  The women in these households are likely to have low potential earnings, and between child care costs and the phaseout of some tax breaks, it probably doesn’t pay very much for them to work.  I would also guess that many of them are from cultures that highly value at-home mothering.  At the other end of the spectrum, married couple families with incomes over $100,000 are slightly more likely than those with incomes between $75,000 and $100,000 to have a SAHM.

Turning to the families with SAHDs, I was surprised to see that they were generally worse off had lower cash incomes than families with SAHMs.  Less than 22% have family incomes above $75,000, and 15.6% were poor.  This presumably reflects the overall lower earnings of women compared to men. But I would have guessed that the influence of selection would have pushed the average family incomes up.

Revised 10-20-2005 to reflect Parke’s suggestion.

On Monday, Philip Klinkner at PolySigh posted a graph of labor force participation rates by race and gender for the past 50 years.

I was generally aware of the overall trends, but was surprised at how low the labor force participation rates were for black women in the 1950s.  Yes, they were a lot higher than for white women, but I’ve often heard comments to the effect of "black women have always worked; they didn’t have the privilege of being stay-at-home mothers even in the 1950s.

That’s certainly the impression that I got from reading The Street.  We returned the book to the library yesterday, so I can’t post a direct quote, but Lutie Johnson (and presumably Ann Petry) lays the blame for most of the ills of black people on the fact that whites wouldn’t give black men jobs, but would hire the women as domestic servants.  So the men felt emasculated and sought to prove themselves by fighting and sleeping around.  And the children were left unsupervised in dangerous neighborhoods, and got sucked in by the attractive menace of the Street.

Petry also suggests that at least some of the huge increase in single-mother households over the past 50 years is illusory.  Almost all of the women in The Street are technically still married, but living on their own or with men other than their husbands.  They’re only married because they can’t afford to jump through all the hoops required to get a legal divorce at the time.

Jen (who often comments here) asked me if I knew how the Boxing Day tsunami compares to other historical natural disasters in terms of the number of deaths.  Here’s the most comprehensive single list I was able to find, going as far back as an earthquake in 1201 AD.  Limiting it to the past 40 years or so, it looks like it falls third, behind only the 1970 cyclone that hit Bangladesh and the 1976 earthquake that hit China.  Wikipedia also has an entry on this, broken out by types of event.

For a different perspective, Nicholas Kristof points out in the New York Times that more people die every month of AIDS (240,000) and malaria (165,000) than died in the tsunami, and almost as many die of diarrhea (140,000).  But it’s hard to stay focused on unglamorous, persistent problems, and so kids die for lack of a few cents worth of oral rehydration salts, or bleach to purify their water.

I’d encourage everyone who is giving so generously to the victims of the tsunami to also consider giving unrestricted funds to an organization like CARE or Doctors without Borders to fight chronic disease and poverty. 

Moreover, a recent report from the UN Development Programme points out that most of the victims of "natural disasters" are also victims of poverty.  Poverty means that people live in places prone to flooding, earthquakes and mudslides, live in flimsier buildings, and so forth.  I’m not sure there’s any building code on earth that could have made much difference against the tsunami, but there’s a reason that similar size earthquakes kill tens of thousands of people in Iran, but only a few in California.

I finally got a chance to look at the new Census report on America’s Families and Living Arrangements.  The data on SAHMs and SAHDs is all the same material that I discussed last month when I discovered the detailed tables on their website, so instead I’m going to talk about the big drop in the fraction of all households that contain children.

According to the Census, in 2003, families with children made up just 32 percent of households, down from 45 percent in 1970.  This is a  big shift, driven by a bunch of factors all working in the same direction:

• More people don’t have children at all.  In 2002, almost 18 percent of women ages 40-44 had never had kids, up from about 10 percent in 1976.  (About 0.3 percent of women have their first child in their 40s.)
• More people delay childbearing (so they’re childless for longer)
• People are having fewer kids, even those from cultures that have traditionally valued large families. (A family with 1 child will have a child under 18 for exactly 18 years, while a family with 3 children, 3 years apart, will have one for 24 years.) 
• People live longer after they’re done having kids.
• Affluence and mobility both result in more single people — both young adults and the elderly — living on their own rather than with their families.  In 2003, over a quarter of all households were people living on their own.  Less than 10 percent had five or more people, down from 20.9 percent in 1970.

I think these trends make it harder to convince businesses that they have to adopt family-friendly policies.  But, as I’ve said before, it strikes me as utterly insane that in a potential working life of 50 or more years, it’s not feasible to take 2 or 3 off to focus on childrearing.

Since I often post about bad news, I thought I’d share this encouraging piece of news that came across my desk today.  The CDC reported that births to very young teen mothers (ages 10-14) are down to 0.7 live births per 1,000 girls, half the rate they were in 1990, and the lowest rate since 1946.  This is good news on all levels — girls this age aren’t physically ready to give birth (it’s dangerous both to the mother and the baby) and they’re not emotionally ready to be parents. 

The interesting thing is that no one really knows why the birth rate has dropped in the last decade — either for this group of very young teens, or for teens in general (the birth rate for girls 15-19 is 41.7 per 1,000, down from 61.8 in 2000).  The story is a complicated one, involving both better birth control (especially long-term hormonal approaches like depo-provera and norplant) and reduced sexual activity.

This week, I was part of a small group that got together at the National Partnership for Women and Families to talk about their work to expand the Family and Medical Leave Act and to extend paid sick and family leave to more workers.  It was a good conversation, and it reminded me that I want to look into California’s paid family leave program (which is funded through an employee tax, not by employers) in more detail. 

Someone asked the question, what fraction of families don’t have a stay-at-home parent.  I thought I had addressed this question in my discussion of the trends in women’s labor force participation (see Who’s "opting out"?)  but when I checked, I discovered I hadn’t.

The latest figure i could find was for 2001, when 68 percent of children had both parents or the only resident parent in the labor force, up from 59 percent in 1985.  Interestingly, in 1985 this was true for 51 percent of children under 6, and 63 percent of children 6-17.  By 2001, the gap had narrowed significantly, to 66 percent of children under 6 and 70 percent of older children.  (These figures come from Table ES.3.1.A in Trends in the Well-Being of America’s Children & Youth: 2003, a handy reference book put out by the fine folks for whom I work.)

The Work-Family program at the New America Foundation likes to refer to these families as "juggler families," which is a nice catchy phrase.  Their talk of how such families have "replaced the traditional family of the breadwinner and the homemaker" is a little misleading, however.  It conveys the impression that all of the working parents in those juggler families are fully committed to the labor force.  But the statistics include a significant number of parents who consider themselves primarily caregivers, but also have some paid work.  I’ve never seen a comparable figure broken out by hours of work — if any of my readers has, please let me know.

The National Partnership does a good job of pointing out that paid leave doesn’t only benefit families without an at-home parent; in fact, families with only one earner are more vulnerable if illness causes that earner to miss work and lose pay.  Unlike increased funding for child care, many social conservatives support paid family leave; however, the business lobby bitterly opposes it.

As described yesterday, I searched all over the internet to try to substantiate the claim that the number of stay-at-home moms (SAHMs) has increased by 15 percent in less than 10 years.

And finally, I found it: Table SHP-1: Parents and Children in Stay-At-Home Parent Family Groups: 1994 to Present. In fact, this table reports that the number of stay-at-home mothers increased by over 19 percent between 1994 and 2003, from 4.5 million to 5.4 million.

I hope that some of you are saying "but…" right now. Doesn’t 5.4 million sound awfully low? For perspective, there were over 93 million women between the ages of 16 and 65. How can this be right? The catch is that Census is using a very narrow definition of what constitutes a stay-at-home parent: you have to be a married parent of a child under 15, out of the labor force for an entire year, say that the reason you’re not working is to care for "home and family" and your spouse has to be in the labor force for the entire year. RebelDad did an excellent job least year of explaining the drawbacks of this definition, so I won’t repeat them.

Even though this definition isn’t perfect, this is the first longitudinal data I’ve seen on the number of stay-at-home dads (SAHDs), applying the same definition to a consistent data series over time. They found 98,000 SAHDs (using this narrow definition) in 2003, down from a high of 106,00 in 2002, but up from just 49,000 in 1996. However, because the number of SAHDs is relatively, there’s a lot of "noise" in the figures — I asked the Census bureau, and they said that the drop from 2002 to 2003 isn’t statistically significant. One way that statisticians deal with this kind of noise is to pool the findings from several years. So I compared the average number of SAHDs for 1994-1996 to the average number for 2001-2003, which suggests a whopping 50.8 percent increase. Just comparing 1994 to 2003
produces a 28.9 percent increase, also quite impressive.

One way to get a sense of the limitations of the definition is to compare this series to a similar one that just looks at married couples, and whether one, both or neither is in the labor force. This comparison indicates that in 70 percent of the married couples where only the husband was in the labor force, the wife met the definition of "stay-at-home mother." But in the married couples where only the wife was in the labor force, only about 10 percent of the husbands met the definition of "stay-at-home father."

One reason for the gap is the requirement that only spouses of year-round workers can count as "at-home parents." I’m not certain, but I think that taking maternity leave is considered as being "not in the labor force." If that’s the case, my husband wouldn’t have counted as being an at home dad last year, because I was on maternity leave for 12 weeks. Adding back in the parents who meet all of the other requirements to be an at home parent would increase the reported number of SAHDs by 60 percent, to 157,000, but the reported number of SAHMs only by 12 percent, to 6 million. I also think men are less likely to say that the reason they’re not working is to "care for family and spouse."

Last week’s 60 Minutes story on (Women) Staying at Home, included the statement that "Census bureau statistics show a 15 percent increase in the number of stay-at-home moms in less than 10 years." I hadn’t seen any hard numbers supporting the claim that there’s been a big increase in the number of women staying home, so I set off in search of this statistic.

First stop was the Bureau of Labor Statistic’s useful databook on Women in the Labor Force. I soon learned that the labor force participation rate (meaning the fraction of the population employed, or looking for work) for women was in 2002 was 59.6 percent, a slight (0.7 percent) decline from the 1990 peak of 60.0 percent, but higher than any year between 1970 (when it was 43.3 percent) 1996.

But, of course, women with children are only a small fraction of all adult women, so there’d have to be a pretty big drop in the number of working mothers for it to show up in the overall labor force participation rate. So I kept looking.

The same databook reports that the labor force participation rate for all women with children under age 18 was 72.2 percent in 2002, down from a peak of 72.9 percent in 2000.* Looking only at women with children under age 3, the rate is 60.5 percent, down from a peak of 62.2 percent in 1998. We’re still looking at changes in the 1-3 percent range, nothing earthshattering. So I kept looking.

Next, I found an interesting article from the Monthly Labor Review, a journal put out by the Bureau of Labor Statistics with the headline: "Are women leaving the labor force?" It includes this quote from Barron’s, claiming that "In just the past two years, a quiet counterrevolution has begun…" The most interesting thing about this article, however, is that it’s from July 1994, shortly before women’s labor force participation hit new all-time highs. (For the record, the author, Howard Hayghe, correctly concluded "it is too early to proclaim that the trend of increasing labor force participation rates of women has been halted.") So, is the claim of the retreat from the workforce just hype? I kept looking.

Moving over to the Census bureau’s bi-annual report on the Fertility of American Women, which turns out to be the source for this Womens eNews story from last year. This report says that of women who had a child in the last year, 54.6 percent were in the labor force in 2002, down from a peak of 58.7 percent in 1998. That’s a 7 percent decline — albeit from an extremely high point. Looking at the breakout by education, it looks like the biggest percentage decrease is for women without a high school degree, and the smallest decrease is for women with a high school degree, but no college.

So what’s going on? I think there are two different stories, at different ends of the labor market. At one end, is the story about the stars and planets aligning in the late 1990s to get more low-income mothers into the labor force than ever before**: welfare reform removed an alternative to working, increased federal and state support for child care made working more possible, the Earned Income Tax Credit made working more profitable, and the strong economy made jobs available. The economy isn’t so strong these days, and it shouldn’t be a surprise that fewer poor mothers are working. (The big unanswered question is what are they living on, because it’s not welfare, but that’s a topic for another day.)

The second story is about well-off well-educated women who have the choice whether or not to work because they have other sources of income, most often husbands. For the last 30 years, this group has been more likely to work, not less, than other women, because they have access to the most interesting, renumerative, and flexible jobs. And it does look like there’s a small increase in the number who have chosen not to work in the past few years. Whether this is a blip in the trends (as the apparent decline in the early 1990s was), possibly caused by the weak economy, or is the start of a real change (post 9/11 reprioritizing?), I have no idea. And no one else does, either, no matter what they tell you.

Coming tomorrow: The plot thickens: another source of data on stay-at-home moms is found. And stay-at-home dads, too!

Footnotes:

* Yes, mothers have a higher labor force participation rate than all adult women; it’s because "all adult women" includes senior citizens.

** I am aware that not all never-married mothers are low-income, but this was the closest graph I could find to what I wanted to show.

The Annie E. Casey Foundation issued a new report this week, called Working Hard, Falling Short. It’s about low-income working families, defined as those earning less than twice the poverty line, or about $36,800 for a family of four. About 1/4 of all working families with children fall into this category, accounting for about 1/3 of the children in working families. Those are pretty grim figures.

Or are they?

Last month’s Washington Post article on the decline of the middle class reported that about 41 percent of all American households (including those without children) make less than $35,000 a year, down from 54 percent in 1967. (All these figures are in inflation adjusted dollars.)

So, is it terrible that so many families are low-income, or encouraging that so few are?

One thing to note is that a significant portion of the gains of the last 30-40 years are due to the increase in women’s labor force participation. A lot of families worked their way out of low-income status by having multiple wage earners contributing to the family income. But that results in child care costs, which come out of the discretionary income. And it doesn’t leave families with as much flexibility to respond to family crises. People feel like they’re working harder just to keep up — and there’s not a whole lot of room in the day for more work.

Are low-wage jobs “worse” than they were 30 years ago? They’re more likely to be in the service sector, less likely to be in manufacturing. They’re less likely to be unionized. They’re less likely to be 9-5 jobs, more likely to be evenings or weekends, keeping the 24-hour economy running. Or they may be variable shift, with the employer deciding how to staff based on the previous week’s sales. I don’t know whether such benefits as employer-provided health insurance and paid leave have gone up or down over that time period — anyone have a data source?

One hint about the problem comes from the title: “Working Hard, Falling Short.” Falling short compared to what? Precisely because average incomes have increased so much, the same amount of money feels like less. And, as Tyagi and Warren argue, it’s not just a matter of envy, but that low-income families have to compete with better off ones for houses in safe neighborhoods, for access to quality schools, and so on.

Following on the heels of yesterday’s review of The Two-Income Trap is this article from today’s New York Times, about the disastrous consequences of Texas’ Robin Hood system to equalize school finances.

The basic problem (which the Robin Hood system was designed to fix) is that most public school funding in the United States comes from local property taxes. Rich areas have much more of a valuable property base, and so are able to put more money into their schools. As a recent study from the Education Trust showed, nationwide, there’s a gap in school funding of more than $1,000 per student between the highest- and lowest-poverty school districts.

The Robin Hood system was implemented after a group of school districts sued the state, claiming the old system of school funding was unconstitutional, and won. It attempts to overcome this inequity by taking money from the wealthiest school districts — those with property value above a threshold — and distributes it to the poorest. The New York Times article, based on an analysis by Harvard economists Caroline Hoxby and Ilyana Kuziemko, explains what went wrong. Hoxby and Kuziemko argue that the problem is that part of what determines the value of property is the quality of the schools. As the quality of the schools went down (for a given level of taxation), the property values also dropped, so the state’s share of taxes went down. In order to provide the level of funding they had promised to the poor districts, the state had to lower the threshold and therefore affect more districts. And the spiral continued. They conclude that the Robin Hood system only reduced inequality between school districts by about $500 per student, but destroyed thousands of dollars per student in property value.

What are the implications of this study?

First, it both supports Warren and Tyagi’s argument that access to quality schools is a significant portion of what drives up property values, and emphasizes how unlikely it is that their proposal to delink residence from access to schools could ever be implemented. The disruption, as home values in previously “good school districts” plummeted, would be monumental. I can’t imagine the circumstances under which such legislation would pass.

Second, it’s a strong rebuttal to those who argue that how much money a school district has doesn’t matter. Homebuyers in Texas certainly valued access to schools in rich districts less when their funding dropped, even though nothing else about the schools had changed. Thus, it could support an argument for either a minimum floor on school spending per student or for some version of equalization or both. Hoxby and Kuziemko are quite clear that they don’t want their paper used as an argument against equalization in general.

Third, as Hoxby and Kuziemko argue, it suggests that lawyers shouldn’t be allowed to design tax systems. They claim that the Robin Hood system was designed solely in order to avoid a legal ban on a statewide property tax, and that it does pretty much everything wrong from the standpoint of creating an efficient tax system.

What none of these studies do is tell you what to do if you’re a parent and a good liberal who believes in public schools and can’t really afford private school anyway and enjoys living in a diverse urban (or semi-urban) area but isn’t quite sure that the schools where you live are as good as you want them to be and doesn’t want to sacrifice your kid to an abstract principle. (Here’s a link to a friend of mine’s much funnier and slightly obscene rant on the same subject. Don’t click if cussing offends you.)

Tomorrow night we’re off to a pizza party being held for parents of preschoolers who are zoned for our local elementary school (which is literally 3 blocks from our house). Part of the goal is to discuss how we can help the school, but another part is just to hold each other’s hands and convince ourselves that we’re making the right choice.