“The car, the house, two kids in private school…”: Who are America’s middle class?

Introduction

The 2016 election laid bare the middle class’ frustration with its economic and demographic development. Middle class Americans no longer believe they embody the spirit of the American dream, but rather the government’s broken promises. But, who are America’s middle class? In this paper, I investigate the American middle class (AMC)’s composition and development from 1960 to 2010. I analyze four dimensions of middle class development: occupational, demographic, economic, and racial. Recent research on the AMC separates economic and demographic analysis from social and racial.  Economic analysis is limited to only “middle income” Americans, whose income is, according to Pew, between sixty-six and two-hundred percent of the median household’s (Pew Research Group 2015, n.p.). Social and racial analyses focus on middle class identity: lifestyles that Americans associate with the middle class. This project builds a foundation for research into the middle class as an identity and as an income bracket. My research is motivated by the following questions: Who are considered middle class in America? What measurable qualities exist to understand the AMC beyond income? What are the long term trends for the AMC along my dimensions of analysis?

Contemporary research indicates that a middle class lifestyle is not as accessible for all Americans as it used to be. The Pew Research Center released a report in December 2015 on America’s middle income population (Pew Research 2015). The report begins with the ominous conclusion that, despite some positive trends, the AMC is on the decline. Notably, Pew’s research is limited to the middle class as “middle income” or families between two-thirds and two hundred percent of the median household income (approximately $42,000 to $126,000). Pew found that the majority of Americans are now either above or below this threshold for the first time.

According to Pew, only about 49% of Americans are middle income, but the Department of Commerce found that nearly 92% of Americans identify as middle or working class (Department of Commerce 2010, 10)[1]. Pew’s analysis of middle income Americans is not a comprehensive analysis of the American middle class. Even the United States government does not simply define middle class through the middle income specification. Rather, the Department of Commerce, in its 2010 report, found that middle class is loosely defined by income, but more accurately characterized by a set of shared values and expectations. The Department’s report cites stability and expectation of a better future as two defining characteristics of the AMC (Department of Commerce 2010, 11).

However, findings within that report and others contradict the assertion that stability and future prosperity are the middle class’ defining characteristics. Both the Commerce Department and Pew write that maintaining one’s position within the middle class, loosely defined, is much more difficult today than twenty-five years ago. Many who consider themselves to be middle class do not enjoy economic stability. Instead, because of increased income inequality or declining wages for low-skilled workers, families find it harder now than at any other time in the last quarter-century to stay in the middle class (NPR 2016, n.p.). A large portion of “middle class” Americans, even those within the “middle income” category, can no longer afford to send their children to college or live in areas with high quality schools (Department of Commerce 2010, 12-17). Thus, the Department of Commerce’s middle class values may be many Americans’ goals, but they are not the middle class’ defining characteristics.

The government’s report did indicate that a worker’s particular sector is an important characteristic for the middle class lifestyle. Income inequality and declining wages are not equally distributed across the middle income. Rather, individuals without college degrees or in declining and older industries are finding it hardest to maintain prosperity (Pew 2015,2). This finding is particularly relevant in light of Donald Trump’s victory in the 2016 presidential election. Resentment within the middle class between professionals and non-college educated whites fueled Trump’s populism and racial animosity (Williams 2016). Occupation changes in the middle class will be an additional economic indicator beyond income in my discussion of changes in the AMC.

Several reports on racial disparity in the AMC contradict one another. The Pew Research Center found that African Americans and Asians made substantial economic progress over the last twenty-five years (Pew 2015, 4). However, the report only described African Americans who already moved into the middle income population. The report has no information on how increased income affected African American movement into middle class neighborhoods. A study conducted by Richard Alba found that African Americans were making progress in wage and wealth growth. As their education and income increased, African Americans moved into less segregated or economically depressed neighborhoods. However, their neighborhoods tended to have less wealth and lower property values than comparable white or mostly-white neighborhoods (Alba 2000, 548). Thus neither income nor housing alone describe African American economic progress in the last fifty years. Rather, income and housing, along with occupation, should be discussed together to form a more comprehensive picture of race in the AMC.

The existing literature has raised several questions about the AMC that guided my inquiry. First, the question of definition. Who exactly constitutes the AMC? Income is an insufficient metric to assess the AMC. However, selecting arbitrary values or expectations to describe the identity is equally insufficient and unempirical. Herein lay the second question from the literature: what are the characteristics of families in the AMC? Through my project, I will put forward a more robust definition of the AMC and show, using census data, the individuals and families who constitute this highly revered and little understood body of Americans. Some Americans distinguish between the middle and working classes in America. However, the Department of Commerce found views on the working class to be less consistent across surveys they conducted (Department of commerce 2010, 3). I will not make a distinction between the middle and working class in this paper, but these divergent identities could be the foundation for additional research.

Data & Methods

Data for all figures come from 1% IPUMS samples for each census from 1960 to 2000. ACS survey data were used for 2010. For 1970 and 1980, 1% metro fm1 and 1% metro samples were used, respectively. I defined the middle class with these criteria: family income was within 67% to 200% of the median family income for that census year, either the head of household or their spouse had an educational attainment no more than two grade levels below the median education for that census year for other heads of households and their spouses, and the family owned or was purchasing their home in the given year.  Income is the total income for individuals from the previous year and standardized to 1999 dollars using CPI99. Education is a person’s highest educational attainment. Median income and education was weighted with HHWT. All population statistics for all years are weighted with PERWT. Methods for each figure are described in their captions.

Replication code available here.

Results & Discussion

Figure 1 shows the percent of the total United States population considered middle class under my definition from 1950 to 2000. Home ownership could not be determined for 1950. As a result, the middle class population is a substantially larger portion of the total population.

Figure 1 shows the percent of the total United States population in each census year considered middle class under my definition from 1960 to 2010. The middle class population includes all individuals in middle class families and totally United States population is the national population in any given year.

Under my definition, the middle class is smaller than both the middle income population and the population who identify as middle class. Pew found that approximately 61% of Americans were middle income in 1970 and 51% of Americans were middle income in 2010 (Pew 2015,2). However, using my definition, the middle class was only 34% of the population in 1970, and 29% of the population in 2010. These numbers are also significantly smaller than the 92% of individuals who identify as middle or working class according to the government (Department of Commerce 2010,10). Middle class Americans are a much smaller portion of the American population then previous scholarship contended, but they are also a declining segment of the population. This finding supports previous middle income and middle class definitions, which describe the middle class in decline.

Broad decline in the middle class is not particularly revealing. I will now turn my attention to trends in occupation, race, nationality, income, education, and geography in the middle class. I found that the middle class, as I have defined it, is whiter than the overall population, and, although well-off and educated as a whole, growing more unequal in both income and education.

Figure 2 shows middle class Americans by occupation as a percentage of total middle class population from 1960 to 2010. Occupation are standardized to their 1990 equivalents using IPUMS-provided categories in OCC1990. Occupations categorized as ‘unknown’ are excluded from the graph as they likely represent children and other dependents.

Figure 2 shows middle class Americans by occupation as a percentage of total middle class population from 1960 to 2010. Occupation are standardized to their 1990 equivalents using IPUMS-provided categories in OCC1990. Occupations categorized as ‘unknown’ are excluded from the graph as they likely represent children and other dependents.

This figure shows the concentration of middle class jobs over the last fifty years into managerial/professional and services jobs. In 1960, these jobs accounted for only 27.6% of all middle class American occupations. By 2010, 41.8% of all middle class individuals had these occupations. Craft/production/repair and operators/fabricators/laborers in the middle class have declined substantially since 1960. By 2010, only 23.3% of middle class individuals held these occupations compared to 34.3% in 1960. Technical/sales/administrative jobs have remained a fairly constant middle class occupation for the last fifty years. This graph illustrates the middle class’ professionalization and its increasing barriers to entry over the last fifty years. Professional and managerial occupations require college-degrees or at least a high school diploma. Thus, one’s advancement into the middle class is now heavily dependent on access to and completion of quality education. On average, service workers and managerial/professional workers had completed at least some college by 2010. This point is illustrated in greater detail in figure 5.

The middle class’ professionalization corroborates Pew’s and the Department of Commerce’s findings that attaining a middle class lifestyle is more difficult now than in the past (Pew 2015, n.p. and Department of Commerce 2010, 23). In 2010, 73.1% of middle class occupations were managerial/professional, technical/administrative, or services. In 1960, they were only 61.3%. Many of these occupations require a college degree, and access to resources that disadvantaged groups may not have readily available.  In Figure 3, I explore the impact that professional barriers to entry have on the middle class’ diversity.

Figure 3 shows middle class Americans by race/ethnicity as a percentage of total middle class population from 1960 to 2010. Race/ethnicity was first determined using the HISPAN variable. If individuals were not Hispanic, then their race was determined using the IPUMS provided RACESING variable, which classifies each individual as only one race. Individuals classified as ‘Other’ are excluded.

The middle class’ professionalization and barriers pose difficulties for disadvantaged groups to gain middle class status. Figure 3 highlights this point. This figure shows the AMC’s lack of racial diversity. The percentage of middle class Americans who identify as white is declining, but not as quickly as the decline in the percentage of white Americans.  Although the AMC is slowly becoming more diverse, it remains an overly white-dominated classification. In 1960, 88.6% of Americans identified as white compared to 94.1% of middle class Americans. Thus, the diversity gap in 1960 was 5.5 percentage points (Hobbs and Stoops 2002, A-20). By 2010, 72.4% of Americans identified as white compared to 74.7% of the middle class (Humes et al. 2010, 4). The diversity gap has shrunk to only 2.3 percentage points over the last fifty years. However, most of the gains for minorities have come in the last ten years. In 2000, the diversity gap was 5.1 percentage points – nearly as wide as in 1960 (ibid).

Figure 3 supports Pew’s findings that African Americans and Asians have made substantial gains into the AMC over the last half century. The percentage of African Americans in the middle class under this definition has nearly tripled. The percentage of Asian Americans has increased by one order of magnitude. However, these gains are outpaced by overall population growth for both races (Pew 2016, n.p.). My analysis refutes Pew’s findings that the Hispanic population is not gaining in the middle class. Hispanic middle class Americans, as a percent of the middle class, increased nearly 700% from 1960 to 2010. As a percent of the total United States population, the Hispanic population only quintupled. Pew asserts that the Hispanic middle income population is on the decline. I found that their gains in the middle class are outpacing their gains in the overall population (Stepler 2016, 4).

Figure 4 shows the middle class by birthplace as a percentage of the total population from 1960 to 2010. Birthplace was determined using the IPUMS-provided categories for the BPL variable.

Middle class identity is not only heavily determined by race, but also by one’s birthplace. This figure illustrates that the AMC is not only racially homogenous, but also homogenous by nationality. The percentage of middle class Americans who were born in the United States has declined since 1960 and fell below 90% in 2010. However, unlike the middle class’ racial diversity, middle class nationality tracks closely with national trends. For example, in 1970, 4.7% of the population was foreign born compared to 3.2% of the middle class. By 2010, 11.3% of the population was foreign born compared to 10.1% in the middle class (Zong 2016, n.p.). The gap in foreign born population has shrunk slightly between national trends and middle class trends, but overall the AMC is more representative of the national immigrant population than of the non-white population. The literature reviewed in this paper did not examine middle class status by nationality. The findings here do not strongly support any notion that middle class nationality is significantly different from the country as a whole.

Figure 5 shows the AMC population by race, occupation, income, population, and education from 1960 to 2010. Median income is used because income is top-coded in most years. Shapes correspond to different occupations and colors correspond to different races. The size of each point corresponds to the number of people of that group in the AMC. Like Figure 2, this graph excludes individuals categorized with “unknown” occupations. The individual symbols on this graph are less important than the movement of the large cluster of groups over time.

Figure 5 combines findings on race and occupation in the AMC, and illustrates several critical economic trends in the middle class. First, middle class wage growth has been sluggish, especially since the 1970s. The largest cluster of occupations/races barely moves from 1970 to 2000. The median middle class income (in 1999 dollars) rose only $1,595 from 1970 to 2000, from $48,805 to $50,400. A stark contrast to the $11,679 increase from 1960 to 1970. Median income declined $5,294 from 2000 to 2010 due to continued sluggish wage growth and the Great Recession. This graph also illustrates how important education has become as a qualifier for a middle class lifestyle. In 1960, the data are much more spread out across the education axis. By 1990, almost all race/occupation pairings in the middle class have a high school education. The single largest grouping, white technical/sales/administrative workers in 2010, have, on average, completed at least one year of college. By 2010, the white, black, and Asian managerial/professional category has, on average, nearly graduated from college. Critically, increases in education clearly correspond to increases in income for each group by 2010.

This figure also illustrates racial disparity in income and education within the middle class. The income and education gaps between managerial/professional whites and Asians and essentially all other categories are substantial. Both of these groups made more than $50,000 on average in 2010. The next highest managerial/professional wage earners in 2010, Hispanics and blacks, both made more than $4,000 less, despite having the same average educational attainment as whites. This finding adds to Alba’s findings that black middle class communities are less wealthy overall than white communities (Alba 548). Not only do African Americans live in different neighborhoods (ibid), they are paid less, on average, for similar occupations despite attaining similar education levels to whites.

Alba’s findings of racial disparities extend beyond the African American middle class. Both managerial/professional white and Asian groups made above $50,000 in 2010, but, on average, the white group had completed one less year of college. To achieve similar standing to the white middle class in the managerial/professional world, Asians had to work harder and become more educated to make comparable income. In general, the graph shows that the white middle class required less education to enter occupations than other races, yet typically made more money.

Figure 6 shows the middle class population by state from 1960 to 2010. Original map data provided by Prof. Merchant.

This figure illustrates that the AMC has primarily lived in the Northeast and Midwest as far back as 1960. Since then, the major gains in middle class population have been in the Southeast. This reflects immense economic growth in the United States’ so-called “Sunbelt” (Glaeser 2007, 610). Geographical analysis of the middle class reveals that the gains in middle class population have not been particularly equal. Middle class concentration has important political consequences. The heavily white and native born middle class population is concentrated in the Midwest and Northeast. Years of stagnant wages, increasing middle class diversity, and higher education costs likely contributed to the nativist themes in both the Democratic and Republican party in the 2016 presidential election (Williams 2016, n.p.).

Conclusion

In this paper, I put forward a definition of middle class that bridges the gap between income-based and identity-based definitions. Contrary to analysis of the middle-income population, the AMC is only about one-third of the total population and has been on the decline since the 1980s. The middle class is whiter and more professional than the country as a whole. However, the middle class is about as native born as the whole nation. Middle class wage growth has been largely stagnant since 1970. There are also significant disparities in earnings and education. Whites tended to make more money and had, on average, less education for the same occupations as non-whites.

Stagnant wages, sustained occupational inequality, and the rising importance of education are not conducive to a thriving middle class or overall society. Easterly writes that middle class “consensus” is critical to a thriving society (1999, 2002). “Consensus” entails minimal ethnic divisions, low inequality, and high share of overall wealth. Data from this paper suggest the middle class has low ethnic divisions because it is fairly homogenous. However, the smaller minority middle class community must be more educated than their white counterparts and sometimes make less money. Findings in this paper also suggest growing income inequality in the AMC. Previous research also found that the middle class no longer holds most of the country’s wealth (Casselman 2016, Bui 2016, Pew 2015).  Thus, the middle class “consensus” achieved in the 1960s has deteriorated in the last fifty years.

[1] Some Americans distinguish between the middle and working classes in America. However, the Department of Commerce found views on the working class to be less consistent across surveys they conducted (Department of Commerce 2010, 3). I will not make a distinction between the middle and working class in this paper, but these divergent identities could be the foundation for additional research.

Works Cited

Alba, Richard, John Logan, and Brian Stults. “How Segregated Are Middle-Class African Americans?” Social Problems 47.4 (2000): 543-58. Google Scholar. Web. 2 Oct. 2016.

Bui, Quoctrung. “Where the Middle Class Is Shrinking.” New York Times. New York Times, 12 May 2016. Web.

Casselman, Ben. “Most Americans Aren’t Middle Class Anymore.” FiveThirtyEight. ESPN, 09 Dec. 2015. Web. 02 Oct. 2016.

Easterly, William. “The Middle Class Consensus and Economic Development.” Policy Research Working Papers (1999): n. pag. Web.

Glaeser, Edward, and Kristina Tobio. “The Rise of the Sunbelt.” Southern Economic Journal 74.3 (2007): 610-43. Web.

Hobbs, Frank, and Nicole Stoops. Demographic Trends in the 20th Century: Census 2000 Special Reports. Place of Publication Not Identified: Bibliogov, 2012. Census 2000 Special Reports. United States Census Bureay, Nov. 2002. Web. 16 Nov. 2016.

Humes, Karen, Nicholas A. Jones, and Roberto R. Ramirez. Overview of Race and Hispanic Origin, 2010. Washington, D.C.: U.S. Dept. of Commerce, Economics and Statistics Administration, U.S. Census Bureau, 2011. United States Census Bureau. Census Bureau, 11 Mar. 2011. Web. 2016.

NPR Staff. “A Portrait Of America’s Middle Class, By The Numbers.” NPR. NPR, 7 July 2016. Web. 10 Oct. 2016.

Pew Research Center. “The American Middle Class Is Losing Ground.” Pew Research Center’s Social Demographic Trends Project RSS. Pew Research Center, 09 Dec. 2015. Web. 02 Oct. 2016.

Stepler, Renee, and Anna Brown. “Statistical Portrait of Hispanics in the United States.” Pew Research Centers Hispanic Trends Project. Pew Research Group, 19 Apr. 2016. Web. 31 Oct. 2016.

United States. Department of Commerce. Economics and Statistics Administration. Middle Class in America. By Rebecca Blank. Office of the Vice President, Jan. 2010. Web.

Williams, Joan C. “What So Many People Don’t Get About the U.S. Working Class.” Harvard Business Review. Harvard Business Publishing, 10 Nov. 2016. Web. 15 Nov. 2016.

Wilson, Reid. “Watch the U.S. Transition from a Manufacturing Economy to a Service Economy, in One Gif.” Washington Post. The Washington Post, 3 Sept. 2014. Web. 24 Oct. 2016.

Zong, Jie. “Frequently Requested Statistics on Immigrants and Immigration in the United States.” Migrationpolicy.org. Migration Policy Institute, 26 May 2016. Web. 24 Oct. 2016.