Skip to main content

A Consumption-Based Definition of the Middle Class


Research on the middle class has typically defined middle class membership in terms of income. In this paper, we develop a consumption-based measure of the middle class that closely follows economic theory of constrained optimization. Overall, we find that only 55% of those considered middle-class by income are also classified as middle-class by consumption, with the remaining 45% divided between the consumption working class (34%) and the consumption upper class (11%). Put differently, a sizable share of Americans—16% of the overall population—are characterized as middle-class but consume like they are working-class, with little capacity to save. We find substantial differences in the demographic makeup of the consumption-based middle class compared to the income-based middle class. Notably, fewer Black and Hispanic Americans are included in a consumption-based measure of the middle class, reinforcing distinctions between income and wealth drawn by prior literature.

This is a preview of subscription content, access via your institution.

Fig. 1


  1. See, for example,

  2. In this work, while the terms “working class” and “lower class” are equivalent, working class is our preferred term. In the US, “working class” typically refers to the socioeconomic status below the middle class, and households in this category may be more likely to self-identify as working class than lower class (e.g. see Bird & Newport, 2017; Jones, 2022; Reeves et al., 2018). We do, however, acknowledge that households in any class may be working.

  3. In cases where lower-income households may be spending a small proportion of their income on necessities, but where that necessity spending is lower than the poverty line, we argue that their spending is insufficient to access a middle-class lifestyle. There is also a small share of households that report very low income and consumption (~ 1% of households in 2019); these outliers may be the result of underreporting of income or consumption, or other circumstances that we cannot discern from the data. By removing households that exhibit consumption poverty, we also exclude these households from the middle class.

  4. A comparison of households with missing vs. non-missing consumption data reveals no statistically significant differences between the two groups by race, age, college degree, having children, and after-tax income. Respondents with missing consumption data are only 0.4% points more likely to be Black and 1.1% points less likely to be college-educated. After-tax incomes are 2.5% smaller.

  5. Within the consumption definition, a breakdown of the share of households in the working and upper class is presented in “Appendix Fig. 2”.

  6. Note: the modest fluctuations in this series are due to differences in sample weights in each year; the series holds exactly at 50% if sample weights are not applied.


Download references


This research was conducted with support from the Rand Corporation’s Lowy Family Middle-Class Pathways Center.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Melissa Haller.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


Appendix 1

See Table 5 and 6.

Table 5 Components of necessity spending from the consumer expenditure survey data
Table 6 Representation of the working and upper class by definition, 2019

See Fig. 2.

Fig. 2
figure 2

Percent of US Population in Consumption Categories, 2004–2019. A All US Households, B US Working Age Households

Appendix 2

Details on Materials and Methods

3.1 Defining the Consumption Middle Class

While most of the households within the consumption-based middle class likely resemble households captured by other definitions, our definition may also include higher income households which spend a large proportion of that income on necessities. However, we do not exclude them from our definition of the middle class on the basis of their higher income.

We do, however, further differentiate between a more traditional middle class and a “flourishing” middle class to capture the heterogeneity of these middle-class groups (see Lacy, 2007 for a discussion of this distinction across Black households). The middle class is the segment of the consumption-based middle class that more closely resembles the household income and consumption patterns we might expect from the middle class. While the flourishing middle class is more financially stable; these households sit on the boundary between the middle and upper classes. To create this group, we separate the middle class in the top quartile of necessity consumption (regardless of income) into the “flourishing” middle class. In other words, we examine the distribution of necessity consumption spending, and classify those who are at or above the 75th percentile as part of the flourishing middle class. In 2019, this represented approximately $55,000 in yearly necessity spending.

For the broader consumption middle class definition, it is important to note that we do not distinguish between the specific allocation of goods consumed. It may be the case, for example, that one household spends all of its income just to afford rent and food, while another household spends all of its income to send its child to a high-priced private school. Some might argue that the second household should not be considered working class, but we believe it should for two reasons: first, although education spending is an investment in the child's future, our definition is only interested in the quality of the household’s life today. Our data does not enable us to trace households across generations. Second, without any leftover to accumulate savings or to allocate towards luxuries, these households better fit our definition of the working class.

Note that the 40–90th percentile thresholds were chosen strategically: we chose a wide initial band to capture as close to half of US households as possible. Other income-based estimates of the middle class typically find that it makes up about 50% of US households; setting up our measure in this way allows us to directly compare how these similarly sized populations of households differ. In this way, we can more directly assess whether households with middle-class incomes also have access to what we think of as a middle-class lifestyle.

3.2 Pew Household Adjustments

Using Pew's definition, we account for household size using two steps: first, we adjust household income by the square root of the number of household residents:

$$Adjusted \;Household \;Income = \frac{Household\; Income}{{\left( {Household\; Residents} \right)^{0.5} }}$$

Unlike a per capita measure of income, this measure is an "equivalence scale" calculation that recognizes that there are economies of scale in consumer expenditures that are increasing in household size. The 0.5 parameter value was set by the Pew Research center. Once we perform this calculation, we then scale each household so that it is roughly equivalent to a three-person household, which is the average household size in the US.

$$3\; Person\; Household\; Income = Adjusted\; Household\; Income*3^{0.5}$$

By adjusting our income data in this way, we approximate Pew’s middle-class measure. For example a one-person household earning $50,000 would be adjusted up to $87,000, while a family of four making $80,000 would be adjusted down to $69,000. For simplicity, we do not further adjust our measure for regional differences in the cost of living, although Pew has made this adjustment in its analyses.

Rights and permissions

Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Haller, M., Wenger, J.B., Zaber, M.A. et al. A Consumption-Based Definition of the Middle Class. Soc Indic Res 164, 1249–1270 (2022).

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Middle class
  • Income distribution
  • Household consumption