PRB

More than 55 million Americans are age 65 or older, according to the Census Bureau’s 2020 population estimates. One-fourth of these older Americans live in one of three states: California, Florida, and Texas. Seven other states—Georgia, Illinois, Michigan, New York, North Carolina, Ohio, and Pennsylvania—account for roughly another quarter of the 65+ population.

These 10 states are also the most populous and include over half of the total U.S. population. Sparsely populated states such as Alaska, North Dakota, Wyoming, and Vermont also have very small older adult populations—less than 130,000 each in 2020.

But the states with the most adults age 65 or older do not necessarily have the oldest population age profiles. California is a relatively young state even though it has the largest number of older residents: Only 15% of the state’s total population was age 65 or older in 2020. In contrast, Maine’s relatively small number of older adults represent 22% of its total population, the highest share of older residents in any state.

States Ranked by Percent of Population Age 65 or Older, 2020

While southern states are regarded as retirement magnets, partly due to their warmer weather and tax benefits for seniors, states in the Northeast and Midwest have among the largest shares of older adults. What’s driving these regional patterns?

Migration, both internal and international, has a large impact on the distribution of older adults. States that have attracted older retirees, such as Arizona, Florida, New Mexico, and South Carolina, have larger proportions of older residents. Many states in the Midwest and Northeast also have large shares of older adults, but for different reasons. As young adults in these states have moved south and west looking for educational and job opportunities, the older population has been left to age in place. In contrast, Texas has been a popular destination for state-to-state and international migrants, which has kept its population relatively young.  Austin-Round Rock-Georgetown was the second-fastest growing metropolitan area in the country between 2010 and 2020, trailing only The Villages in Florida.

The share of older adults will continue to increase as more members of the large baby boom cohort reach retirement age. By 2030, 26 states are projected to have age profiles similar to those of Florida and Maine today, with at least 20% of their residents age 65 or older. This demographic shift has implications for many federal and state programs that support older adults. As more Americans become eligible for federal entitlement programs like Medicare and Social Security, spending reductions and tax increases may be inevitable.

Excerpted from PRB’s Population Bulletin, “Elderly Americans,” by Christine L. Himes, and updated in 2021.

Coronavirus Pandemic Temporarily Shortens Average U.S. Lifespan by About a Year

While the pandemic is shortening the average U.S. lifespan—temporarily—its effects will be felt most heavily by Black Americans, whose mortality rate in ordinary years is higher than the rate for white Americans during the pandemic. Each coronavirus-related death will likely impact about nine close family members.

These impacts are among the findings of new research supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) that aims to put the pandemic’s magnitude in context and inform responses.

Extreme Racial Inequality in COVID-19 Deaths Layers on Top of Existing Stark Disparities

Black Americans experience a higher mortality rate every year than white Americans are experiencing during the coronavirus pandemic, finds Elizabeth Wrigley-Field of the University of Minnesota.1 Her analysis focuses on death rates and compares the scale of this pandemic to racial inequality, which she calls “another U.S. catastrophe.”

Using demographic models, Wrigley-Field estimates how many deaths of white Americans would be needed to raise the white age-adjusted mortality rate to the best-ever (lowest) Black age-adjusted rate.

At least 400,000 excess deaths of white Americans—deaths above and beyond the number expected in a non-pandemic year—would be needed to reach the best mortality rate ever recorded for Black Americans, which occurred in 2014, she finds.

Black Americans’ age-adjusted, confirmed COVID-19 deaths are more than 2.5 times higher than that of white Americans, she reports.2

Social factors rather than innate vulnerabilities drive these mortality differences: Mounting research suggests these stark disparities are driven by differences in exposure to the coronavirus. In particular, Black Americans are overrepresented in service jobs with high public contact and are particularly overrepresented among low-paid workers who may lack the power to demand adequate protection.3

For white mortality to reach levels that Black Americans experience outside of the pandemic, excess mortality in 2020 for white Americans would need to increase by 5.7 times the level of cumulative COVID-19 mortality reached in July 2020 (when the research findings were published), Wrigley-Field reports.

Final analysis of 2020 is likely to reveal “a deadly pandemic causing a spike in mortality for whites that nevertheless remains lower than the mortality Blacks experience routinely, outside of any pandemic,” she suggests.

This disparity in mortality rates has an impact on life expectancy during the pandemic as well. For white Americans, life expectancy in 2020 will remain higher than life expectancy for Black Americans has ever been unless nearly 700,000 excess white deaths occur, Wrigley-Field finds.4

“If Black disadvantage operates every year on the scale of whites’ experience of COVID-19, then so too should the tools we deploy to fight it,” she argues. “Our imagination should not be limited by how accustomed the United States is to profound racial inequality.”

COVID-19 Expected to Shorten the Average U.S. Lifespan in 2020

With the U.S. population as a whole experiencing nearly 350,000 COVID-19 deaths in 2020 and more to come in 2021, life expectancy may appear to be plummeting.5

But in estimating the magnitude of the pandemic, demographers at the University of California, Berkeley have found that COVID-19 is likely to shorten the average U.S. lifespan in 2020 by about a year.6

In July 2020, demographers Ronald Lee and Joshua Goldstein calculated the consequences of U.S. lives lost to COVID-19 that year in order to put COVID-19 mortality rates into historic, demographic, and economic perspective. They used two scenarios: One based on a projection of 1 million deaths for the year, the other on 250,000 deaths, which is closer to the current estimate of 345,700 deaths by Johns Hopkins University.7

One million deaths in 2020 would cut about three years off the average U.S. life expectancy, they conclude, while 250,000 deaths would reduce lifespans by about 10 months.

That said, without the societal efforts implemented to lessen COVID-19’s impact, 2 million deaths were projected by the end of 2020—a reduction of the average U.S. lifespan by five years, the researchers point out.

Their estimated drop in life expectancy is modest, in part because 250,000 deaths is not a large increase on top of the 3 million non-COVID-19 deaths expected for 2020. The study also notes that older people, who typically have fewer remaining years of life than others do, represent the most COVID-19 fatalities.

Still, while COVID-19 mortality rates in general remain lower than those of the 1918 Spanish flu pandemic, the toll of the coronavirus in the United States could be just as devastating as the country’s longer-lasting HIV and opioid epidemics if mitigation efforts fail, the researchers said.

“The death toll of COVID-19 is a terrible thing, both for those who lose their lives and for their family, friends, colleagues, and all whom their lives touched. Those are real people, not abstract statistics,” says Lee.

“But the population perspective helps put this tragedy in a broader context. As we work to contain the coronavirus, it is important to know that the United States has been through such mortality crises before,” he adds.

About Nine Close Relatives Suffer Grief With Each COVID-19 Fatality

The ripple effects of each COVID-19 death will impact the mental and physical health of about nine surviving close family members, a study of kinship networks shows.8

For example, when 190,000 people were dead from the disease in September 2020, 1.7 million Americans experienced the loss of a close relative, explains Ashton Verdery of Penn State University. A kinship network includes grandparents, parents, siblings, spouses, and children. Black Americans had a slightly higher number of close relatives than white Americans, averaging an estimated 9.2 people compared with 8.9, they found.

If 1 million people eventually die from COVID-19, then 8.9 million—or about 3 out of 100 Americans—would be in mourning.

These findings can help raise awareness about the scale of the disease and the ripple effects that deaths may have on a community, as well as prepare officials and business leaders to manage those effects, according to Verdery.

“It’s very helpful to have a sense of the potential impacts that the pandemic could have,” he says. “And, for employers, it calls attention to policies around family leave and paid leave. At the federal level, it might inform officials about possible extensions for FMLA (Family and Medical Leave Act). There could also be some implications for caretaking. For example, a lot of children grow up in grandparent-led houses and they would be impacted.”

Many people are also facing the loss of a close loved one at a younger age because of the disease, according to Verdery, who worked with Emily Smith-Greenaway of the University of Southern California, Rachel Margolis of the University of Western Ontario, and Jonathan K. Daw at Penn State.

“There are a substantial number of people who may be losing parents that we would consider younger adults and a substantial number of people may be losing spouses who are in their 50s or 60s,” he suggests.

Their findings could help local officials understand and prepare for the waves of grief that may affect specific geographic areas and regions of the country.

This article was produced under a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). Yasmin Anwar of the University of California Berkeley and Matt Swayne of Penn State University contributed to this article. The work of researchers from the following NICHD-funded Population Dynamics Research Centers was highlighted: University of California, Berkeley (2P2CHD073964-05A1); University of Minnesota (5P2CHD041023-19); and Penn State University (5P2CHD041025-19).

A list of newly published research on the pandemic by NICHD Population Dynamics Research Centers can be found here.

To combat obesity and diabetes, lawmakers in a number of U.S. cities have taxed sodas, sports drinks, and sweetened tea, and many are now considering health warning labels.

Growing evidence suggests that both strategies—taxes and warning labels—can reduce the purchase and consumption of sugary drinks, research supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) shows.

Health-Warning Labels Influence What People Buy and Consume

Even brief exposure to health warnings on sugar-sweetened beverages reduces purchases of those beverages, providing evidence that such warnings can promote healthier drink choices, a new study demonstrates.¹

A team of researchers from the University of North Carolina at Chapel Hill (UNC)—including Anna Grummon, Lindsey Smith Taillie, Shelley Golden, Marissa Hall, and Noel Brewer—examined how health warnings influence what consumers actually buy in real settings. This randomized controlled trial assigned 400 consumers of sugary beverages to groups that saw either a health warning or a label that looked like a barcode.

“We worked in a convenience-store laboratory that allowed us to control whether the sugary drinks had warnings,” explains Grummon, now at Harvard University. “We are also one of the first studies to measure what consumers actually buy after seeing warnings, when they have their own money on the line.”

Participants who saw the health warning labels purchased about 22% fewer calories from sugary drinks compared with participants who saw a neutral label. The study also found that the warnings were influential across diverse groups: The effect of health warnings on beverage purchases did not differ by participants’ race/ethnicity, education, age, gender, sexual orientation, income, body weight, or health-literacy level.

According to Grummon, critics of health warning labels argue that consumers won’t notice or pay attention to the warnings. However, three-quarters of the participants in this study reported noticing the health warnings, and most of those participants also reported that they read and looked closely at the labels.

In another study, Grummon and Hall synthesized the findings of 23 studies and found that health warnings labels not only reduced purchases of sugary drinks but also caused stronger emotional responses, increased perceptions that sugary drinks contribute to disease, and reduced intentions to purchase or consume sugary drinks.² All these responses are key indicators when it comes to long-term behavior change, they note.

“Our findings suggest that sugary drink warnings help consumers better understand products’ healthfulness and encourage them to make healthier choices about what drinks to buy,” says Grummon.

In a related mathematical simulation, UNC researchers show that a national policy requiring health labels on sugar-sweetened beverages could reduce obesity prevalence by about 3.1 percentage points over five years, if sustained.³

“While three percentage points might sound modest, on a national scale it equates to more than five million fewer people with obesity,” says Grummon. “Warnings are a highly scalable strategy for helping consumers make healthier choices. These findings suggest that warnings are also promising for addressing obesity in the U.S.”

Improved Child Health Projected in Wake of Mexico’s Soda Tax

The Mexican government enacted the first national tax on sugar-sweetened beverages after a 2012 study indicated that more than 70% of the country’s population was overweight or obese, and that in excess of 70% of the added sugar calories in the Mexican diet were coming from sugary drinks.

In the two-year period spanning 2014 to 2015, a research team that included Barry M. Popkin and Shu Wen Ng of UNC found that:

• The one-peso-per-liter excise tax on sugar-sweetened beverages in Mexico resulted in a 6% reduction in purchases of taxed beverages during the first year and continued to decline, with a 10% decrease in purchases in the second year.
• During the same study period, purchases of untaxed beverages such as bottled water increased 2.1%.
• Residents of households with lower socioeconomic levels, for whom health care costs are most burdensome, reduced their purchases of sweetened beverages the most.⁴

The findings run counter to initial reports from the sugar-sweetened soda industry, which said that the purchases of sugary drinks actually went up after the initial tax year. However, the researchers found those reports did not account for numerous significant factors, including inflation and shifts in population.

In addition, a new analysis co-authored by Popkin estimates that Mexico’s sugar-sweetened beverage tax could result in meaningful weight control and life-long health benefits for the country’s children and adolescents, particularly those who had been high consumers of the beverages before the tax.⁵ Childhood obesity is a strong predictor for obesity later in life, which can also lead to chronic illnesses such as diabetes, hypertension, and heart disease, the researchers emphasize.

To estimate the one-year effect of the tax on the body weight of children ages 5 to 17, by taking into account patterns of childhood growth and obesity in Mexico and assuming that the known reductions in sugar-sweetened beverage purchases would reflect changes in consumption.

Findings show that one year after the implementation of the current tax, children and adolescents should experience an average reduction in body weight of 0.26 and 0.61 kg (one kilogram equals about 2.2 pounds). For those who had been high consumers of sugary drinks, the team estimates the positive impact on body weight would be even greater, with an average body weight reduction of 0.50 kg for children and 0.87 kg for adolescents. Sustained over several years, such weight loss could mean some children and adolescents would not longer be considered obsese.

“Taxation represents one of the most effective ways to reduce consumption of unhealthy sugar-sweetened beverages, which can make a meaningful impact on future excessive weight gain and significantly reduce the long-term risks of becoming obese,” says Popkin. “If the taxation revenue is used to support child and adolescent healthy eating, then the benefits of such taxes are enhanced.”

Public Support Is Key to Policies Limiting Sugary Beverages

For taxes on sugary beverages to become a widely used strategy for improving public health, public support and acceptance are key.

Public opinion on the policies’ unintended consequences may affect attitudes toward the policy, argue Melissa Knox, Jessica Jones-Smith, and Vanessa Oddo of the University of Washington, who analyzed perceptions of the effects of Seattle’s 2017 sugary beverage tax.⁶

“We find that a majority of participants (59%) support the sugary beverage tax in Seattle and correspondingly, most people believed that the tax will positively impact health, and will not negatively affect general and personal economics in Seattle,” they report. “However, lower-income, versus higher-income, respondents were more concerned about the possible negative economic consequences of the tax,” such as job loss or increased financial costs for their family and friends.

A related study shows that attitudes toward sugary beverage taxes may be difficult to accurately estimate in phone surveys.⁷ Phone respondents (but not web respondents) under-report their sugary beverage consumption by about 25% and over-report positive attitudes toward the tax by about 11%, the researchers determined. These differing results likely reflect respondents’ answering interviewers’ questions in ways they believe are more socially desirable or acceptable rather than choosing responses that reflect their true thoughts or feelings, a tendency known as social desirability bias.

The researchers offer advice to lawmakers implementing soda taxes.

• Policymakers “should be wary of solely relying on self-reported measures of intake when evaluating the effectiveness of these policies,” they write, noting that consumers may consume more sweetened beverages than they report.
• Lawmakers should strengthen “their public messaging regarding the health and economic benefits of sweetened beverage taxes, even if they believe that attitudes are generally positive. Without a pro-tax messaging campaign that informs the public about the positive health and economic effects of these taxes, the taxes may eventually lose public support.”

The researchers point out that “recent successful efforts to block U.S. municipalities from enacting future beverage taxes by banning the taxes at the state level have relied heavily on informational campaigns that focused on the negative economic effects of the taxes. These campaigns, often funded by the beverage industry, may ultimately shift social norms in the direction of more favorable attitudes toward sweetened beverages, with unpredictable effects on public health.”

This article was produced under a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). The Communications and Marketing team at the Gillings School of Global Public Health at the University of North Carolina at Chapel Hill contributed to this article. The work of researchers from the following NICHD-funded Population Dynamics Research Centers was highlighted: University of North Carolina at Chapel Hill (P2CHD050924) and University of Washington (5P2CHD042828-18).

People of any age with obesity face a greatly increased risk of severe illness and death from COVID-19, the disease caused by the novel coronavirus, confirms new research supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD).¹

An analysis of published data led by Barry Popkin of the University of North Carolina at Chapel Hill found that people with obesity—defined as a body mass index (BMI) of 30 or higher—who were infected with the virus had a higher risk of hospitalization, admission to intensive care, and death compared with those without obesity (see Figure 1).

FIGURE 1. 
People With Obesity Face Higher Health Risks From COVID-19 Than People Without Obesity

Source: Barry M. Popkin et al., “Individuals With Obesity and COVID‐19: A Global Perspective on the Epidemiology and Biological Relationships,” Obesity Reviews 21, no. 11 (2020): e13128, https://doi.org/10.1111/obr.13128.

Obesity is an epidemic of its own in rural America: In a recent interview Lee pointed to analysis showing that 43% of Americans in nonmetropolitan areas were obese in the period from 2013 to 2016 compared with 35% of their peers in large metropolitan areas, a full eight percentage-point difference.²

FIGURE 2.
Obesity in the United States Is More Prevalent Outside of Metropolitan Areas

Source: Craig M. Hales et al., “Differences in Obesity Prevalence by Demographic Characteristics and Urbanization Level Among Adults in the United States, 2013-2016,” JAMA 319, no. 23 (2018): 2419-29.

“A large coronavirus outbreak in a rural area could lead to a high level of severe cases and greater mortality because the rural population has more underlying conditions, including obesity, and tends to be older,” Lee says. “There are fewer health care resources such as ICU beds, and distances make accessing care more difficult.”

Recent research by Lee and other NICHD-supported investigators examines the dynamics of rural obesity, underscoring the importance of prevention efforts.

Living in Rural Areas Linked to Weight Gain and Obesity

Not only are rural residents more likely to be obese than urban residents, people who move to rural counties are more likely to become obese than their peers who remain in cities, Lee reports.³

These findings come from a study based on 1986 to 2014 data from the National Longitudinal Survey of Youth, which began regularly surveying a nationally representative sample of Americans ages 14 to 22 in 1979 about important life events such as employment, education, health problems, and marriages. The analysis takes into account differences in the makeup of rural and urban populations.

Previous research may explain some of these patterns: Rural areas have a lot of open space, but the roads are not conducive to physical activity, and sidewalks and parks are limited, Lee says. Many rural residents have long commutes and rely on cars for all of their errands.

Grocery stores selling high-quality fresh produce are also more limited. The U.S. Department of Agriculture identified census tracts that have limited access to healthy and affordable food and found that rural tracts were nearly twice as likely to have limited access compared with urban tracts, at 15.9% and 8.2% respectively.⁴

“Convenience stores that sell highly processed items and limited perishable foods like Dollar General are the largest retailers in rural America,” Lee points out.

Given the increased risk of weight gain in rural areas, residents would benefit from policies that address obesity and exercise such as expanding access to quality food and opportunities for physical activity, he notes.

Residents in More Remote Rural Areas Face a Higher Risk of Obesity

The more remote the rural area, the higher the obesity risk residents face, another recent study showed.

Rural Americans living in remote areas are more likely to have obesity and diabetes than their counterparts whose homes are closer to highways and near small towns, according to research conducted by a team from the universities of Maryland, Utah, and California at San Francisco.⁵

The researchers analyzed neighborhood characteristics nationwide using a comprehensive set of 16 million Google Street View images linked with health data from the U.S. Centers for Disease Control and Prevention. Their analysis took into account differences in county-level demographics, economic characteristics, and population density.

The greater abundance of services and facilities found in areas with more roads may offer residents greater access to places and resources that promote health, they suggest.

“Poorer health in rural areas is not inevitable,” the researchers argue. Addressing obesity and promoting health may involve advocating for roads and community resources such as parks and sidewalks—especially in more resource-poor areas, they suggest.

Rural-Urban Life Expectancy Gap Widens

Over the past three decades, and particularly since 2010, gains in life expectancy have slowed or even reversed in some places for non-Hispanic whites, particularly those living in nonmetropolitan areas, new research shows.⁶

Deaths related to the opioid epidemic explain part of this trend, but rising obesity may also play a role, suggest Irma T. Elo, Yana Vierboom, and Samuel H. Preston at the University of Pennsylvania, Arun S. Hendi of Princeton University, and Jessica Y. Ho of the University of Southern California.

The team of researchers analyzed death data from 1990 to 2016 by age, sex, race/ethnicity, and cause of death for American adults ages 25 and older compiled by the National Center for Health Statistics, then estimated death rates by age, year, and geographic region. They grouped the data into four categories based on geographic characteristics: large central metropolitan areas, large metropolitan suburbs, small/medium metropolitan areas, and nonmetropolitan areas.

As part of the comprehensive analysis, they compared the change in life expectancy with the change in obesity rates in 40 geographic areas. They found that decreases in life expectancy were associated with increases in obesity more closely than any other variable they examined.

Obesity is a risk factor for cardiovascular disease and other chronic conditions, they note, and suggest that it “plays some role in the trends we observe.”

Living in Disadvantaged Areas Increases the Odds of Obesity Among Adults

Might the higher prevalence of obesity in rural America be related to the level of disadvantage in some rural areas?

A team of researchers examining risk factors for cardiovascular disease found that living in a socioeconomically disadvantaged areas early in life was linked to high blood pressure in adulthood, while living in a disadvantaged area in adulthood was linked to obesity.⁷ They categorized disadvantaged areas based on levels of household income, unemployment, and residents without a high school diploma.

These researchers, from Brown, Johns Hopkins, and Harvard universities, based their analysis on longitudinal data from the New England Family Study, which tracked 671 people in Massachusetts and Rhode Island over 46 years, from 1961 to 2007.

Their findings provide additional evidence that living in a low-income area is associated with an elevated risk of obesity, suggesting that people living in disadvantaged rural areas are an important target for obesity prevention initiatives.

Given that large shares of rural Americans have—or are at risk of—obesity and other medical conditions that can make COVID-19 severe, regular preventive health care is more important than ever, Minnesota’s Lee argues. Such care includes flu shots and other immunizations, blood pressure checks, blood sugar and cholesterol monitoring, and mental health screening.

This article was produced under a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). The work of researchers from the following NICHD-funded Population Dynamics Research Centers was highlighted in this article: University of Minnesota (P2CHD041023), University of Pennsylvania (P2CHD044964), University of Maryland (P2CHD041041), Brown University (P2CHD041020), and University of North Carolina at Chapel Hill (P2CHD050924).

More unmarried couples today are living together, and doing so for longer than in the past, but fewer of these relationships lead to marriage, new research finds. This change may in part reflect shifting attitudes toward cohabitation, and it results in more separations and re-partnering during young adulthood.

Most young women today will live with a romantic partner at least once, compared with just one-third of young women in the late 1980s.¹ During that decade, most cohabiting relationships were short-lived and frequently led to marriage.

The new research, conducted by graduate students and faculty at the Center for Family and Demographic Research at Bowling Green State University, examined how cohabitation and marriage patterns have changed for young women over the past four decades. Their research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD).

In their study, Esther Lamidi, now at the University of Colorado Colorado Springs, and colleagues Wendy Manning and Susan Brown at Bowling Green, drew on data from the National Survey of Family Growth (NSFG) to compare women ages 15 to 39 who lived with a first romantic partner in 1983-1988 and in 2006-2013.² They examined changes in whether couples who lived together had married or split up within five years.

They found that while cohabiting relationships are still relatively short-lived, couples today are cohabiting longer—increasing from about 12 months in the 1983-1988 cohabitation cohort to 18 months in the later cohort—and that this longer duration is linked to couples delaying or forgoing marriage altogether. After five years, similar shares of women in both cohorts were still living with their partner, but the distribution of those still cohabiting as compared to those who had married had shifted. Among the early cohort, 23% of women were still cohabiting five years later, and 42% had married their partner. These shares were reversed among the later cohort—43% were still cohabiting and only 22% had married.

Women With Less Education Experience More Changes in Cohabitation

Over the past five decades, changes in family behaviors such as declining rates of marriage have been more pronounced among women with less education compared with women who have more education. Lamidi and her colleagues confirmed this divergence—similar to what’s been observed in other family behaviors and frequently termed “diverging destinies”—when they examined patterns of cohabitation across different sociodemographic groups.

Their analysis found that the more recent cohort was much less likely to marry their cohabiting partner, and while this pattern was observed across all sociodemographic groups, it occurred more frequently among women with less education.

After accounting for women’s educational attainment, their results show that between the two cohorts only women with less than a college education experienced a decline in marrying their cohabiting partner. In addition, women having one or more children while cohabiting—an occurrence more common among women with less education—delayed or inhibited marriage more for the later cohort than the earlier cohort, they found.

Cohabitation Changes Reveal a Widening Social Class Divide

Sociodemographic characteristics are associated with the pathways out of cohabitation—break ups or marriages—and changes among the cohabiting population’s characteristics can be reflected in changes in cohabitation outcomes. Yet while the researchers noted that the cohabiting population grew in size, became more racially and ethnically diverse and more highly educated, and had more births while living together, they found these compositional changes had little impact on the changes in cohabitation outcomes across the two cohorts.

What does this finding mean? The researchers conclude that the limited impact of population composition changes on cohabitation outcomes, combined with the decline in marrying a cohabiting partner among women with less education, suggests that the social class divide in the American family appears to be widening.

Their findings also “diminish the traditional view of cohabitation as a prelude to marriage” for women with less education and show, particularly for this population, that “cohabitation is increasingly serving a role similar to that of traditional marriage in offering a viable context for childbearing and child-rearing.”

Young Women Today Are Increasingly Likely to Experience a Breakup

Although cohabiting relationships may be lasting longer, they remain relatively unstable. Kasey Eickmeyer, now at the Center for Policing Equity, reports, “Millennials experienced more relationship instability during young adulthood than earlier birth cohorts of women.” She found that cohabitation experience accounted for this instability.

Eickmeyer asked whether young women see their intimate live-in relationships (either marriage or cohabitation) end more frequently today than earlier generations.³ She analyzed data from multiple cycles of the NSFG to examine women’s experience of ending marriages and cohabiting relationships when they were ages 18 to 25 across several five-year birth cohorts from 1960 to 1985.

She found that among women who had ever married or cohabited, the share breaking up with a live-in partner increased from 31% among women born between 1960 and 1964 to 44% among women born in 1985 to 1989.

Cohabitation explains this increasing likelihood of experiencing a breakup. Compared to women in the 1985-1989 birth cohort, women in the earlier birth cohorts from 1960-1964 through 1975-1979 were significantly less likely to have one or more live-in partnerships end. Once Eickmeyer accounted for women’s cohabitation experience, she found that young women’s increased likelihood of having an intimate partnership end is because union formation during young adulthood shifted from marriage—a relatively stable union—to cohabitation, a relatively unstable union.

More Breakups and Re-Partnering in Young Adulthood Suggest Changing Attitudes About Cohabitation

As more young women enter into and end cohabiting relationships, they have more opportunities to live with multiple partners in a pattern of serial cohabitation. The growing practice of serial cohabitation reflects in part changing attitudes about couples living together without marriage.

Eickmeyer and Wendy Manning wanted to know whether contemporary young adult women who had ever cohabited are more likely to re-partner than prior cohorts of young women.⁴ Using data from the 2002 and 2006-2013 NSFG, they compared the cohabitation experience of young women ages 16 to 28 across five-year birth cohorts beginning in 1960 through 1980 to examine trends in serial cohabitation.

They found that early Millennial women (born 1980-1984) were 53% more likely to live with more than one romantic partner during young adulthood compared with the late Baby Boomers (born 1960-1964), even after taking into account sociodemographic characteristics such as race and ethnicity and educational level, and relationship characteristics such as their age when their first cohabiting relationship ended and whether they had children.

Not only were early Millennial women more likely to live with more than one partner without marriage, they also formed subsequent cohabiting relationships more quickly than the late Baby Boomers—dropping from nearly four years between live-in relationships to just over two years.

The characteristics most strongly associated with serial cohabitation—such as identifying as non-Hispanic white, having less than a college education, and growing up with a single parent—remained stable across birth cohorts, Eickmeyer and Manning found. And, much like the cohabiting population, the composition of women who had previously lived with a partner changed across cohorts, but this shift does not explain the increase in serial cohabitation.

The researchers conclude that the increase stems from more young adults cohabiting, the continued instability of cohabiting relationships, the increasing length of time between first cohabitation and first marriage, and the growing acceptance of cohabitation during young adulthood.

Their findings highlight the instability in many contemporary young adults’ lives and the increasing role cohabitation plays in relationship churning. Although multiple live-in romantic relationships could have negative consequences for young adults’ well-being (and any children they may have), Eickmeyer and Manning suggest “that young adult relationships may be evolving, and young women may be learning to end coresidential relationships that are not working.”

This article was produced under a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). The work of researchers from the NICHD-funded population dynamics research center at Bowling Green State University (P2CHD050959) was highlighted in this article.