Data from the Survey of Consumer Finances suggest that family-owned businesses are experiencing a decline. There appear to be as many households earning some income from proprietary businesses, but the proceeds from these businesses are falling. There are fewer businesses that earn a living wage, and the prevalence of high-earning proprietary businesses also seems to be falling.
Proportion of Households Earning Business Income
Consider the figure below, which shows the proportion of US households earning (1) any income, (2) at least a median income, and (3) at least a 90th percentile income from a proprietary business.
Between 8% and 10% of US households receive some business income, but between a two-thirds and three-quarters of these households fail to secure the equivalent of a median household income through personally-owned businesses. In 1992, about 3.3% of US households received at least a median household income from personally-owned businesses. By 2013, just over half as many households – 1.8% – received a median income from such businesses. The proportion of high-earning personal businesses has fallen even more sharply, from 1.3% in 1992 to 0.4% in 2013.
On one hand, these reduced earnings could be the byproduct of random variation. For example , business earnings took a bit of a dip in 1995 and 2004, but then recovered. This could be a product of sample variability or a natural minor fluctuation in small business profitability. However, that seems less likely when we look at the wider distribution of personal business earnings.
Distribution of Business Earnings
The figure below shows the distribution of business earnings from 1992 and 2013. It describes the 25th, 50th, 75th, 90th, and 95th percentile earnings from proprietary businesses.
This figure shows a long-term decline in proprietary business earnings, which seems to have begun after the 2001 recession. In 2001, the median household business earned $26,664. This figure fell regularly through 2013, where the median stood at $16,400. This represents a fall of about 38%.
A similar decline occurred among the higher ranks of the business income scale. From 2001 to 2013, 75th percentile income fell from $79,200 to $40,000. Ninetieth percentile income fell from $217,800 to $87,000. Ninety-fifth percentile income from $370,920 to $156,600. These are staggering losses.
What Does It Mean?
If these figures accurately reflect changes in personal business earnings, it suggests that their earnings are falling quickly. Why might this be happening? It may be that small business faces mounting pressures from many quarters. Retailers may have increasing difficulty competing with big box stores and online retailers. Small manufacturing outfits may have trouble competing with foreigners. It might be that many small businesses are being replaced with automated substitutes (e.g., TurboTax and LegalZoom are killing small accountants and lawyers).
Whatever the cause, it seems quite clear that small businesses (at least unincorporated ones) are doing badly in the US.
More elderly people receive social assistance, and their social assistance is far more generous.
There is a strong strain of US political culture that is vehemently opposed to social spending. These opponents generally imagine welfare recipients to be able-bodied, lazy young people who simply
don’t want to work. Conservative opponents of social programs often seize on this imagery, who try to inflame voters into supporting efforts to dismantle the welfare state.
What is interesting is that a core constituency of these anti-welfare initiatives are in fact society’s biggest welfare recipients: the elderly. We can illustrate that fact by digging into data from the Survey of Consumer Finances.1
Varieties of Welfare
Our first step involves asking, what is welfare? Here, it includes economic transfers from taxpayers to households that qualify for government programs designed to help families who do not earn enough money on their own. There are three major types of social payments programs in the US:
Workers’ Programs, payments to workers’ whose employment is disrupted, like Unemployment Insurance or Workers’ Compensation.
Social Security, which provides payments to the elderly and disabled.
Many readers might object to the inclusion of Social Security as a type of welfare program, on the grounds that it is a system into which its recipients have paid. In fact, Social Security is a PAYGO system, in which today’s seniors’ checks are principally funded by the payroll deductions taken from the working populations’ paychecks. Today’s recipients’ payroll deductions were used to pay for the Social Security checks of those who were elderly during their working years. These are current transfers from the working to (largely) non-working population.
Social Security is by far the most expensive, absorbing nearly one-quarter of the federal budget, along with a substantial part of state budgets. It roughly costs as much as all health care spending (post-Affordable Care Act) and more than double the amount of all remaining social safety net programs combined.
Number of Recipients
The figure below describes the percentage of households receiving any money from these three social payments programs:
The number of people receiving all three types of assistance fell in numbers over the 1990s. In part, this was a byproduct of an extraordinarily long prosperity and strong labor markets. When jobs are bountiful and better paid, fewer people need public assistance.
Since 2001, America’s welfare system – at least the part of it that provides money to poor people – expanded considerably. Social Security recipients grew as a group since the mid-1990s. The proportion of households receiving welfare as cash payments (e.g., TANF, SNAP, traditional welfare checks) has grown steadily since 2001. Workers’ benefits have grown since the 2008 crisis.
The figure below describes the median take from each of these three program types. This is the median receipts from each program type among all households receiving any money.
The median take from Social Security is by far the biggest and fastest-rising. In 1992, it was just over $8100. By 2013, it stood at $16,740. Payments more than doubled, and did so at a time when incomes stagnated.
The take from workers’ programs also doubled, from $2,480 to $5,120. This growth has been fueled by largely temporary expansions after the 2008 financial crisis (e.g., extending eligibility for Unemployment Insurance). Many of these program extensions have since expired, and they will probably be lower when the 2016 survey’s results are released.
Welfare, on the other hand, has not grown. In 1992, median welfare take stood at $2940. By 2013, it was $2,660.
Poor Young People Aren’t Major Social Assistance Recipients
America has a large welfare state, but it primarily serves the elderly. Social payments to the working age population and the poor are far less generous. Payments to the poor are low and have been shrinking, probably as a result of the political demonization of these programs.
In contrast, senior-targeted social programs have been generous. This generosity has been growing at a time when the government has been trying to cut other assistance programs.
These analyses were aided by the excellent R resources made available by Anthony Damico↩
Over the past several decades, America’s financial sector has boomed. Financial assets have appreciated rapidly The range financial investment options available to consumers has multiplied. One might assume that this has led to a boom in financial income for households. It has not. Today, households receive less money from financial investments than they did thirty years ago. The typical US family earns much less in financial income than twenty years ago.
The Financial Boom, and Rising Incentives to Invest
America’s financial sector is much larger than it was forty years ago. There is much more money flowing through financial markets. Financial markets now make up a greater proportion of overall economic activity. Financial enterprises now employ far more people. The rising stature of financial markets is widely known as financialization.
There is a lot of debate about whether or not financialization benefits the average American household. Some argue that expanding financial markets have created new opportunities to borrow or invest, which theoretically leads to more businesses and jobs. Moreover, it is though to have provided US households with great opportunities to accumulate wealth. From the 1980s until the 2008 crisis, financial markets were very prosperous, and those who were invested in it reaped the benefits. Those with the resources and foresight to invest in financial markets enjoyed excellent returns.
In the midst of this financial boom, many government programs sought to encourage households to invest more in financial markets. Programs like the IRA, 529s, 401k’s, or the more recent myRA give people financial incentives to invest. To those who see financial investment as a boon to households, such programs help people help themselves.
What is interesting is that, although financial markets have grown considerably, financial returns have been high, and government incentives to invest have grown, the typical family makes much less off of financial investments than 25 years ago.
Financial Income is Widely Falling
The table below compares the size and composition of financial earnings among US households in 1992 versus 2013. Data are from the Federal Reserve Board’s Survey of Consumer Finances.1 It describes the percentage of US households receiving various forms of income, and the median take of each investment type. The median take is the median value of this income among all households receiving any financial income of the sort in question. Note that these figures are in constant dollars.
Median Income (all households)
% Receiving Financial Income
Median Total Financial Take
% Receiving Interest Income
Median Interest Income Take
% Receiving Dividend Income
Median Dividend Income Take
% Receiving Capital Gains
Median Capital Gain Income Take
% Receiving Tax Exempt Bond Income
Median Tax Exempt Bond Income Take
The table suggest that fewer people receive financial income, but those who have been able to invest in stocks are taking in more money.
Fewer Households Receive Any Kind of Financial Income
The table shows that the proportion of households receiving any kind of income has fallen considerably. In 1993, 44% of US households received any kind of financial income. By 2013, about half as many households received anything. Even though far fewer households receive any kind of income, the median take hasn’t moved. In effect, the typical person who sits in the top 25% of society in terms of financial income now receives as much from their investments than someone in the top 50% person twenty years ago.
The fall in household incomes from finance is mainly the result of low interest rates. Most of the households receiving financial income did so through the interest that they received on deposit accounts or certificates of deposit. Today, those accounts yield next to nothing in the present investment environment. In other words, less than half as many people receive interest income today as twenty years ago, and those who are fortunate enough to keep getting this kind of income are receiving far less than in the past.
Similarly, the prevalence of holdings in other financial income categories has fallen, though less dramatically. For those who continue to earn income from these sources, the typical take has risen. Capital gains were buoyed by spectacular gains in the stock market by 2013. Dividends have also risen. For tax-exempt bonds (like munis), the typical take has risen, but it is probably a byproduct of fewer, bigger investors being left in those markets.
This fall is one source of consternation among wealthier older people, who were relying on their accumulated wealth to yield low-risk income through these kinds of investments.
The Median Take from Stock Market Investments
Those who have managed to maintain an investment in stock markets have seen their takes rise. The typical take on capital gains has roughly doubled in the past quarter-century. Dividends’ take has more than doubled.
What is interesting, however, is that the franchise of these investments has not expanded much. Despite the massive development in private financial markets, and multiple government programs designed to encourage financial investment, the proportion of households receiving financial income has more or less stayed the same, if it has not fallen. Moreover, it is not as if this exponential growth has created a massive stream of income for those who are invested in financial markets. Households who are invested in financial markets are generally wealthier, such that these typical takes are not huge. An elite 6% of US society is receiving any kind of dividend, but that dividend amounts to about $100 a month – far less than what Social Security delivers.
Investment Markets are Not Critical to the Vast Majority of US Households
While we may be impressed by the massive growth in financial markets, the historical returns that have accrued to shareholders, and the government’s policy of aggressively encouraging financial investment, financial income accrues to a very small proportion of US households. Moreover, most of those who receive financial income are not receiving much, relative to other common income sources like labor or government payments. The franchise of financial investment is simply not expanding. Financial income is much more critical to very wealthy households, and its role in shaping income is marginal to zero for the vast majority of Americans.
Median incomes have been stagnating for a long time, but what about those closer to the top and bottom of the income scale?
Median wages have been stagnating over the past fifteen or so years. This suggests that those who live near the middle of America’s income scale haven’t experienced serious income growth for a long time. But medians only capture what is happening to those who are at the very middle of the income scale. What is happening to those who are in the higher and lower ranks of that scale?
On one hand, we might expect those with lower incomes to experience faster income gains, because their incomes have more “room to grow.” Presumably, it is easier for someone to double their income by moving from a $15,000 to $30,000 a year job than it is for someone who earns $150,000 to find a $300,000 a year job.
On the other, most of us have some sense that those with more money tend to prosper more. Those with higher paying jobs tend to be better trained, their work is less likely to be outsourced to low-income foreign countries, and their overall income is probably supplemented by the investment opportunities aforded by their higher incomes (e.g., stock dividends, capital gains, business income)
The figure below uses Census Bureau data1 to depict changes in household incomes at different points of the country’s income scale. The lines depict changes in the lines that divide income quartiles. Those who earn less than the 20th percentile income are earning less than 80% of the population. Those whose earnings surpass the 95th percentile are in the top 5% of earners. And so on.
The graph shows how high income earners have seen their incomes grow faster over the 46 years depicted. In 1967, an income of at least $60,813 would have put you in the top 5% of earners. This translates into an income of $115,863 in 2013 dollars. By 2013, 95th percentile incomes grew by nearly 69%, to $196,000. By comparison, 80th percentile income thresholds rose by about 47% during this period, 60th by 29%, 40th by 12%, and 20th by 14%.
These differences were produced by minor differences in year-on-year income growth. Incomes at the top of the pyramid enjoyed growth rates that were less than one percentage point higher than those at the bottom, but the cumulative effect of these different growth rates are large. The table below describes how mean annual real income growth rates varied across the income scale.
Mean Annual Growth Rates of Real Household Income, US, 1967 – 2013
1967 – 2013
1967 – 2000
2000 – 2013
In addition to the observation that higher income people tend to earn more money, the data also suggest that incomes have been stagnating to falling across the income scale since 2000. Mean income growth rates were negative across the board, although the upper 20 percent’s negative rates are barely distinguishable from zero. Still, no one seems to be getting a raise in the 21st century economy, except perhaps those at the very, very top of the income scale. One possibility, which I will discuss at a later date, is that those closer to the top of the income scale are getting richer, but they are accumulating wealth through the appreciation of assets they own, as opposed to earning higher incomes.
A look at household income growth in the United States, 1947 to 2012
Over the past fifteen or so years, median real household incomes have stagnated. This observation is widely evoked in arguments that America’s middle class is faring poorly. It is used as a basis for asserting that the US economy does not serve regular Americans’ economic interests well.
Real income means inflation-adjusted income, a metric that is used in an attempt to differentiate genuine raises from those that merely reflect changes in general prices. For example, the median household in 1947 received about $3 thousand in income, but that money could purchase more then because prices were generally lower. At current prices, the Census Bureau estimates that this $3,000 (the prevailing nominal income at that time) is equivalent to (a real income of) about $27,000 today.1 This implies that the median household, whose income was about $60 thousand in 2012, has somewhere between twice and three times (rather than twenty or so times) the purchasing power of its post-WWII counterpart.
The figure below depicts canges in median US real household income from 1949 to 2012. Data are drawn from the Census Bureau.2
Between 1950 and about 1970, median household incomes rose quicky and steadily. Median real incomes grew at an average annual rate of about 3.5% per year, compared to an average rate of 0.9% from 1970 to 2000 and -0.1% from 2001 onwards.
These seemingly modest differences in growth rates amount to big income differences over time. For example, had real income continued to growth at its mid-century pace, real median income in 2012 would have been somewhere around $236 thousand, instead of $64 thousand. Had it maintained its 0.9% annual growth rate from 2000 onward, median incomes would have been somewhere around $75,320 by 2012. Had incomes not fallen since 2000, the median household would have about $5000 income in inflation-adjusted terms.
Over time, median wages have slowed. Their decline over the past 15 years makes the slow, steady growth of the 1970s through 1990s look like the “good old days.” However, the middle class prosperity of the past 40 years was far inferior to what was experienced in the 1950s and 1960s, at least insofar as real incomes are concerned.
Over the past several decades, the US economy has largely failed to secure rising incomes for its middle class. Policy has failed to create an environment in which regular Americans have been able to earn more money relative to prices. Of course, discerning how it has failed is a much more complicated (and different) topic.
How much money does it take to be part of the upper class? The lower class? These types of questions are troublesome. There are undeniably wealthy people who think that they are middle class, and some patently poor people who self-identify as part of the middle class. Drawing economic class lines is complicated.
Still, it is interesting to draw hypothetical lines. The exercise gives us some indication of what different classes’ personal finances might look like, and where someone seems likely to lie on the country’s economic hierarchy.
According to a 2012 survey by the Pew Research Center,1 about 32% of US society self-identifies with the lower or lower-middle class. Another 49% identify with the middle class proper. A further 15% identify with the upper-middle class, and 2% with the upper-class. Without data on these respondents’ personal finances, we cannot create a profile of how households’ finances differ across (self-identified) economic classes.
Drawing Class Lines with Data
Perhaps another solution is to ask how household finances differ across classes if people were to correctly identify their own economic class. For example, if the 2% of respondents were correct that they were at or above the 98th percentile of income or wealth, then how much would they earn or possess? We can answer this question using data from the Survey of Consumer Finances2. The table gives the implied income and net worth ranges:
Net Worth Range
(min to 7th pctl.)
(7th to 32nd pctl.)
$11,566 – $30,435
-$9,019 – $18,210
(32nd to 82nd pctl.)
$30,436 – $109,914
$18,211 – $483,420
(82nd to 98th pctl.)
$109,915 – $417,175
$483,421 – $4.4 million
(above 98th pctl.)
above $4.4 million
I would imagine that the income figures make sense to readers, but not the wealth figures. In general, people seem to draw economic class lines based on income, and they often fail to appreciate how wealth varies and the impact of wealth on a household’s overall economic situation. Conretely speaking, the middle class has wide ranges in wealth, which sit between the rough equivalent of a year’s worth of poverty line income and nearly a half million dollars.
An alternative, ad hoc division might look like this. The typology captures how a very large proportion of society has next to no wealth. While 7% of society’s households are “poor” in income terms, nearly a quarter of them are “wealth” poor. The middle 50% has some wealth, but not enough to cover a few years at the poverty line. Like income, wealth is concentrated in the top quarter, and most of this is concentrated in the higher ranks of this top quartile: