Inequality in the United States PowerPoint Presentation

Slide 1 -

US Inequalityalong numerous dimensions Gaetan Lion, June 1, 2022

Slide 2 -

Introduction Objective Studying trends in US inequality along several social dimensions including education, ethnicity, percentiles, and work status. We don’t explore gender because it is not disaggregated within the mentioned data that focuses on families (fairly similar to households). Data source US Government Survey of Consumer Finance (SCF) data. The SCF aggregates financial data on US families every three years. And, it discloses a time series from 1989 to 2019. Semantics I use the data terms, including ethnic groups, used by the SCF. 2

Slide 3 -

3 A first measure of Inequality: Between-Difference Between-Difference This is the difference between two different categories. Let’s say the median income of college graduates is $100K vs $50K for high school graduates. Here we would measure this between-difference by observing the multiple between the two: $100K/$50K = 2. In other words, college graduates median income is twice the one of high school graduates. Observing this multiple over time reveals if inequality between the two is rising or not. Here, we focus on the median because we want to remove the noise associated with the average or the mean when including very high income individuals & families (maybe we could call that the Elon Musk effect).

Slide 4 -

4 A second measure of Inequality: Within-Difference Within-Difference This is the difference within the same category. Let’s say college graduates have a median income of $100K and a mean income of $200K. Here we would measure this within-difference by observing the Mean/Median multiple: $200K/$100K = 2. In other words, college graduates average or mean income is twice the college grads median income. Observing this multiple over time reveals if inequality within college graduates has increased or not over time. Contrary than for the between-difference, with the within-difference we purposefully want to capture the Elon Musk effect since it is representative of the inequality within college graduates. By the way, just to make sure … Elon Musk indeed graduated from college. We know it is not always true of billionaires (Mark Zuckerberg did not).

Slide 5 -

5 The mechanics of Within-Difference. Exploring the Mean/Median multiple We start with a very simple model that splits a population into 5 different quintiles. All quintiles have a symmetric distribution. They range from $25K to $125K in income. They are space $25K apart. And, this population has both a median and mean income of $75K. The resulting Mean/Median multiple is equal to 1. Next, we sensitize the income of the 4th and 5th quintile in order to increase the Mean/Median multiple. In all cases, the median income would remain unchanged at $75K. But, the mean income would be the specified multiple higher. The main point of this model is to illustrate how high the upper quintiles have to reach in order for the mean/median multiple to reach a level of 2 to 4. As we will see this multiple level is common in the data.

Slide 6 -

6 Inequalities studied

Slide 7 -

7 Definitions

Slide 8 -

8 Net worth section At times, we will compare the behavior of change in difference multiples over time with stock market movements. I hypothesized there would be a fairly close correlation between the two (as the stock market rises rapidly, I would have expected measures of inequality to rise too … and vice versa). But, as we will see … at least visually (looking at time series graph), this does not appear to be the case.

Slide 9 -

9 Net worth. Overall. All families This within–difference jumped from a mean/median multiple of 4.6 in 2007 to 6.4 in 2010. Contrary to what we expected, it was not due to a rise in the stock market that remained flat when going back to 2007 and going forward to 2010 (on an inflation adjusted basis). In between those two years, it experienced a severe Bear market (Housing Bubble & Great Recession).

Slide 10 -

10 Net Worth. Education. College vs. High School graduates Another between-difference that is pretty high, between 4 and 5 x since 2001. But, the rise in this between-difference do not appear to visually correlate with stock market movements.

Slide 11 -

11 Net Worth. Education. College graduates This within-difference among college grads is as big as the between-difference between college and high school grads. This suggests that college majors and related career paths make a huge difference in net worth accumulation. However, again the rise in this within-difference does not seem to correlate closely with stock market movements.

Slide 12 -

12 Net Worth. Ethnicity The between-differences between White and Black and White and Hispanic are huge. However, both between-differences have declined markedly since their respective 1989 levels. Also, they have dropped rapidly since 2013. Thus, even if contemporary data shows large differences. The contemporary trends are positive as such differences are declining.

Slide 13 -

13 Net Worth. Ethnicity - White This within-difference among White is surprisingly high and rising. Yet, again this within-difference does not appear to be closely correlated with stock market movement. Look at the period from 2010 to 2019, where this within-difference remained pretty much flat around 5 x. Meanwhile, the stock market, even after adjusting for inflation, went on a rapid Bullish run.

Slide 14 -

14 Net worth. Percentiles of net worth The between-differences between the high percentiles (75th and 90th) vs. the low one (25th) are huge as can be expected. Both those between-differences jump between 2007 and 2010, which we know was not related to stock market movements as shown earlier. * The table discloses the precise definition of the percentile buckets. We used the 75/25 and 90/25 difference specification as a short cut to render some portion of the data table and the graph legend more parsimoniously readable.

Slide 15 -

15 Net worth. Percentiles of net worth, 75th percentile Here is the first difference of any kind that actually shows very little difference. Indeed, the Mean and the Median of this 75th percentile have remained closely aligned around a multiple close to 1 over the entire 1989 to 2019 period.

Slide 16 -

16 Net worth. Percentiles of net worth, 90th percentile The within-difference among the 90th percentile has remained relatively low and stable around 2 x during the entire 1989 to 2019 period. This low multiple does not signify that there could be some material differences within this group if we looked at more granular data (remember the earlier slide ‘The Mechanics of Within-Difference’).

Slide 17 -

17 Net worth. Work status Self-Employed vs. Employee Another pretty large between-difference that is rather expected. Business owners accumulate more net worth capital than employees to pretty much remain in business. However, again the trend in this between-difference multiple does not track the stock market.

Slide 18 -

18 Net worth. Work status. Employee This is another large and rising between-difference. But, again it does not closely visually correlate with stock market trend (again focus on the 2007 – 2010 period; also look at the 2010 – 2019 period).

Slide 19 -

19 Net worth. Work status. Self-employed Notice that this within-difference among the self-employed is rising faster and is much larger than the between-difference between self-employed and employee reviewed on the prior slide. Similarly, the rising trend in the self-employed within-difference does not correlate closely with stock market trend (again look at 2007 – 2010, also 1995 – 1998).

Slide 20 -

20 Pre-tax income section

Slide 21 -

21 Pre-tax income. Overall. All families This is a very steady within-difference remaining within a very narrow range of 1.5 to 1.9 times since 1989. Notice how much lower this within-difference is compared to the same within-difference when focused on Net Worth.

Slide 22 -

22 Pre-tax income. Education. College vs. High School graduates This also a rather steady between-difference around 2 x. Notice how much lower is this between-difference than the same one focused on Net Worth. Notice that either measure of pre-tax income has not materially changed between 1989 and 2019, on an inflation adjusted basis.

Slide 23 -

23 Pre-tax income. Education. College graduates This is another rather steady within-difference, that is also much lower than its counterpart focused on Net Worth. Again, these measures have not increased much between 1989 and 2019 on an inflation adjusted basis.

Slide 24 -

24 Pre-tax income. Ethnicity Notice how these between-differences are so much lower than their counterparts focused on Net Worth, reviewed earlier.

Slide 25 -

25 Pre-tax income. Ethnicity - White This is another within-difference that is relatively low and steady, and certainly much lower than its counterpart focused on Net Worth, reviewed earlier.

Slide 26 -

26 Pre-tax income. Percentiles of income These Pre-tax income between-differences are still relatively high. But, they are a lot lower than the same between-differences focused on Net Worth.

Slide 27 -

27 Pre-tax income. Percentiles of income, 80th percentile Within this percentile bucket there is pretty much no within-difference. Over the entire period, the Median and the Mean remain very close to each other so the related Mean/Median multiple is always very close to 1 time. Remember, that the situation was similar when looking at this same within-difference focused on Net Worth.

Slide 28 -

28 Pre-tax income. Percentile of income, 90th percentile This is another relatively low and steady within-difference. And, it is not all that different than the same within-difference when focused on Net Worth.

Slide 29 -

29 Pre-tax income. Work status This between-difference is surprisingly small given that its counterpart when focused on Net Worth was so much larger.

Slide 30 -

30 Pre-tax income. Work status. Employee This is a relatively stable and low within-difference. Again, it is surprising given that the same within-difference was so much larger when focused on Net Worth.

Slide 31 -

31 Pre-tax income. Work status. Self-employed After an abrupt drop between 1989 and 1992, this within-difference overall trend suggests a slow increase. Notice that the self-employed pre-tax income has virtually not increased from a begin- to end-point basis at around $75K in $2019 dollars.

Slide 32 -

32 Stock holdings section The differences, as measured, are a lot greater when focusing on Stock holdings instead of Net Worth and Pre-tax Income. This is an instance where we expected that long term trends in the reviewed between- and within-differences would track stock market movements. But, for the most part we could visually see that these differences did not track the stock market.

Slide 33 -

33 Stock holdings. Overall. All families See the steady rise in this within-difference. In 2019, the Mean stock holdings is nearly 10 x greater than the Median. However, the rise in this within-difference does not correlate closely with the stock market trend.

Slide 34 -

34 Stock holdings. Education. College vs. High School graduates This is another between-difference that does not track the stock market (see 2007 – 2010 and 2010 to 2019).

Slide 35 -

35 Stock holdings. Education. College graduates This within-difference trend over time has a U shape that is completely different than the stock market trend.

Slide 36 -

36 Stock holdings. Ethnicity 1989 White/Hispanic between-difference is very much an outlier. Taking this outlier out, the trend for both ethnicity match-ups sort of zig-zag up and down without a clear direction.

Slide 37 -

37 Stock holdings. Ethnicity - White This within-difference has really two periods. It is rather flat from 1989 to 2013. And, then it goes on an accelerated rise from 2013 to 2019. Yet, it often diverges from stock market trend (see 1989 – 1998, 2007 – 2010).

Slide 38 -

38 Stock holdings. Percentiles of net worth These between-differences are huge, as expected. Notice how the 75/25 multiple has peaked in 2001. Meanwhile, the 90/25 has kept on rising beyond its 2001 peak over the 2010 – 2019 period. These trends do not much correspond to trends in the stock market as shown on earlier slides.

Slide 39 -

39 Stock holdings. Percentiles of net worth, 75th percentile This within-difference has remained very steady around 1.3 to 1.4 since 1998.

Slide 40 -

40 Stock holdings. Percentile of income, 90th percentile This within-difference has remained fairly steady since 2001 hovering within a fairly narrow range (1.9 to 2.4).

Slide 41 -

41 Stock Holdings. Work status The self-employed own typically 3 to 4 times as much stocks as the employees. But, this multiple has peaked in 2001. And, again the trend in this between-difference really does not relate much to stock market trends.

Slide 42 -

42 Pre-tax income. Work status. Employee This within-difference is pretty high and has risen rapidly since 2007. But, again it does not follow the stock market trend (see 1989 – 1998; 2007 – 2010; 2016 – 2019).

Slide 43 -

43 Pre-tax income. Work status. Self-employed This is a very large and rising within-difference. Notice the rise from 4.5 in 2001 to 10.1 in 2019. This difference does not track the stock market from 1989 – 2001. But, it seems to track it reasonably from 2001 to 2019.

Slide 44 -

44 Special Section: 55-64 Years Old The 55-64 years old should have a strong balance sheet to be ready for retirement. Given that, within this section we will focus on retirement accounts, net worth, and financial assets. Our focus here is not only to evaluate inequality as we have, but even more so to evaluate retirement-readiness in terms of financial reserves.

Slide 45 -

45 Additional Definition

Slide 46 -

46 55-64 years old. Retirement accounts The current Census estimates that a 60 year old has a remaining life expectancy of 21 years. Either the Median or Mean level of retirement funds are far from being adequate to support close to a couple of decades in retirement. This denotes a nationwide failure in facilitating our elderly retiring somewhat comfortably. It is unclear what is the solution given rising fiscal pressures at every level of Government.

Slide 47 -

47 55-64 years old. Net worth The Median net worth seems grossly inadequate to support a family or household into its retirement years. The Mean net worth is more reasonable (yet probably does not make for a very comfortable and secure retirement either for a family with a remaining life expectancy of a couple of decades). The between-difference is pronounced and rising. And, it underlines how un-ready families at the Median level are to face decades of retirement.

Slide 48 -

48 55-64 years old. Financial Assets Similar comments as for the two previous slides. The Median financial asset level is grossly inadequate for families nearing retirement. The between-difference is rising spectacularly fast since 2004. But, it is mainly due to the decrease in the financial assets Median!. Meanwhile, the Mean has not materially increased since 2001 That trend is a real concern at the nationwide level.