Donald Trump and the Media Agree on Middle-Class Income, and They are Both Wrong

Agresti, J. D. (2016, July 27). Donald Trump and the Media Agree on Middle-Class Income, and They are Both Wrong. Retrieved from
Agresti, James D. “Donald Trump and the Media Agree on Middle-Class Income, and They are Both Wrong.” Just Facts. 27 July 2016. Web. 16 July 2024.<>.
Chicago (for footnotes)
James D. Agresti, “Donald Trump and the Media Agree on Middle-Class Income, and They are Both Wrong.” Just Facts. July 27, 2016.
Chicago (for bibliographies)
Agresti, James D. “Donald Trump and the Media Agree on Middle-Class Income, and They are Both Wrong.” Just Facts. July 27, 2016.

By James D. Agresti
July 27, 2016

During his acceptance speech at the Republican National Convention, Donald Trump stated that “household incomes are down more than $4,000 since the year 2000.” Many “fact checkers” of Trump’s speech declared this statement to be true, including those of the New York Times, PolitiFact, Vox, and NPR.

In reality, the data cited by Trump and these media outlets excludes 24% of household income. More comprehensive data recently published by the Congressional Budget Office (CBO) reveals that median household income has climbed by $5,000 since the year 2000.

CBO’s Data

In June of 2016, CBO released a report that details inflation-adjusted household incomes over the past several decades. This data shows that median household income rose by $5,000 between 2000 and 2013 (the latest available data). This occurred in spite of the fact that household income has stagnated since the Great Recession of 2007–2009. These trends are even more vivid when subtracting out federal taxes in order to estimate take-home income:

median_household_income_1979-2013Unlike the source cited by Trump and the media, CBO accounts for nearly all forms of income in its analysis, including wages, capital gains, Social Security, and unemployment insurance, as well as non-cash benefits like food stamps, Medicare, and employer-provided health insurance.

Also unlike the source cited by Trump, CBO uses multiple sources of data to achieve accuracy, including:

  • a large sample of individual income tax returns, which is its “core data” source.
  • survey data collected by the U.S. Census Bureau.

Nevertheless, such economic data is almost always imperfect, and CBO notes that its analysis:

  • does not capture all government spending on social programs, because people tend to underreport this income and because some of this money goes to individuals who are “outside the scope of the survey data.”
  • tracks a new set of households each year, and hence, it cannot tell us what happens to actual households and actual people, who may move to higher or lower income groups from year to year.
  • does not account for 7% of federal taxes, which CBO was unable to accurately allocate.

To account for inflation, CBO uses the Personal Consumption Expenditure price index. This index tends to show slightly less inflation than the commonly used Consumer Price Index. CBO considers the Personal Consumption Expenditure price index to be more accurate, in part, because “it more fully accounts for the adjustments that consumers make to their spending patterns as some prices change relative to other prices.”

After translating CBO’s data to the Consumer Price Index, it is still at odds with Trump’s claim that income has fallen by $4,000 since 2000. In fact, it shows median household income rising by $1,500 and approximate take-home income rising by $5,000 since 2000.

Trump’s Source

The income figures cited by Trump and the media come from the U.S. Census Bureau’s Current Population Survey. This data indeed shows that inflation-adjusted median household income declined by $4,000 between 2000 and 2014 (the latest available year).

However, the data is incomplete and misleading if used uncritically. In the words of the Census Bureau, this is because it:

  • is “based solely on money income” and does “not include the value of noncash benefits,” such as “food stamps, health benefits, subsidized housing,” and “full or partial payments by business for retirement programs.”
  • excludes “certain money receipts such as capital gains.”
  • uses the Consumer Price Index to adjust for inflation.

After listing these caveats and others, the Census Bureau warns that “data users should consider these elements when comparing income levels. Moreover, readers should be aware that for many different reasons there is a tendency in household surveys for respondents to underreport their income.”

Likewise, CBO uses this Census data as a supplementary source for its income analysis but explains that it is not a comprehensive measure of income. CBO notes that while the tax return data it uses as its primary source and the Census data each have limitations, together, “the two sources can form a more complete picture.”

This larger picture becomes especially important when comparing income over extended periods, because the Census data’s shortcomings accumulate over time. For example, the sources of income excluded by the Census survey have become increasingly significant over the past decades, growing from 15% of middle-class household income in 1979 to 24% in 2013. This causes the Census data to show less income growth than actually occurred. Hence, it should be used thoughtfully and with its caveats explained, especially when it conflicts with more complete sources of data—as it does in this case.

False Narratives

Like many political narratives that diverge from the truth, this particular one serves the interests of those who spread it.

In the case of Trump, it reinforces his assertion that Hillary Clinton “has supported virtually every trade agreement that has been destroying our middle class.” Along the same lines, Trump says that “powerful special interests” have “rigged our political and economic system for their exclusive benefit,” and he contends he will stop this if voters give him the presidency.

In the case of media outlets like the New York Times and PolitiFact—both of which have records of propagating left-leaning falsehoods—Trump’s claim furthers the notion that the middle class is suffering, and the government should expand social programs to help. It also allows them to say, as PolitiFact did, that Trump’s income “statistic came amid a series of shots at President Barack Obama,” but the “majority of the period Trump was referring to came under the presidency of George W. Bush, a Republican.”

Contrary to these false narratives, the most comprehensive available data on median household income shows that it generally rose for the quarter century preceding the Great Recession of 2007–2009. Since then, it has been nearly flat for as long as this data currently extends.

It is important to note that these facts don’t prove that the economic policies of Bush and previous presidents were better for middle-income households than those of Obama. This is because it is impossible to objectively separate the impacts of a president’s policies from numerous other factors that determine the incomes of Americans.

At the same time, several factors that negatively affect median income have reached record or near-record highs in recent years, and Obama has implemented policies that have contributed to these outcomes. This includes but is not limited to increasing the national debt, propagating regulations, and fostering immigration of low-skilled populations into the United States.

  • July 27, 2016 at 2:30 PM

    It can be argued that Social Security, and unemployment insurance are earned income but still do not reflect in a real current way what the a working age middle class worker is facing in terms of flat or declining wages.
    The non-cash benefits like food stamps, Medicare, and employer-provided health insurance may be a help to families but no working age middleclass worker would consider it income in the sense of spendable or saveable take home salary or wages.

    Trump is right on this subject and no Government mumbo jumbo, gabblygoop, statistical manipulation by including additional items will change that real fact that in today’s market the american worker see a flat and declining wage.

    • July 27, 2016 at 2:49 PM

      Trump and the media said “income,” not “wages,” and that is precisely the data they cited.

      Your comment provides no evidence that wages have declined since 2000.

      It does not matter whether a middle-class worker would “consider” health insurance and other non-cash benefits to be income. The fact of the matter is that they are.

      • September 12, 2017 at 3:19 PM

        Did you factor in devaluation of the dollar’s buying power over the same time period?

  • July 27, 2016 at 9:55 PM

    Appreciate your honesty. No one can say you’re coddling Trump!

  • July 28, 2016 at 11:01 AM

    “food stamps, health benefits, subsidized housing,” I cannot view these as income, because I define income as money exchanged for something of value. The better question is to look at welfare programs and see if their contribution to household income has grown over time and by how much. Also, if you back welfare out, what does income look like then?

    • July 28, 2016 at 11:48 AM

      Whether your employer gives you a health plan as part of your compensation or gives you money that you use to buy a health plan, it is still income. The same applies to government benefits like food stamps.

      I have the data to back out all government benefits and plan to do so in a future article. It is important to note that government benefits have perverse effects on income, because they give people the means and motive not to work:

      • November 30, 2016 at 9:17 AM

        I do believe we have catered to certain groups so much, that it has made them dependant on welfare, however since it is a long standing policy it would be rather unfair to cut them off. Our goverment has cause this situation, and it will be difficult to change it. First of all , you would have to take big bussiness out of it. They are responsible for sending manufacturing jobs overseas for cheap wages and cheaper made products. CONGRESS IS THE ENABLER here.
        The deals they make puts money in thier pockets.
        This is not greedy citizens , it is the responsibility and fault of Congress.
        Do we really realize the damage politicions do to our economy? Goverment statistics are not as reliable as some of you think, I don’t believe these statistics are that close to the truth, prehaps Trump’s are a little off, but closer then the rose closed picture the Goverment claims.

        • November 30, 2016 at 9:29 PM

          Citizens elect Congress; so to absolve citizens and blame Congress is nonsense. People complain about politicians, but they are simply a reflection of the will of the people.

          You present no evidence for your claims about the government’s and Trump’s numbers, and hence, your claim can be dismissed without evidence.

    • March 7, 2017 at 10:04 AM

      Your analysis is the basis upon which I also would consider as to whether incomes have risen or declined. Govt handouts aren’t income, they redistribution of “real” earned income. Trump’s assessment captures the crux of this issue. Some have no problem riding on the coat tails of others. This entitlement attitude has encouraged dependency and damaged personal accountability and entrepreneurial spirit.

  • July 28, 2016 at 1:10 PM

    Health insurance premiums have nearly tripled since 2000. Since employer contributions for the insurance are income, how much of the rise in median “income” can almost be entirely contributed to that?

    The average health insurance premium as percentage of income has risen at an astonishing rate (14% to 21% from 2003-2013; 2000-2016 is likely MUCH larger). Deductibles also doubled and that’s probably also a larger number now, as my “low” deductible plan’s deductible has doubled within two years. Also, looking at some older data, there was a 4% rise in employer’s share of health insurance contribution from 2014 to 2015. The trend is likely accelerating, but time will tell.

    So essentially, wages are stagnating. Other benefits are staying pretty much the same, in terms of total % of my compensation. But the insurance premiums are going through the roof. Within 20 years, over 1/3 and probably closer to 1/2 of my total compensation will be health insurance. So it looks like I’m receiving more income, but I’m literally receiving the same thing. You do mention non-cash benefits are growing, but you don’t really examine it more; nor do you realize that that itself could pose a large issue. At the rate we’re going, My grandchildren are going to have an inverse compensation scheme: 15-25% wages and 75-85% employer health premiums. But economists are going to argue their income is skyrocketing.

    You can continue to beat your fists on the table and shout “IT’S INCOME” because of the obtuse CBO definition of income. But it’s not income in the commonly understood sense, and Trump’s claim has more in common with what the average person sees as income. Not what you do.

    • July 28, 2016 at 1:52 PM

      Inflation accounts for growth in healthcare costs, and CBO’s data accounts for inflation. Hence, your argument about the growth of healthcare costs is meritless in this context.

      Moreover, as documented in the article, the Census data Trump relied upon has other flaws.

      CBO’s definition of income is a practical, commonsense measure that captures practically every form of revenue that households receive. The definition you propose is narrow and flawed.

  • July 28, 2016 at 3:21 PM

    It is helpful to parse the stats and show discrepancies in rhetoric and reporting. What about the fact of people moving in and out of these income groups? Over 16 (or 13 available data) years is pretty long. I like Dan Mitchell’s take on the country’s economic condition and how he compares it to other nations. He agrees with the $4000 decrease, but IMO, it doesn’t matter that much. To some an increase of $4000 over 16 years is significant and may prove the middle class is healthy, but I must disagree. You rightly use incomes such as gov’t help and investment returns, but middle class expenses are very high and we get very little help if in upper mid class. Especially if you live in a high tax state and city. You cannot convince me using this macro data that a lower middle class family is going to be happy even if they obtained $4000 IN CASH over 16 years. Bush and Obama were/are statists. Their overall economic records reflect policies and conditions that have limited the potential for economic growth. Mitchell proves this case and I agree. He shows how the Clinton years were good for growth precisely b/c overall gov’t spending was lowered during his years as president. Please give us a President and politicians who would stop interfering (not that Trump would be) and follow the SIMPLE advice that Mitchell has shown enabled many countries (AND THE USA AT ONE TIME) to grow best. Yes the rich will probably get richer, but it is a necessary condition. Redistribution is not the answer, GROWTH is. Repubs used to say these things, but did not deliver. (IMO a reason why Trump is the R-nominee) The leaders and candidates are beholden to special interests & class warfare inhibiting them from lowering spending, especially Dems, which is a simple yet significant way to get more growth. He shows how by design the gov’t manipulates budgets to prevent REAL fiscal reform. If they cannot be honest with the data AND the population we will be a fiscal mess and ALL Americans will suffer.

  • July 30, 2016 at 5:19 PM

    I tend to agree with the other commentators. Wage income is declining. When you add in government transfer payments the income decline should be less and as CBO attempts to point out there may even be in increase. For example, Healthcare benefits consume more of individual’s possible wages each year (as healthcare goes from 18% to 20% of GNP). However, income (and productivity growth) is key. The other non-wage benefits are mainly income transfers. They should be a temporary help when times are hard. They are increasingly a permanent feature of life and are becoming unaffordable because the people being taxed have less income to be taxed in each cycle.

    • July 31, 2016 at 12:07 AM

      Wage income may or may not be declining, but that does not mean household income is declining. In fact, median-income households are measurably better off than there were in 2000, despite factors that are placing downward pressure on wages, such as:

      • masses of baby boomers retiring.
      • increasing levels of broken families, which spreads workers’ wages over an increasing number of households:

      I agree with your conclusion that income transfers “are increasingly a permanent feature of life and are becoming unaffordable,” but middle-income federal taxes have declined since 2000. Instead, these burdens are being thrown onto workers and future generations via the federal debt:

      • October 12, 2016 at 3:06 PM

        How does this stack up with the latest Census Bureau survey which put median household income at $33,544, and showed household net worth to be down 91% in the last 12 years (and negative for Black households)?

  • July 30, 2016 at 5:50 PM

    1. In the real world, as opposed to the tax code, my income is what I can spend as I choose.

    2. That more people are living better on the taxes paid by their fellow citizens is not good news.

    3. If you’re suggesting that the CPI exaggerates inflation, you’re suggesting the opposite of the truth.

    • July 30, 2016 at 11:54 PM

      1. In the real world, you cannot make up your own definition of income.

      2. Agreed, but as I’ve already pointed out, Trump’s figure includes unemployment, cash welfare, and other government entitlements. The dividing line between Trump’s statistic and CBO’s data is not government subsidies. As explained in the article, it includes a host of items that the Census data does not capture.

      3. Your statement is an unsubstantiated assertion that is at odds with a concrete fact documented in the article.


Make a Comment

Your email address will not be published. Required fields are marked *

Articles by Topic