All Policy Briefs

Policy Brief 28: Why, When, and How to Teach the Fundamentals of Inequality in Principles

PDF Version

Can Erbil and Geoffrey Sanzenbacher

August 2020

Why should we teach inequality?

The continued rise of economic inequality in the U.S. has been on the radar of economists and policymakers for the last few years. New initiatives and platforms like EfIP (Economics for Inclusive Prosperity) have emerged, where economists work on creating fresh and useful policy ideas to promote more “inclusive” models of generating prosperity. Yet, students still face a lot of confusion, “noise,” and disagreement when it comes to the topic of inequality.[1] Especially in the age of easy access to information online, students not fully literate in economics struggle to sift through the literature and separate rigorous scientifically based research from opinion pieces. This article discusses a way to introduce some recent trends in U.S. inequality to students, and then delve into a specific topic (e.g., racial inequality) in a two-lecture module at the end of an introductory economics course.

After all, it is not that the students are not interested in this topic. But, left alone, many of them become confused. Students do not make distinctions between different broad types of inequalities, such as income inequality, wealth inequality, or inequality of opportunity. They do not understand differences in the underlying causes of inequality, which vary depending on whether the topic is inequality across educational groups, gender-based groups, or racial/ethnic groups. And, they make basic mistakes in reading articles or viewing data, such as mixing-up correlation with causation or looking at graphs with different scales. Fortunately, introductory economics courses have a chance to reach students early in their careers, and correct many of these deficiencies.

Indeed, recent events in the US compel us to cover inequality and its consequences in our introductory classes. The economic impact of the COVID-19 epidemic falls significantly more on minorities, women, people with lower income levels and people of color.[2] And, the disparities in poverty and opportunity rates between black and non-black populations is at the root of racial discrimination which re-ignited the BLM movement.[3] Indeed, economic inequality is also related to other disciplines in Social Sciences, such as Sociology, Political Science and even Liberal Arts courses such as History and English. However, Economics faculty are well-equipped to cover the empirical outcomes and the underlying economic models – and lay out their strengths as well as short-comings.

Given that teaching about economic inequality is an important task, this article will first discuss when it makes sense to teach it. The article will then spend the majority of its time providing some tips for covering the material, as well as some suggested assignments.

When should we teach inequality?

Many of the introduction to economics courses are structurally very similar, focusing on delivering basic fundamentals and simple economic models with “representative agents.” With respect to inequality, a brief discussion often occurs at the beginning of the semester about the trade-off between efficiency and equity, and about how markets, left alone, deliver the most efficient outcome. The rest of the semester typically covers efficient economic models and the behavior of an average (representative) household or firm, without touching the relevant and important topic of income and wealth distribution and what happens to economic agents that are “not average.”

The lack of emphasis on inequality is partly due to the time constraint of the faculty who is trying to cover a very wide variety of topics (especially in introduction to economics courses where micro and macro are taught together in one semester) and partly due to how the course material has been written and structured. Recently, there have been open discussions between high profile economics faculty on how an introduction to economics course should be (re)structured.[4]

The good news is that during most introductory courses, students learn many tools that can help them understand the basics of economic inequality. For example, if the supply and demand for labor was covered, then the idea of productivity differences driving wage differences can be understood. As can discrimination, when identically productive workers are treated differently. And, if the idea of capital versus labor have been covered, wealth versus income inequality can be touched on. None of these ideas can be covered in a tremendous amount of detail, but the bones exist for a discussion.

Thus, our proposal is to add a short two-lecture module to the end of every introduction to economics course to provide basic information about the level and consequences of economic inequality in the US. The next section focuses on ways in which we have accomplished that in our classes.

How should we teach inequality?

This section lays out some of the basic logistics of the module, before turning to the meat of what exactly to cover. We conclude the discussion with some ideas for short empirical projects that can help students get their hands into data.

i) The Logistics and Design of a Module

This module is designed to be delivered in two 45-50 minute lectures. The first lecture will focus on several broad facts that describe the economy over the last four decades, the second will yield more detail on one of these facts, as chosen by the instructor. The lectures will not be a replacement for more in-depth “economics of inequality” courses, but rather a “preview”. Students who want to learn more on this topic will be encouraged to take the corresponding elective course in their department to advance their knowledge in this field. The module will equip the students with the necessary tools to navigate through available information and current (elementary) research in the field.

The following are some of the essential design properties of the proposed module:

  • The module needs to be elementary. Since it will be part of an introductory class, the faculty cannot assume prior economic knowledge beyond that covered in the class.
  •  It needs to be data driven and based on sound economic research. Faculty will curate carefully selected, up-to-date research on the topic.
  • It needs to be relevant to students.
  • It needs to contain a hands-on component (we have suggested exercises below), exposing intro-level students to cutting edge, applied empirical research.

ii) The Structure and the Content of Teaching Material

One of the biggest problems faced by an instructor teaching about inequality is what to cover in the first place. Students will jump to different places on hearing the word inequality. Some might think immediately about CEO pay, others about racial or gender inequality, and others about issues related to the fact that high-income kids go to higher-quality schools. It can therefore be helpful to begin with a quick discussion of what students think of when they hear the word “inequality.”

This discussion can be followed by the introduction of the concepts of inequality in economic outcomes versus inequality in opportunity. The gender wage gap or the growth of the income share going to the top 1 percent represent inequality in economic outcomes. The fact that parental income is correlated with child’s income is an example of inequality in opportunity. By the last two lectures of your intro class, students will probably be better equipped to think about inequality in outcomes – but many students default to the issue of opportunity. The discussion can help tease this issue out.

The next challenge is to cover some of the major inequalities students will encounter in the outside world. We suggest that during the first class you can accomplish cover some of these through the presentation of “stylized facts” – facts about the economy that are not controversial, even though the underlying causes and solutions to the inequality might be. For data reasons, it is often easiest to focus on the period since the mid-1970s, and we usually focus on the U.S. Four facts on economic outcomes and one fact on opportunity can introduce students to a world of economic thought.

  1. The median male worker has not seen any real wage growth since the mid-1970s, while workers slightly higher up in the distribution (e.g., the 80th percentile) have seen growth.[5]
  2. The median woman, working full-time and full-year, still only makes 80 percent as much as the median male, with little progress in the last two decades.[6]
  3. The median Black worker has seen no progress in terms of earnings relative to the median white worker since the 1970s.
  4. Very high-earning workers (often defined as the top 1 percent) have seen a large growth in income, much more than those around the 80th percentile and much, much more than those at the median.
  5. A strong correlation exists between parent’s income and child’s educational attainment and income.

These five facts can be displayed quickly during the first class of the module in just five slides if desired (we have provided example slides with some sources in the Appendix). However, preceding these slides with a few key pieces of information can provide context.

First, depending on the mathematical prerequisites of the class, it can be useful to proceed the discussion with some information on measuring inequality. A primer on percentiles at least can be useful.

Second, a discussion of economic data can be useful. Data on workers from the main part of the income distribution can often be obtained from household surveys (we used the Current Population Survey) and discussion of these sorts of sources can be informative. For example, the CPS is a household-level survey of addresses, and ignores institutional populations. This precludes the data from allowing an analysis of inequality in incarceration, and the data will understate inequality in employment, since those in prison by definition are not working.

On the other hand, data on the Top 1 percent often come from estimates based on national accounts (we used the World Inequality Database), and students should know why a household survey is insufficient. In particular, pointing out that these household surveys are statistically unlikely to capture very high earners, and in any case top-code data would be important. Finally,

data on economic mobility will likely come from Opportunity Insights, and a discussion of the administrative data they use can also be helpful. Again, we have included some example course material on each of these points. At the end of the first class, it can be useful to have students chime in on what they think might be driving the facts discussed, and if any of the facts were particularly surprising.

For the second lecture in this two-lecture series, we recommend focusing on a single one of the facts of the instructor’s choosing. Focusing on a single fact allows an important nuance of research on inequality to shine through — economists do not always agree on causes. Using the lecture to present a few cutting-edge explanations can be more fruitful than providing a little detail on each fact.

Turning to what to cover, we have recommendations based on experiences teaching a full course on the Economics of Inequality in the U.S.

  1. Falling Median Earnings among Men — The most common explanations for the absence of earnings growth for middle-income men are: 1) technological change/automation; 2) trade; and 3) declining bargaining power. If students have been exposed to the model of supply and demand for labor, then technology can be framed as having two effects. First, it increased the productivity and thus demand for higher earners more than lower-earning workers (“skill bias”). Second, it has tended to substitute and thus reduce demand for middle-income workers specifically (polarization).  Case studies from Fernandez (2001) and Autor et al. (2002) can provide examples for class, and Autor and Dorn (2013) can yield some empirics on the effect of automation.  On trade, the effect can be summarized most easily as a decrease in demand for domestic middle- and low-income workers who are more prevalent among our trading partners. Empirical work from Autor, Dorn, and Hanson (2016) providing some useful evidence from trade with China.  Talking about bargaining power can be the hardest in an intro class.  However, pointing out that unions negotiate for higher wages and providing data on the union wage premium and trends in unionization can make the point.  Recent work by Farber et al. (2018) is especially useful.
  2. Gender Wage Gap — When teaching about the gender wage gap, it can be useful to talk about two phases in time: 1) when it closed from 60 to 80 percent between the 1970s and 2000; and 2) the period since then, what the gap has been stuck at 80 percent. The closing of the gap was fueled mainly by increased education and work experience among women. The effect can broadly be framed as an increase in productivity due to “Human Capital”, or a boost in labor demand.  Providing data from the census on women’s education and labor force participation is useful.  Goldin and Katz (2002) on the Power of the Pill can help students see an interesting cause of educational change, and provides a nice way for them to think about how technology can alter the costs and benefits of education.  Greenwood, Sehardi, and Yorukoglu (2005) can give good insight on how technological change in the household allowed married women to work more.  Regarding the remaining gap, women’s responsibilities as caregivers are the most common Goldin (2014) attributes this to compensating differentials in a way that can be intuitive to students. And, using BLS data on the labor force participation of young mothers to young fathers can make it clear there is a work experience gap that develops around motherhood.
  3. The Racial Wage Gap – Here, discrimination and the Human Capital gap are important topics to hit. In an introductory course, pointing out that discrimination occurs when workers are paid less than their marginal product is the natural point to start. Empirically, this definition means workers with similar productive characteristics in similar jobs should get paid the same amount regardless of race. Covering a few studies can be helpful to show the existence of discrimination, but also the trouble economists go through to identify it. Bertrand and Mullainathan (2004) is a classic experimental example, and Fryer, Pager, and Spenkuch (2011) a nice regression example. Turning to the Human Capital gap, using Census Data to show racial graduation rates and the difference in earnings between graduates and non-graduates is a good place to start. The discussion can then be extended to get into some of the underlying reasons for the Human Capital Gap. A common focus point might be ability to pay as captured by loan inequality, and Scott-Clayton and Li (2016) have some interesting data. Housing discrimination, residential segregation, and school quality could involve a more detailed discussion. A recent Newsday study out of Long Island by Choi et al. (2019) provides recent examples of housing discrimination and the resulting exclusion out of neighborhoods that might have higher quality schools.
  4. The Rise of the Very Rich – When discussing the rise of incomes for the top 1 percent, a key distinction to have your students make is between labor and capital income.  As Piketty, Saez, and Zucman (2019) point out, the rise in top income over the last several decades has been driven by both income sources, albeit at different times.  Through the 1990s, it was mostly growth in labor income. The underlying causes of growth in labor income are not fully understood, so we usually give two examples of causes. One is so-called “Superstar Theory,” namely that technology allows people with marginally higher productivity to capture larger shares of markets than in the past. Gabaix and Landier (2008) provide a nice discussion in the context of CEO pay.  The other is a potential failure of corporate governance, which has led to rising executive salaries often paid in the form of equity. More recently, Piketty, Saez, and Zucman point out that capital income growth has taken over.  Here, one main point worth making is that capital is extremely concentrated, and has become even more so, as in Saez and Zucman (2016). If you have time, highlighting the growth in the capital share that may occur as a result of slowing growth in developed countries — as highlighted in Piketty’s Capital — can give students a sense of the potential for future growth in inequality.
  5. Opportunity – If focusing on inequality of opportunity, the first thing you want to establish is a definition of what equality of opportunity would look like. Having a discussion about what is within a person’s control and what is outside of it is key. In the opportunity literature, what is within control is often referred to as “effort” and outside of it “circumstances” (see Roemer and Trannoy, 2016). Discussing with students the kind of things they think of as circumstances is important, as is discussing the extent to which “effort” is really within someone’s control. For example, low-income kids tend to put less time into schoolwork, but more time into paid work (e.g., see Porterfield and Winkler, 2007). Would we say they are exerting less effort towards school work when their time must be split? Students will see that it’s complicated. In the end, when we say we want equal opportunity, we usually mean that circumstances do not dictate economic outcomes. So, parents’ income should not affect a child’s propensity to go to college or their income as an adult. Once these definitions are laid out, using data from OpportunityInsights.org can be used to illustrate the lack of Equality of Opportunity within the U.S. We have provided an example module in the Appendix, courtesy of the researchers at Opportunity Insights.

iii) Examples of Empirical Research Projects

The assignments below are designed to accomplish two things. First, they all expose students to newspaper articles, think tank pieces, and perhaps some academic writing on the topic. Second, they force students to work with data, and produce a visualization to make their arguments. Below, we lay out one assignment for each of the four facts on economic outcomes from the first lecture. We recommend picking the fact you focus on from the second lecture. If you focus on opportunity, we have included an assignment from Opportunity Insights in the Appendix.

  • Example Assignment: Occupational Wage Growth
    • To the Instructor: This assignment gives the students a chance to see the effects of Skill-biased Technological Change and/or Polarization first hand. The goal of the assignment is to have the students identify one occupation that has been negatively impacted by automation, and one that has benefited from technology.  In completing the assignment, the student will be exposed to articles on economics, and will practice displaying data visually.
    • Example Assignment Prompt: As we learned in class, median incomes for full-time workers have been stagnant over the last several decades, but at higher points in the distribution they have risen. Two related explanations put forward by economists to explain this phenomenon are Skill-biased Technological Change and Polarization. In this short-paper assignment, your goal is to find evidence of: 1) a middle-income occupation where automation has tended to replace workers (Polarization); and 2) a higher-income occupation where technology may make workers more productive.
    • Example Assignment Details: The paper should include at most 2 pages of text 1.5 spaced, as well as one data-based figure.  As evidence for the two occupations you choose, you should one high-quality source each (i.e., two sources total).  These sources can include government websites, academic papers, or policy briefs from non-partisan think tanks (e.g, The Pew Charitable Trusts, The Brookings Institute).  The figure should show the trajectory of median earnings within the occupation over the last several decades, to see if the damaging or improving aspect of technology argued in your sources play out in the data (it is fine if they do not, although you may want to hypothesize as to why).  The figures do not count against the page limit, and should include clearly marked labels on the axes, a legend if necessary, and a clear title.
  • Example Assignment: The Gender Wage Gap…Closing then Stalling
    • To the Instructor: The goal of this assignment is to expose students to economic articles and data, and familiarize them with constructing data visualizations in the context of the gender wage gap.  In general, students will likely find that the closing of the gender gap during the 1970s, 1980s, and 1990s was driven by increased education, increased work experience, and/or a decline in occupational segregation.  Explanations for the remaining gap often center on child care responsibilities (either by reducing work experience or by introducing a compensating differential for flexibility) or on outright discrimination.
    • Example Assignment Prompt: Even today, the typical woman working full time makes about 80 cents for every dollar made by a full-time working man.  However, this gap was larger in the past.  In 1975, the median full-time working woman made about 58 cents for every dollar a man made.  In this short paper assignment, your goal is to offer: 1) an explanation as to why women have partially caught up to men; and 2) an explanation as to why women’s income still lags behind men’s. You should make an effort to tie your explanations to concepts learned during the semester.
    • Example Assignment Details: The paper should include at most 2 pages of text 1.5 spaced, as well as two data-based figures, one for each argument.  To support your arguments and provide data, you need at least two high-quality sources.  These sources can include government websites, academic papers, or policy briefs from non-partisan think tanks (e.g, The Pew Charitable Trusts, The Brookings Institute).  The figures should augment the argument being made and can be drawn solely from the two sources required or from additional sources.  For example, if you are arguing women make more because they enter high-paying STEM fields at a higher rate today than in the past, your figure might show the share of men and women majoring in STEM majors over time.  The figures do not count against the page limit, and should include clearly marked labels on the axes, a legend if necessary, and a clear title.
  • Example Assignment: Identifying Discrimination
    • To the Instructor: The goal of this assignment is to expose students to the definition of discrimination, and to how researchers go about identifying it. Students will be asked to find an article from a government source, think-tank, or academic journal on discrimination. They will then be asked to determine how the authors determined some effect (often lower income, but possibly something else like housing inequities) was discrimination. The assignment also allows the students to construct a data visualization.
    • Example Assignment Prompt: Discrimination occurs when two people of nearly identical characteristics receive differential because of some group characteristic (e.g., race).  For example, a Black person with the same productive characteristics may be paid less than a similar white person. Identifying discrimination is difficult, because it requires comparing two otherwise similar individuals. Find a high-quality source purporting to identify discrimination, discuss how the authors attempted to identify, and argue whether or not you think they did a good job.
    • Example Assignment Details: The paper should include at most 2 pages of text 1.5 spaced, as well as one data-based figure illustrating the nature of the discrimination.  Your source can include government websites, academic papers, or policy briefs from non-partisan think tanks (e.g, The Pew Charitable Trusts, The Brookings Institute).  The figure does not count against the page limit.
  • Example Assignment: The top 1% income inequality
    • To the Instructor: The goal of this assignment is to expose students to economic articles and data, and familiarize them with constructing data visualizations in the context of income inequality at the top 1% level. The top 1% of American earners have nearly doubled their share of the national income since the 1970s, according to Saez’s analysis. Americans in the top 1 % average over 39 times more income than the bottom 90% and 85 times as much as the bottom 20% (CBO data). Even the after-tax income of the top 1% has been growing much faster than the rest, making this enormous gap even bigger.
    • Example Assignment Prompt: There is a significant gap between the incomes of the top 1% and the median income of the American population. This discrepancy is also present, in varying degrees, across the US. This assignment will compare the income of a top 1% earner in your own county with the income of a median earner in your home county (or, if from abroad the county where your college sits). In this short paper assignment, you should: 1) determine the income gap between the 1-percent income in your county and the median; 2) determine how the median differs by race and gender within your county; 3) determine how a 1 percenter would need to spend their money in a day so that they did not accumulate any additional wealth (see DeLong (2019) for example); 4) determine how many minimum wage workers it would take in your county to hit the 1 percent income; and 5) analyze and critique a policy platform that attempts to address this sort of inequality.
    • Assignment Details: The paper should include at most 2 pages of text 1.5 spaced.  You should also construct a data-based figure comparing income for the 1 percent in your county to the median Black female, white female, Black male, and white male.  The figures do not count against the page limit, and should include clearly marked labels on the axes, a legend if necessary, and a clear title. For this assignment you will need utilize the following data sources:
      1. Economic Policy Institute
      2. Opportunity Atlas
      3. Economic Policy Institute – Minimum Wage Tracker

Conclusion

Teaching about inequality in an introductory course is both imperative and difficult. We believe by focusing first on some agreed upon facts, and then focusing in on one of those facts you can achieve two goals. First, you can give your students some experience with data on inequality, and exposure to some statistics in a controlled environment. They may not agree on whether or not inequality is a problem, or what to do about it, but they will at least know the relevant trends and the current state of the world. Perhaps more importantly, you can show them how economics can help them think about one particular source of inequality. Hopefully the second lecture can push some students to think about economic policy options and how economics can be used for inclusive prosperity.

Selected Resources, Readings and Data Sources to Build Your Own Module

There are many other excellent, publicly available sources for faculty to utilize when they are building their own modules. Some of the ones below may be helpful as readings an assignment to students. These are selected with considering that the students are in an introductory class and don’t have much prior background in economics. We found them to be accessible, insightful and useful for a module like the one proposal in our policy brief.

See the appendix of the attached PDF for an example of how such a module can be successfully integrated into a large introduction to economics course. The complete set of course material can be found here.

  1. CORE-US – Chapter 19 – Economic Inequality This is an excellent source for anybody who wants to cover economic inequality in an introduction to economics course. It provides many teaching tools and also addresses difficult questions like “what (if anything) is wrong with inequality?” and “how much inequality is too much (or too little)?” – these are questions most of the introduction students are struggling with. Economic analyzes as well as empirical data are utilized to answer each question.
  2. Economic Policy Institute
  3. Washington Center for Equitable Growth
  4. https://inequality.org
  5. Combating Inequality: Rethinking Policies to Reduce Inequality in Advanced Economies, Peterson Institute for International Economics Webcast, 2019
  6. How economic inequality harms societies, TED talk by Richard Wilkinson, 2011
  7. The Triumph of Injustice, by Gabriel Zucman, March 20th, 2020, Social Europe, podcast
  8. Tackling Inequality from the Middle, Dani Rodrik, Project Syndicate, 2019
  9. Why some countries grew rich, Suresh Naidu, 2015
  10. Good and Bad Inequality, Dani Rodrik, Project Syndicate, 2014
  11. The Inequality Trust
  12. The Spirit Level: Why More Equal Societies Almost Always Do Better, Kate Pickett and Richard Wilkinson, 2009 – Powerpoint slides
  13. Suresh Naidu on Capitalism, Monopsony, and Inequality, MindScape Podcast, 2020
  14. Should we worry about income gaps within or between countries?, Dani Rodrik, Social Europe, 2019
  15. The New Economics: Data, Inequality and Politics, 2019, John Cassidy, The New Yorker
  16. Inequality is
  17. https://www.millionairesforhumanity.com
  18. How Pandemics Leave the Poor Even Farther Behind, IMF Blog, 2020
  19. https://www.project-syndicate.org/commentary/inequality-data-and-denialism-by-facundo-alvaredo-et-al-2019-12?barrier=accesspaylog
  20. https://reason.com/2020/01/25/the-truth-about-income-inequality/
  21. https://www.bloomberg.com/opinion/articles/2020-01-07/economists-debating-piketty-care-too-much-about-inequality
  22. https://www.wsj.com/articles/the-truth-about-income-inequality-11572813786
  23. https://www.barrons.com/articles/the-pandemic-is-weakening-americans-retirement-security-51588086298
  24. https://progressless.org/2020/05/21/the-covid-19-recession-versus-the-great-recession-in-one-chart/

Endnotes

[1] Disagreements stem from disagreements over trends in inequality to whether or not inequality should be a focus of discussion. For example, see articles by Piketty, Chancel, Alvaredo, Saez, Zucman (2019), Henderson (2020), Bershidsky (2020), or Gramm and Early (2019). We even had a student e-mailing me a link to a YouTube video of Milton Friedman from 1979 as a
counterpoint, without really understanding the point being made.
[2] See Kapadia (2020), Sanzenbacher (2020), and Fisher and Bubola (2020) for some analyses that illustrate these points.
[3] Chetty, Raj, Nathaniel Hendren, Maggie R. Jones, and Sonya Porter (2020).
[4] Matthews (2019) – The radical plan to change how Harvard teaches economics
[5] One note: keeping the fact simple and focusing on males eliminates many other moving parts, and highlights the deteriorating position of workers who have not seen many other fundamental changes (unlike women, who achieved higher levels of
education and work experience over this time period).
[6] Again, focusing on those working full-time and full-year eliminates the moving part of growing hours.

All Policy Briefs