Component 5: Pathways to the Middle Class and the Geography of Economic Mobility

Are pathways to the middle class getting narrower and harder to climb?

This section examines the factors that shape upward economic mobility and access to the middle class. It includes both macro-level measures, which indicate whether the broader economy is structured to support upward mobility, and micro-level measures, which assess whether conditions within communities provide residents with meaningful access to opportunity. While data limitations and time lags make it difficult to measure intergenerational economic mobility directly, the indicators selected here are informed by research demonstrating that childhood environments can have a significant impact on educational attainment and economic outcomes later in life.

Introduction

Earlier sections examined what it costs to maintain a middle-class standard of living and how many Massachusetts households can afford it today. This section turns to a different question: How strong are the pathways that allow people to enter the middle class and remain there over time?

Economic mobility depends on more than individual earnings or educational attainment. It is shaped by how the gains from economic growth are distributed, how much economic output reaches workers through compensation, and whether families can translate their incomes into stable housing, independent household formation, business ownership, and long-term assets. It also depends heavily on place. Residential segregation and concentrated poverty can produce sharply different access to safe neighborhoods, strong schools, employment networks, institutions, and other resources that support upward mobility.

The measures that follow examine these conditions from four perspectives: the distribution of income and economic opportunity; the geography of opportunity; education and employment pathways into middle-class occupations; and opportunities for economic independence and asset-building. No single measure can determine whether an individual will reach the middle class. Taken together, however, they show whether Massachusetts is building broadly accessible pathways to realizing the American Dream.

I. The Distribution of Income and Economic Opportunity

The strength of pathways into the middle class depends partly on who benefits from economic growth. The measures in this section examine this question from two perspectives: differences in income among households and the share of economic output that reaches workers through wages, salaries and employer-provided benefits.

Income inequality

Research suggests high levels of income inequality reduce intergenerational mobility across countries and communities within the United States. Income inequality also shapes access to the middle class by generating inflationary pressures disproportionately borne by those in the lower half of the income distribution and limiting the resources families can invest to build wealth. Recent modeling suggests that these inflationary pressures may be especially consequential in high-cost housing markets.¹

Measure 1A showing changes in household level income inequality by combining Gini coefficient values from decennial census data with annual ACS estimates beginning in 2010. The Gini ranges from zero, where all income is spread equally across all households, to one, where income is concentrated in a single household. The measure does not adjust for household size or composition, so changes in household formation can affect the results. It also excludes capital gains, understating disparities in the total resources available to households at the top. Because it summarizes differences across the entire income distribution, seemingly small movements can be meaningful: a change of 0.01 equals one full point on a 0 to 100 scale and can reflect a broad shift in how income is distributed.

Income inequality has grown substantially both in Massachusetts and across the nation. In 1979, the national Gini coefficient was 0.415, compared with 0.398 in Massachusetts. Massachusetts caught up by 1999, when both stood at 0.463. The sharpest increases occurred during the 1980s and 1990s: Between 1979 and 1999, the Gini rose nationally by 0.048 points, or 12 percent, and in Massachusetts by 0.065 points, or 16 percent. Inequality has continued to increase since then, but much more slowly as measured by the household income Gini coefficient. Between 1999 and 2024, the Gini rose by approximately 0.02 points, or about 4 percent, both in Massachusetts and across the nation.

Massachusetts had the fifth-highest Gini coefficient in the nation in 2024, trailing only New York, Connecticut, Louisiana, and California. It also had the third-highest cost of living for a single adult. Across all states, inequality and living costs were positively but only modestly correlated, indicating that high costs are not confined to the most unequal states. Massachusetts nevertheless occupies a particularly challenging position near the top of both distributions. This combination means that households lower in the income distribution must contend with prices shaped by one of the nation’s most expensive markets without sharing proportionately in the purchasing power concentrated among its highest earners.

How economic growth reaches workers

For most working-age adults, access to the middle class depends primarily on earnings from employment. Among Massachusetts adults in the middle half of the income distribution, wages and salaries account for 91 percent of total income. Yet economic growth does not necessarily translate into greater compensation for workers: State economic output can increase even when wages, salaries, and employer-provided benefits do not keep pace. Employee compensation’s share of gross state product measures how much of the value generated by the economy reaches workers.

The measure does not show how compensation is distributed among employees. A state can have both a relatively high employee-compensation share and substantial household inequality if those gains are concentrated among highly paid workers. However, read alongside the Gini coefficient and the later measures of occupational and educational access, 1C helps distinguish between two questions: how much economic output reaches employees, and how broadly access to that compensation is distributed.

Since 1970, employee compensation has generally declined as a share of economic output both in Massachusetts and across the nation. By 2024, employee compensation accounted for 55 percent of economic output in Massachusetts and 51 percent nationally, down from 62 percent and 57 percent in 2000, respectfully.² Massachusetts nevertheless ranked fifth among states in employee compensation’s share of economic output in 2024.

Taken together, income inequality and employee share of economic output reveal an important tension. Massachusetts directs a comparatively large share of its economic output to employee compensation, yet household income remains among the most unequal in the country. This combination is consistent with an economy in which compensation is relatively strong in aggregate but unevenly distributed, making access to well-paying occupations and the education they require central to upward mobility.

II. The Geography of Opportunity

Research finds that rising income inequality contributes to residential sorting, particularly the geographic isolation of affluent households.³ This separation can, in turn, perpetuate inequality across generations: Where children grow up affects their economic prospects, and separation across neighborhoods and schools can limit access to resources and cross-class relationships that support upward mobility.⁴ This section examines the extent to which Massachusetts residents of different incomes live in separate communities and how unevenly children are exposed to concentrated poverty and adults with higher levels of educational attainment.

Income segregation

Research by Raj Chetty and colleagues increasingly identifies cross-class relationships as a central mechanism connecting place to upward mobility. Their measure of “economic connectedness”—the extent to which people with lower socioeconomic status form friendships with people of higher socioeconomic status—is more strongly associated with upward mobility than other forms of social capital and accounts for much of the observed relationship between economic mobility and racial segregation, concentrated poverty, and inequality.⁵

Income segregation is tracked with the Reardon index, a tool widely used by economists to gauge how evenly families across the income distribution are distributed among geographic areas. Similar to the Gini coefficient, a value of 0 means that every neighborhood has the same income distribution as the broader region, while higher values indicate greater residential sorting by income.

By this measure, Massachusetts had the sixth-highest neighborhood-level income segregation in the nation in 2024. Only Virginia, New Jersey, New York, Texas, and Connecticut recorded higher levels. This pronounced residential sorting mirrors the commonwealth’s relatively high level of income inequality.

Similar to the Gini, income segregation was largely stable during the 2010s, both in Massachusetts and across the states. However, a starkly different pattern has emerged since the COVID-19 pandemic. Massachusetts’s Reardon index has risen sharply, while the median level across other states rose after the pandemic but has since trended downward toward pre-COVID levels. Although comparisons across the 2020 break should be treated cautiously due to changes in Census tract boundaries and other potential sources of statistical noise, the steady increase over several years merits closer attention.

Income segregation varies considerably across Massachusetts metropolitan areas. In 2024, the Boston metropolitan area had the highest neighborhood-level income segregation, with an index of 16.3—more than twice the national metropolitan median of 6.9. Springfield and Worcester also recorded comparatively high levels, at approximately 13.2 each, followed by Providence-Warwick—which spans several southeastern Massachusetts cities, including Attleboro, Fall River, New Bedford, and Taunton—at 12.2. Barnstable was close to the national median, while Amherst-Northampton and Greenfield were below it.

Most of these regions also became more segregated between 2009 and 2024. The largest increases occurred in Boston and Barnstable, where the index rose by 3.5 and 3.2 points, respectively. Worcester and Amherst-Northampton experienced increases of about 2 points, while Providence-Warwick and Greenfield changed relatively little. Springfield was the only region where measured segregation declined, falling by 0.7 points. Because the series crosses the 2020 transition to new census tract boundaries, these long-term changes should be interpreted with some caution.

Income segregation is less pronounced between school districts than between neighborhoods, partly because school districts encompass larger and more internally diverse areas. Nevertheless, Massachusetts had the third-highest level of between-district income segregation in the nation in 2024, behind only New Jersey and Connecticut. Its Reardon index of 0.09 was nearly twice the median across states. The result suggests that Massachusetts families of different incomes are unusually likely to live on opposite sides of school-district boundaries, producing large differences in the populations that local school systems serve.

The commonwealth’s district-level Reardon index fluctuated within a narrow range, from 0.087 in 2009 to 0.090 in 2024, while the median across states remained near 0.046. Unlike the neighborhood-level measure, the school district series shows no apparent discontinuity in 2020 and no sustained post-2020 increase. Instead, it suggests that Massachusetts’s unusually high between-district segregation is a longstanding feature of its residential and school district geography.

Concentrated poverty

Researchers have long demonstrated a connection between growing up in neighborhoods with high concentrations of poverty and upward economic mobility. To assess the extent to which low-income households are geographically isolated in Massachusetts, we use two complementary measures.⁶

The poor-nonpoor dissimilarity index tracks how segregated poor households (those with income below the federal poverty threshold) are from non-poor households. A score of 0 means the two groups are fully integrated, while a score of 1 means the two groups live in entirely separate neighborhoods. The poor-nonpoor dissimilarity index has gradually declined in Massachusetts, falling from 0.396 in 2009 to 0.353 in 2024, an approximately 11 percent reduction. This suggests that residents with incomes below the official poverty rate became more evenly distributed across census tracts.

Despite this apparent progress, poverty remains more geographically segregated in Massachusetts than in most states. Massachusetts had the eighth-highest poor-nonpoor dissimilarity index in 2024.

High-poverty neighborhoods—areas where more than 30 percent of residents live below the federal poverty threshold—are heavily concentrated in Massachusetts’s urban communities, particularly its Gateway Cities. In 2024, Gateway Cities contained 57 of the state’s 99 neighborhoods with concentrated poverty, or 58 percent. Another 35 were located in the city of Boston.

More than one in five Gateway City residents lived in a census tract with a poverty rate above 30 percent. Statewide, these neighborhoods were home to approximately 115,000 residents in poverty—17 percent of Massachusetts’s population in poverty. Nearly 80,000 lived in Gateway Cities, representing 68 percent of all Massachusetts residents experiencing both individual and concentrated poverty. By comparison, none of the state’s rural municipalities contained a high-poverty tract.

The share of Massachusetts census tracts experiencing concentrated poverty rose following the Great Recession, increasing from 7.9 percent in 2009 to a peak of 10.4 percent in 2013. It subsequently declined for much of the next decade, reaching 6.0 percent in 2020. Since then, the share has fluctuated within a relatively narrow range, ending at 6.2 percent in 2024. Overall, concentrated poverty is substantially less geographically prevalent than at its post-recession peak, but recent data suggest that progress has largely leveled off rather than continued.

Children’s exposure to highly educated adults

Research increasingly suggests that neighborhoods impact upward mobility at least partially through the social connections they afford to people with access to resources, including experience with higher education.⁷To measure this dimension of children’s residential context, we examine the share of working-age adults with a bachelor’s degree in each neighborhood.

Massachusetts children are especially likely to live in neighborhoods with high levels of adult educational attainment. In 2024, nearly half (47 percent) lived in census tracts above the 75th-percentile of educational attainment nationally—the largest share among US states. Massachusetts is also among the states with the lowest shares of children living in neighborhoods with limited exposure to highly educated adults (census tracts below the 25th percentile threshold nationally). However, while only 10 percent of children in the commonwealth currently live in such neighborhoods, this share is increasing faster in Massachusetts than in the US overall.

The proportion of children living in tracts that fell within the lowest national quartile of adult educational attainment increased from 8 percent in 2009 to 10 percent in 2015 and has generally remained at this level over the past decade. In 2024, approximately 104,000 Massachusetts children lived in census tracts with low exposure to college-educated adults. The vast majority of these children (86 percent) lived in Gateway Cities, which contained 125 of the state’s 148 low-exposure tracts.

In 2024, approximately 104,000 Massachusetts children lived in census tracts below the national 25th-percentile threshold for working-age adult bachelor’s-degree attainment. The vast majority of these children (86 percent) lived in Gateway Cities, which contained 125 of the state’s 148 low-exposure tracts.

Low educational exposure and concentrated poverty frequently overlap, but not always. Of the state’s 99 high-poverty tracts, 42 also had low educational exposure. Gateway Cities contained 37 of the 42 tracts meeting both criteria. By comparison, Boston alone accounted for 31 of the 32 tracts with concentrated poverty only. Thus, high-poverty neighborhoods in Gateway Cities were considerably more likely to combine economic deprivation with limited neighborhood exposure to college-educated adults, while Boston’s generally higher educational attainment meant that most of its high-poverty tracts did not fall below the national educational-exposure threshold.

III. Educational and Employment Pathways Into the Middle Class

Reaching the middle class also depends on the pathways available through education, employment, and small business formation. This section examines the credentials associated with middle-class occupations, whether residents’ education aligns with available jobs, where those jobs are located, and whether self-employment provides a viable alternative route.

Educational requirements of middle-class occupations

Recent analysis by the Aspen Institute suggests 68 percent of living-wage jobs in Massachusetts require a bachelor’s degree or higher.⁸ But earning a living wage is far from achieving the American Dream. Our middle-class cost estimate gives us a new benchmark to estimate how central a bachelor’s degree has become to making it into the middle class.

Affording a middle-class lifestyle requires an annual salary of at least $90,000 for a single adult in Massachusetts and roughly the same amount per adult in a family of four. In Component 2, we show that Massachusetts had approximately 1.04 million jobs with median salaries at or above this threshold. Of those jobs, 960,230, or 92 percent, were in occupations that typically require a bachelor’s degree or higher according to the BLS. (Occupations calling for a bachelor’s degree alone accounted for 76 percent, while those typically requiring a master’s, doctoral, or professional degree accounted for an additional 16 percent).

Nondegree pathways represented a smaller but still important share of employment in middle-class occupations. Positions that the BLS classifies as requiring only a high school diploma accounted for 5 percent of employment, approximately the same share as occupations requiring a master’s degree. However, the educational categories used in federal data systems do not capture the full range of educational pathways required for entry. Many occupations in the high school diploma category also require apprenticeships, occupational licenses, trade school, or substantial work experience. Taken together, the findings underscore both the central role of bachelor’s degree attainment in access to occupations associated with middle-class wages and the continued importance of skilled career pathways that are not fully reflected in conventional measures of educational attainment.

Massachusetts is more dependent on a bachelor’s degree as a pathway to middle-class employment than any other state. In 2024, the share of middle-class jobs in occupations typically requiring a bachelor’s degree or higher ranged from 52 percent in North Dakota to 92 percent in Massachusetts. A national benchmark holding occupational mix, wages and middle-class costs at typical national levels was 85 percent. Relative to that benchmark, Massachusetts’s occupational mix increased the bachelor’s-dependent share by 3.7 percentage points. Higher wages reduced it by 9.3 percentage points by bringing more occupations within reach of the middle-class threshold. But the state’s higher middle-class cost of living increased it by 13 percentage points, more than offsetting that wage advantage.⁹

This disparity reflects a broader pattern across states. In states where middle-class costs were 10 percent higher, wages were just 6 percent higher in occupations that do not typically require a bachelor’s degree, compared with nearly 9 percent higher in occupations requiring a bachelor’s degree or more. In other words, below-BA wages appear to adjust less fully to geographic differences in living costs. High costs may therefore compound existing inequality by narrowing access to middle-class earnings more severely for people seeking careers that do not require a four-year degree. This result is consistent with research documenting increasingly skill-biased wage gains in large, high-productivity labor markets.¹⁰

Educational attainment

Massachusetts also stands out nationally for the educational attainment of its working-age residents. In 2024, it ranked first among the 50 states in the share of adults ages 25 to 64 with a bachelor’s degree or higher—14.5 percentage points above the state median and 3.4 points above second-place Colorado.

However, while half of Massachusetts adults ages 25 to 64 have a bachelor’s degree or higher, only 38 percent of the state’s nonfarm wage and salary jobs are in occupations for which BLS identifies a bachelor’s degree or higher as the typical education needed for entry. This discrepancy suggests that expanding pathways to the middle class will require more than simply increasing the number of college graduates. It will also require better alignment between workers’ skills and available jobs, along with accessible pathways to the career-specific training, experience, and licensing that many occupations require.

Educational attainment has increased across every demographic group, but the gains did not eliminate—and in some cases widened—existing differences. Between 2009 and 2024, bachelor’s degree attainment rose from 42 percent to 53 percent among women and from 40 percent to 47 percent among men, increasing the gender gap from two to six percentage points.

Attainment also increased among every racial and ethnic group. By 2024, however, the share of working-age adults with a bachelor’s degree or higher ranged from 69 percent among Asian residents and 55 percent among White residents to 34 percent among Black residents and 24 percent among Hispanic residents. The distance between the highest- and lowest-attainment groups remained within a narrow band over this time period, indicating that broadly shared progress has not necessarily produced convergence.

Educational attainment also increased in every labor market serving Massachusetts, but substantial regional differences remain. Among the larger metropolitan areas, the share of adults with a bachelor’s degree or higher reached 56 percent in Boston-Cambridge-Newton and 54 percent in Amherst-Northampton in 2024. The corresponding shares were 42 percent in Worcester, 39 percent in Pittsfield, 37 percent in Providence-Warwick, and just 30 percent in Springfield. These differences have grown rather than narrowed. In 2009, Boston’s bachelor’s degree attainment rate exceeded Springfield’s by 19 percentage points; by 2024, the gap had widened to 26 points.

Geographic access to middle class jobs

Middle-class jobs are concentrated in Greater Boston both in absolute terms and relative to the region’s overall employment base. Boston accounted for 78 percent of jobs meeting the middle-class income threshold in Massachusetts, compared with its 64 percent share of all jobs. Every other region accounted for a smaller share of middle-class employment than its share of overall state employment.

These estimates apply the report’s statewide-average middle-class income threshold across all regions, though housing costs vary geographically. Some jobs below the threshold may provide a middle-class standard of living in lower-cost regions, while some above it may fall short in more expensive areas. The figures should therefore be understood as showing the distribution of jobs meeting a common statewide benchmark rather than a locally adjusted one.

The limited availability of middle-class jobs outside Greater Boston cannot be attributed simply to a lack of highly educated residents. Greater Boston had approximately 1.7 working-age adults with a bachelor’s degree or higher for every middle-class job in 2024. The corresponding ratio was 2.8 in Providence-Warwick, which includes Attleboro, Fall River, New Bedford, and Taunton; 3.0 in Worcester; 3.4 in Amherst-Northampton; and 5.5 in nonmetropolitan Massachusetts.

These results suggest that educational attainment and jobs meeting the statewide middle-class threshold are not geographically aligned. Regions outside Greater Boston have considerable stocks of highly educated workers, but their economies generate fewer jobs meeting this common benchmark. Although lower living costs may reduce the income needed to achieve a middle-class standard in some of these regions, the imbalance may still encourage some residents to commute toward Greater Boston or relocate closer to its employment base, simultaneously lengthening commutes and intensifying housing demand in the region.

Self-employment as an alternative pathway to the middle class

Formal education and access to well-paying jobs are not the only potential routes into the middle class. Some workers attempt to climb the economic ladder through self-employment, whether by operating an incorporated business, working as an unincorporated sole proprietor, or by providing services as an independent contractor. The ACS identifies these workers according to the class of work associated with their primary job or business, but it does not measure the value of the business or the wealth its owner has accumulated through it.

We therefore treat self-employment as an employment pathway rather than a measure of business ownership or asset building. To assess the strength of this pathway, we distinguish between how common self-employment is and how often it produces sufficient earnings. Because the cost of a middle-class lifestyle varies geographically, we compare workers’ earnings with the annual wage required to support one adult living independently in their state.

Self-employment accounts for 8.5 percent of employed working-age adults in Massachusetts, compared with a median of 9.9 percent across US states. Massachusetts also falls below the state median in the share of self-employed residents earning at least the middle-class benchmark: 28.5 percent, compared with 32.1 percent nationally. However, self-employed residents were only slightly less likely than wage-and-salary employees to reach the Massachusetts threshold. Self-employment therefore appears to be a comparatively uncommon pathway in Massachusetts, but not necessarily a substantially less viable pathway to middle-class earnings for those who pursue it.

The longer-term trend raises concerns about the strength of self-employment as a pathway to the middle class. In 2005, self-employed workers accounted for 10.5 percent of employed working-age adults in Massachusetts and 11.3 percent nationally. By 2024, those shares had declined to 8.5 percent and 10.1 percent, respectively. In Massachusetts, the decline was concentrated in unincorporated self-employment—which includes sole proprietors and independent contractors—while incorporated self-employment remained approximately stable. However, it’s important to note that tax rules and the administrative advantages of operating through a formal entity may have encouraged some successful sole proprietors to incorporate, particularly through S corporations.

At the same time, the earnings associated with self-employment became less likely to support a middle-class standard of living. In 2005, approximately 41 percent of self-employed Massachusetts residents earned enough to support one adult living independently, similar to the national rate of 40 percent. By 2024, the Massachusetts share had fallen to 29 percent, compared with 30 percent nationally. These descriptive trends do not establish why self-employment has become less common or why fewer self-employed workers reach the threshold. Changes in the occupational composition of self-employment, the characteristics of workers entering it, and the relationship between business earnings and living costs may all contribute. Nevertheless, the combined decline in participation and earnings attainment suggests that self-employment has contracted as a viable middle-class pathway rather than expanding as an accessible alternative to conventional employment.

Self-employment encompasses several distinct occupational and industry mixes in Massachusetts. Construction, trades, repair, and home services represented 27 percent of self-employed workers in pooled 2022 to 2024 data, making this the largest category, followed by retail, food, and personal services at 20 percent and professional and business services at 17 percent. Among self-employed residents earning enough to meet the middle-class threshold, however, professional and business services—including legal, accounting, consulting, and technical work—accounted for 31 percent. Construction and trades still represented a substantial 20 percent, followed by health and education practices at 15 percent, and finance, insurance, and real estate at 12 percent.

Only one-quarter of middle-class earners in the construction and trades category held a bachelor’s degree, compared with approximately 90 percent in professional and business services and health and education. Self-employment therefore provides a meaningful nondegree pathway into the middle class through the skilled trades, even though the residents most likely to reach middle-class earnings through self-employment remain concentrated in degree-intensive professional fields.

Self-employment is sometimes portrayed as a comparatively accessible pathway for workers who face barriers in conventional employment. In Massachusetts, however, pooled 2022 to 2024 data do not show greater racial, ethnic, or gender diversity among the self-employed. Non-Hispanic White workers accounted for 70 percent of all self-employed workers, compared with 67 percent of wage-and-salary workers, and their share rises to 82 percent among self-employed workers with incomes at or above the middle-class threshold. Self-employment was also disproportionately male: Men represented 58 percent of all self-employed workers and 70 percent of those reaching middle-class earnings. These patterns suggest that lower formal barriers to entering self-employment do not necessarily translate into equal access to successful self-employment.

While data only permit us to examine self-employment is as a pathway to a middle-class income, to the extent that self-employment also reflects business ownership and investment in a long-term financial asset, the patterns in these data do not suggest that this wealth-building pathway has been expanding for residents of Massachusetts.

IV. Household Formation and Access to Homeownership

Education and employment provide resources that can expand access to the middle class, but household formation and homeownership reveal whether economic and housing conditions allow residents to achieve important dimensions of middle-class security. It is true that living independently and owning a home are not universal aspirations, and shared households can reflect cultural preferences, caregiving, or mutual support. At the population level, however, household formation and homeownership indicate whether adults have access to residential independence, stability, and a potential wealth-building pathway.

Household formation

We focus our here on adults ages 26 to 34, a time in the lifespan when many make longer-term decisions about partnership, family, and where to settle. Because conventional homeownership rate calculations include only households that have already formed, they can leave out adults who remain in someone else’s household. We therefore examine household formation and homeownership together to assess whether housing costs and economic conditions are delaying these milestones and constraining broader life choices.

We define an independent household as one in which the person is the householder, or the householder’s spouse or partner, and they are not living with a parent or parent-in-law or sharing the household with unrelated adults or extended adult relatives.

Independent household formation among adults ages 26 to 34 has declined substantially over the past two decades. In Massachusetts, the share living in an independent household fell from 71 percent in 2005 to just 58 percent in 2024. The decline corresponds with both a large increase in the share living with parents (from 14 percent to 23 percent) and with roommates (from 15 percent to 19 percent). Although the trend has not been entirely linear—including a partial recovery after 2017—the 2024 rate remained more than 12 percentage points below 2005 levels.

Massachusetts also consistently trails the nation on this measure. The state’s independent-household formation rate was lower than the national rate in every year examined, with an average gap of approximately 4 percentage points. In 2024, 58 percent of Massachusetts adults ages 26 to 34 lived in independent households, compared with 64 percent nationally. In 2024, the gap primarily reflected the greater prevalence of shared non-parent households in Massachusetts—19 percent compared with 15 percent nationally—along with a somewhat larger share of young adults living with parents. These data cannot determine whether such arrangements reflect preference or necessity, but their growing prevalence is consistent with young adults taking longer to establish households of their own.

Homeownership

For much of the 20th century, homeownership offered middle-class households both residential stability and a primary means of accumulating wealth. Whether residents can still reach that milestone is therefore central to assessing the condition of the middle class. But the conventional homeownership rate provides an increasingly incomplete answer because it counts only households that have already formed. Adults who cannot afford to establish a household of their own disappear from the calculation altogether.

That omission matters more as household formation declines. Massachusetts’ conventional homeownership rate changed relatively little between 2005 and 2024, falling from 64.0 percent to 62.3 percent. Yet the share of all adults age 26 or older who had attained homeownership—as a householder, spouse, or partner in an owner-occupied home—fell from 64.2 percent to 59.2 percent. A similar divergence occurred nationally. What appears to be a modest decline in homeownership among existing households therefore masks a larger decline in the share of adults who have reached homeownership at all.

Large racial and ethnic disparities are evident under both measures of homeownership. In 2024, 68 percent of White non-Hispanic adults age 26 or older had attained homeownership, compared with 57 percent of Asian or Pacific Islander adults, 42 percent of other or multiracial adults, 35 percent of Black adults, and 30 percent of Hispanic adults. Traditional homeownership rates are somewhat higher but show the same general pattern.

Differences between the two measures should not, however, be interpreted exclusively as evidence of delayed household formation. The gap is largest among Asian or Pacific Islander adults, but adults ages 26 to 34 in this group are not less likely than White adults to participate in an independent household. The difference may instead partly reflect a greater prevalence of multigenerational living arrangements in which an older parent or another relative lives in an owner-occupied home.

This distinction motivates a third, complementary measure: the owner-occupancy exposure rate, or the share of residents who live in owner-occupied housing, regardless of whether they personally hold an ownership interest. Although exposure does not imply that every resident is accumulating housing wealth, it captures access to the potential stability associated with an owner-occupied home and accommodates multigenerational living arrangements that household-based measures may obscure.

This perspective is particularly revealing across geographies. In 2024, 53 percent of Gateway City residents lived in owner-occupied housing, compared with 86 percent of residents in suburban municipalities and 81 percent in rural municipalities. The Gateway City-suburban disparity is therefore 33 percentage points under the population-based exposure measure, compared with 27 points under the traditional measure. The places where many lower-income residents and residents of color live therefore provide substantially less exposure to one of the country’s principal sources of residential stability and household wealth.

Miles Corak. “Income Inequality, Equality of Opportunity, and Intergenerational Mobility.” Journal of Economic Perspectives (2013); Raj Chetty and others. “Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States.” The Quarterly Journal of Economics (2014); John Jerrim and Lindsey Macmillan. “Income Inequality, Intergenerational Mobility, and the Great Gatsby Curve: Is Education the Key?” Social Forces (2015); Xavier Jaravel. “The Unequal Gains from Product Innovations: Evidence from the US Retail Sector.” The Quarterly Journal of Economics (2019); Harrison Campbell. “Income and cost of living: are less equal places more costly?” Social Science Quarterly (2021); Thomas Goda and others. “Absolute Income Inequality and Rising House Prices.” Socio-Economic Review (2020).
↩︎
Economists have proposed several explanations for the long-term decline in labor’s share, including globalization and offshoring; automation and capital substitution; declining unionization and worker bargaining power; increasing market concentration and the rise of high-markup firms; and financialization, including growing shareholder payouts and financial claims on nonfinancial corporations. The relative importance of these mechanisms remains unsettled. See Michael W. L. Elsby, Bart Hobijn, and Ayşegül Şahin, “The Decline of the U.S. Labor Share,” Brookings Papers on Economic Activity (2013); Loukas Karabarbounis and Brent Neiman, “The Global Decline of the Labor Share,” Quarterly Journal of Economics (2014); Petra Dünhaupt, “Determinants of Labour’s Income Share in the Era of Financialisation,” Cambridge Journal of Economics (2017); Anna Stansbury and Lawrence H. Summers, “The Declining Worker Power Hypothesis,” Brookings Papers on Economic Activity (2020); and David Autor et al., “The Fall of the Labor Share and the Rise of Superstar Firms,” Quarterly Journal of Economics (2020). ↩︎
Reardon, Sean F., and Kendra Bischoff. “Income inequality and income segregation.” American Journal of Sociology  (2011). ↩︎
Chetty, Raj, et al. “Where is the land of opportunity? The geography of intergenerational mobility in the United States.” The Quarterly Journal of Economics (2014). ↩︎
Raj Chetty et al., “Social Capital I: Measurement and Associations with Economic Mobility.” Nature (2022). ↩︎
Poverty status is determined using family income and family composition, but the tract poverty rate measures the share of individual residents classified as living in poverty. ↩︎
James W. Ainsworth, “Why Does It Take a Village? The Mediation of Neighborhood Effects on Educational Achievement,” Social Forces (2002); Ann Owens, “Neighborhoods and Schools as Competing and Reinforcing Contexts for Educational Attainment,” Sociology of Education (2010); Raj Chetty and Nathaniel Hendren, “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects,” The Quarterly Journal of Economics (2018). ↩︎
See: https://www.mass.edu/strategic/documents/MassTransfer%20Meeting%20Materials/Meeting-2/Aspen-CCRC_Transfer%20Playbook%20Presentation.pdf ↩︎
An employment-weighted log-log regression with detailed-occupation fixed effects and state-clustered standard errors found that wages were significantly less responsive to state required-income thresholds in below-BA occupations than in BA+ occupations (difference in elasticities = −0.245; SE = 0.061; p < 0.001). The difference remained significant when each occupation-state observation received equal rather than employment-based weight (−0.140; p = 0.027). Estimates are cross-sectional and not causal. Education refers to the BLS typical education needed for entry, not workers’ actual attainment. ↩︎
Nathaniel Baum-Snow and others. “Why Has Urban Inequality Increased?” American Economic Journal: Applied Economics (2018). ↩︎