The Census Bureau recently released results from its alternative measure of poverty, which uses a wider range of factors than the official metric to account for people’s living expenses and the money or other resources they have to pay them. Most of the reaction to the alternative measure focused on its overall impact on poverty rates, but the new metric also is notable for breaking with tradition in the way it assumes cohabiting couples share their money.

Under the traditional measure of poverty, unmarried couples who live together are counted as separate units.  Under the alternative metric, called the Supplemental Poverty Measure, the assumption is that cohabiting couples pool their funds and share expenses just as married couples do. The result: A lower share of cohabiting couples is considered poor under the alternative metric than under the official measure.

This is because living as a couple is more economically efficient than living as two single people. A couple does not need twice the income of two single people. For example, two single roommates might prefer a two-bedroom apartment but a couple can manage with a one-bedroom apartment.

Cohabiting partners are not the only people who become part of family economic units under the alternative metric. The Census Bureau also decided to include foster children and other unrelated children living with families, who are not part of family economic units under the official measure. In 2010, the bureau’s analysis indicates that 33% of the people living in these new family economic units were considered poor using the official measure, compared with 21% under the alternative measure.

The bureau did not publish statistics that only compare cohabiting partners under the official and alternative measures, but cohabiters appear to be driving the difference in poverty rates for these new family units.  Other Census Bureau research indicates that most of the units that change under the alternative measure by adding at least one person do so because of the presence of a cohabiting partner.

History of Poverty Measure

The Census Bureau has been developing an alternative measure of poverty for more than a decade in an attempt to produce a more sophisticated version of the metric that has been in use since the 1960s. The alternative metric, sometimes described in census publications as an “experimental” or “research” measure, will not replace the official measure, at least for now. For the population as a whole, the alternative metric produced a higher poverty rate in 2010 (16%) than the official measure (15.2%).

In setting poverty thresholds, the alternative measure accounts for a wider range of expenses than the official measure, as well as variation in expenses by age, geography and other factors. In determining the resources people have to meet those expenses, the alternative measure includes not just cash income, but also non-cash government benefits such as food stamps and tax credits; as noted, it also broadens the definition of the family economic unit that shares resources.

The notion of including cohabiting couples in a widened definition of a family economic unit was recommended in 1995 by a National Academy of Sciences panel. The NAS report was the foundation for much of the later work that led to the Supplemental Poverty Measure, and its recommendation on including cohabiters in a broadened family unit was echoed in a 2010 report by an interagency technical working group that provided a blueprint for the new census poverty metric. The NAS report recommended more research into how unmarried couples manage their money.

“Families should be defined to include cohabiting couples,” the report said. “Such couples typically pool resources, and many of them exhibit considerable stability, so that it seems to make sense to treat them like married-couple families for purposes of poverty measurement.”

Cohabitation has become a much more common living arrangement since the original census poverty definition was established in the 1960s. The National Academy of Sciences report quoted an estimate of 523,000 couples in 1970. In 2011, according to the most recent Census Bureau report on families and living arrangements, an estimated 7.6 million unmarried couples lived together, including many who are raising children. Still, long-term cohabitation is rare; most couples break up or get married within three years.

What Does Research Say?

Research generally has found that cohabiting couples are less likely than married couples to pool all of their money, but there is evidence that many, if not most, fit the definition of “consumer units” because they share major expenses.

One study of couples with young children found that 73% of married couples and 52% of cohabiting couples combine all their money; the share rises to 75% of cohabiters (and 83% of married couples) by including those who keep their money separate but split household and child expenses 50-50. The study cited earlier work concluding that couples who have children together may be more likely to share their money than couples who do not.

Some researchers argue that couples who expect their relationship to last are more likely to pool resources. Along these lines, a Pew Research Center survey in 2010 found that among those currently living with a partner, 53% say they see it as a step towards marriage.

The alternative measure’s poverty thresholds are based on data from the Bureau of Labor Statistics Consumer Expenditure Survey on out-of-pocket spending on food, clothing, shelter and utilities. The Bureau of Labor Statistics survey assumes that unmarried couples are a single “consumer unit” that splits living expenses. A consumer unit is defined as people sharing at least two out of three major outlays–food, housing or other living expenses. Census Bureau officials cited the goal of consistency with BLS practice in their justification for assuming that unmarried couples share resources as married couples do.

Future Research

The Census Bureau plans to do more work to refine the alternative metric, as resources permit. In releasing results of its alternative poverty metric for 2010, Census Bureau officials continued to acknowledge that more research is needed to parse how cohabiting couples share income and expenses.

Resource sharing by cohabiting couples is one of the “unresolved issues for future consideration” that is included in a recent paper by a Census Bureau researcher asserting that there may be variation in the extent to which unmarried couples share money. The paper suggests, for example, that couples who recently moved in together may not be as likely to pool their resources as would longer-term couples.

The assumption that couples share resources is based on the observation that they have “fairly stable relationships,” according to the paper by Ashley J. Provencher, but “this assumption may be only be valid for particular couples.”