Are Wages, Salaries and Incomes Keeping Up with Inflation?

The fairest answer is resoundingly "yes", but below we'll try and assess this all-important question in as objective a way as we can.

You've probably read and heard it a thousand times, maybe even in 2024 alone: wages and incomes aren't keeping up with inflation, and too many Americans are being left behind. Few people actually seem to have looked at the data on this, however, or they're just giving you a selective portion of it, simply because almost all the data we have indicate that this just isn't true.

Below I pull the following data and compare it to the Consumer Price Index (otherwise known as "CPI"), all of which can be found at the Federal Reserve Economic Data (FRED) website. In this post, I will compare all of this wage, salary and income data to the "All-Items" CPI, which includes food and energy.

• Median Personal Incomes. This looks at the median gross income for individuals in the US. This has data back to 1974.

• Income Before Taxes, looking only at Wages & Salaries. I do this by income quintile (bottom 20%, 21-40%, 41-60%, all the way up to the top earners). This has data back to 1984.

• Income Before Taxes, All Income, again by income quintile (and also back to '84)

• Income After Taxes, All Income, again by income quintile (and again back to '84). This helps account for changes in tax policy that helps (or hurts) our take-home pay.

• Average Hourly Earnings for Private Workers, All Kinds of Employees (this is also provided in the BLS Jobs Report each month). This data only goes back to 2006.

• Average Hourly Earnings for Production and Nonsupervisory Workers (also from the BLS Jobs Report each month). This goes back to 1964.

• Employment Cost Index: Wages and Salaries: Private Industry Workers (this is a key index that the Fed focuses on and is a good barometer of labor costs in the economy). This data goes back to 2001.

• Employment Cost Index: Total compensation: All Civilian. This data also goes back to 2001.

A couple things before we continue.

First, you might notice that I didn't use household income figures in my analysis. The reason I don't do this is because for a variety of reasons (lower marriage rates, fewer kids, etc.), household sizes in the US are shrinking. though they seem to have stabilized since dropping below 2.0 persons per household from 2001-05 (down from 2.4 in 1950, and 2.07 in 1980). With fewer earners in a household, it's naturally going to distort household income growth. Better to alleviate this by looking just at individual incomes.

Additionally, those of you who read my posts on the monthly CPI releases each month might be asking why we're using "headline" inflation here rather than "core", which excludes food and energy prices (as we usually do in the monthly posts). The answer here is because this is very much a political question, so I use the "All Items" CPI because it's what everyone actually pays. In our monthly inflation posts, I tend to focus more on core inflation because 1) the Fed focuses more on it and 2) it's actually a better predictor of where headline inflation will be in the future (which is why the Fed focuses on it, in addition to the fact that it's less volatile). The realities or future of inflation are not the subject of this post though, so we'll use "headline" or "all items" CPI instead, since again, that's what everyone actually pays.

Rather than arbitrarily selecting the years or time periods (something I think people do often when evaluating this question), I looked at all the data above in three ways:

• First, I look at wage or income growth (as the case may be in each data set) each year relative to what inflation did. Then I look at what percentage of the years the relevant wage and income benchmark exceeded inflation. So if we were looking at the period of 1984-2023, and the relevant compensation data exceeded CPI increases 60% of the time, the number in the data table or chart would be 60%.

• Second, I do the same thing, except on rolling three year periods, to try and account for the fact that sometimes inflation is more volatile than wages and incomes.

• Third, I do the same thing again, except on rolling five year periods.

The results are actually quite similar across all of these data sets. The answer seems to be that wages, salaries and incomes, across all income groups, have kept up with inflation about 60-70% of the time. This has been more than enough to meaningfully improve standards of living over time, for all segments of the population.

Let's start with median incomes, which since 1974 have exceeded inflation 63% of the time on a one-year basis, 60% of the time on a rolling three-year basis, and 69% of the time on a rolling five-year basis.

It probably won't surprise anyone then to see the longer term chart of real median income over the last 50 years generally going up and to the right, though there are periods where it flattens out for a time.

The data shows the same thing when you look at wages, salaries and incomes by quintiles: they've exceeded inflation about 60-70% of the time over the last 40 years. Higher quintile workers tend to skew towards the upper end of this range, while workers in the bottom two quintiles (the 0-20% of earners and 21-40% of earners) skew more towards the lower end. As the below three charts show, however, this is definitely not always the case. For example, looked at in one-year increments, income growth for the country's highest quintile of earners actually doesn't keep up with inflation as often as the bottom quintile's does, either pre or post-tax. This flips, however, when you look at multi-year increments.

This chart shows that for the bottom quintile of earners, wages and salaries exceeded CPI increases in about 55-60% of years. Each quintile after that shows the same thing, all in one year increments. So the first bar in the 0-20% section indicates that wages and salaries exceeded CPI in 56% of the years from 1984 to 2023, the second shows the same data for pre-tax incomes for the bottom quintile, and the third bar shows post-tax incomes for the bottom quintile. Each section of the X-axis after that corresponds to the same data for that income quintile.

Now this chart does the same thing, except on a rolling three year basis. The corresponding bars in this chart are almost always higher than in the first chart, for the reason we mentioned above: sometimes inflation spikes and it takes wages and incomes a little longer to catch up. But they usually do.

This last chart shows the same thing, except on rolling five year intervals. Interestingly, on rolling five year intervals, wages and incomes have tended to exceed inflation less often than either in annual increments or on rolling three year intervals. Notably though, this was not the case with the lowest earners, but even then, the five year intervals for wages and incomes again still exceeded inflation about 60-70% of the time. That's pretty darn good.

Now here's the table with the data from the charts above, for reference.

Now let's look at the data from the monthly BLS Jobs Reports on Average Hourly Earnings. The story generally remains the same, with some exceptions:

Average hourly earnings from the BLS monthly jobs reports also indicates that wages are keeping up with CPI about 60-70% of the time, whether that be for one year increments or for three or five year rolling periods. Note that for average hourly earnings for all private employees, since that data only began in 2006, the first five year period began in 2011, so we only have about 15 years worth of 5 year intervals in that case. That makes the 96% figure above less significant.

Lastly, here's the same chart using the Employment Cost Index data sets we mentioned above. Here again, the conclusion is similar: wages, salaries and incomes have exceeded inflation on an annual, rolling three, and rolling five year basis about 60-70% of the time.

Below is that data table with the figures from the above two charts for reference as well:

To reiterate, wages, salaries and incomes are keeping up with inflation about 60-70% of the time, almost regardless of what time period you look at, and almost regardless of what income or wage data you choose to look at. We deliberately chose this methodology to remove the ability to selectively pick years or time periods to try and show that wages and incomes either are or are not keeping up with inflation. By looking at things on an annual, three-year rolling, and five-year rolling basis, it is far more intellectually honest, and eliminates the selective disclosure that we so often see from the press (and politicians) on this issue to tell you that you're being left behind.

Wages, salaries and incomes exceeding inflation 60-70% of the time (and over long periods of time) indicates we are consistently improving our standards of living. We all like to talk about how we're getting left behind, but when you ask people if they'd actually want to go back and trade living life today for life in a previous decade, no one ever says yes. The answer to that thought experiment alone should probably be indicative enough that we don't really believe things are worse today than they were in the past. Yet we don't act on this, especially not during political season.

So why has this fib been perpetuated upon us for so long? It's true that it's much easier to sell newspapers and unseat the incumbent when you complain about how bad things are. It might even be a (sad) fact of human nature that we constantly adjust to our current conditions, and therefore maybe unintentionally depreciate how good we have it, always wanting more because we've acclimated to our new situation. This is not dissimilar to when you give a child a toy that they've been begging for, only to see them finally get it, play with it for five minutes, and then promptly discard it and ask for something new that they don't have again. Children, after all, eventually become adults.

Whether it's human nature then or whether it's because this negative attitude has a certain allure to it, the press and those running for office continue to peddle this falsehood and exploit it. Years and years of hammering this into our brains has seemingly made this notion so widely believed that few incumbents ever dare utter the words you are actually better off than you were four, eight, ten, twenty or however many years ago you want to pick, even though that's exactly what reality and the data says. The question remains then: will we ever acknowledge it?

1/8/25 Update:

One interesting supplement to this exercise is to look back at the last time Americans' confidence in the economy was particularly high, and then look at how wages, salaries and incomes were growing relative to inflation in those periods, and compare it to other periods.

So the first question is, when did Americans last feel particularly good about the economy? The answer seems to be most noticeable on two occasions: first, the 90s, and particularly the late 90s. Second, the mid to late 2010's.

Below are two charts, both from Gallup. The first is a stacked area chart showing the percentage of Americans responding to the question "How would you rate economic conditions in this country today -- as excellent, good, only fair, or poor?" Each part of the chart represents the portion who answered accordingly: the bottom blue portion would be "excellent," the orange area above it would be the portion who responded "good," and so on and so forth. The second chart attempts to clean up the noise from the first chart by giving each answer a weight, so as to create a "Current Conditions Index." I assigned a score of 4 to the responses saying "Excellent", 3 to "Good", 2 to "Only Fair" and 1 to "Poor." I then summed these to produce one number.

Though it's noisy, the charts clearly show the periods where "current economic conditions" are excellent or good were in the late 90's and the late 2010's (just before COVID). The next chart will show this more clearly.

This chart takes the above data and simplifies it by creating an Index. Answers for "Excellent" are given a score of 4, "Good" are given 3, "Only Fair" 2, and "Poor," 1. The percentages are then summed to produce an aggregate figure.

So now the question should be, were wages growing significantly faster than inflation during those periods relative to other periods, particularly compared to the last 2 years? Or asked differently, was "real" income growing significantly faster during those periods than it is today? The answer appears to be generally, yes, but the data here is definitely mixed. Regardless, the differences certainly do not appear to be so significant that they justify people claiming the current economic climate is about as bad today as it was during the Great Financial Crisis years of 2008-2009, and actually worse than the recessionary years of 2001 to 2003 (our composite index from above bottomed out at 185 in February of 2003, exactly where the last reading indicated in Dec. 12-18 in 2024).

The table above compares various measures of real income growth (incomes adjusted for inflation) by period, to see if the data during the periods where Americans claimed to feel much more confident about the "current conditions" in the economy than they do now was supportive of such enthusiasm. The data does seem to indicate that prior periods of higher enthusiasm was generally supported by stronger real income growth. But that being said, the degree of un-enthusiasm for today's economy feels out of whack with the data in the '23-24 period compared to the '96-01 or '15-19 periods. Even if it is not quite as good as those other periods, the sentiment suggests we're not just in a recession, but a bad one, which obviously isn't true.

Bottom line, in the two earlier periods where Americans felt the best about the economy, real income growth was stronger than it's been of late, generally speaking. As the table above shows, however, this is not an across the board thing. In some cases the last two years of the "Biden Economy" seem to have actually been better for real income growth than either the Clinton economy of the late 90's or the Obama / Trump economy of the mid to late 2010's.

It's possible it just takes time for enthusiasm to catch up with reality. This seems to be what happened in both the 90's and in the mid 2010's. In the 90's, enthusiasm about the economy didn't really take off until 1995 and 1996, despite median incomes growing faster than inflation in both 1993 and 1994 (albeit not by quite as much). Median incomes would end up growing faster than inflation for eight consecutive years in the 90's, from 1993 until 2001.

The same general thing seemed to happen in the mid 2010's. Median income growth was noticeably above inflation in both 2013 and 2014, but it wasn't until 2015 where confidence in the economy really started to pick up. In that case, the American economy had a streak of seven consecutive years where median incomes grew faster than inflation, and 2020 had a great chance to tie what is probably the record from the 1990's, if it hadn't been for COVID.

In each of the cases of the late 90's and mid to late 2010's, the lag may very well be explained by the fact that the unemployment rate was elevated until 1995 and 2015, when in those respective years, it ended the year at 5.6% and 5.0% after peaking at 7.8% in June of '92 and 10% in October 2009, again respectively. High unemployment of course would not explain the poor sentiment around today's economy, however, since the unemployment rate stands at 4.2%, and was well below that for most of the last four years. This is why this analysis has focused so much more on real income growth. In theory, people will care less about real income growth if they don't have jobs where they can benefit from that real income growth. Once they're employed, however, their primary focus should be on real income growth, since that is what improves standards of living. This was not the case during the Biden years, however, when unemployment spent most of the time below 4% and when in each of the last two years real incomes grew very solidly. Something else was going on.

It's for this reason that surveyors ought to include questions about political affiliations (or at least, political leanings) when doing these surveys. If the economy has simply become another issue which we view through our private political prism, then that will change a lot. Though this isn't apples-to-apples with our Gallup Poll question (which, for what it's worth, did not show a meaningful uptick in the "Current Conditions" question after the election), Morning Consult put out a chart showing wild swings in sentiment about the economy after both the 2020 and the 2024 elections (the chart below can be seen in this New York Times article), with Democrats spiking and Republicans falling in 2020, and just the opposite happening after the election this time around. While it's not surprising we might feel better about the economy if our preferred political choice regains the White House, the spikes in the data seem excessive, which is why it will be important for Gallup and others to ask these "Current Conditions" questions in ways that will allow us to see how people of different political persuasions answer, thereby giving us a better idea of how politics may be knowingly or unknowingly tainting our view of the economy.