DIFFERENT MEASURES OF INCOME

This section is divided into two parts: first, a definition of concepts, and second, a review of income charts commonly used in this debate.

Definition of Concepts

Unless you're familiar with the basics of income comparison, it can be dead simple to be fooled by supply-side statistics that "prove" everyone's incomes rose during the 80s. Don't be taken. Here is a primer on navigating your way safely through the controversy:

Average income: Also frequently called mean income. Although everyone knows what this is, it can be deceptively used. To arrive at average income, you add everyone's income together and divide by the number of incomes. The problem with average income is that it does not tell you how equally it's distributed. On a deserted island of 10 people, one lucky soul may make the entire island's income of $30,000, and the other nine survivors nothing. Does this mean that the other nine destitutes make an average of $3,000? Of course not. Which leads us to:

Median income: Median income is somewhat more indicative of income distribution. If you have three incomes lined up from smallest to largest -- say, $10, $100, and $1,000 -- then median income is the one in the middle. Notice that this is a crude measure of income inequality as well, since it reveals nothing about income disparities at either end of the scale. However, comparing average and median income is often useful, because the greater the difference, the more that income is concentrated at the top end.

Quintile Income: Economists frequently use this because it depicts income inequality much more accurately. A quintile is one-fifth of the population. The first quintile is the poorest fifth, the second quintile the second poorest fifth, and so on. Sometimes economists also use "deciles," which break populations down into sections of 10 percent.

Current dollars: Also called nominal dollars. These are the prices you pay and incomes you make in any given year. Some old-timers remember when candy bars sold for 5 cents back in 1945; but in 1996, your local theater sells the same bar for $2.25. These prices are expressed in current dollars, and rise every year due to inflation. Some unscrupulous souls (like Rush Limbaugh) sometimes use current dollars to claim that total tax collections rose spectacularly in the 80s. But what they should really use is:

Constant dollars: Also called real dollars. These are dollars that take inflation into account. In the above example, a 1945 candy bar costs $2.25 in "1996 dollars," and a 1996 candy bar costs 5 cents in "1945 dollars." Economists build large tables, called Price Indexes, to convert these prices into constant dollars. Economists almost always use constant dollars to compare prices across years, and if you run into anyone who doesn't (like Rush Limbaugh), your BS antenna should go up.

Individual income: this is the safest measure of comparison, one that sidesteps many statistical tricks often associated with household and family income. As the previous chart for nonsupervisory workers shows, the vast majority of individuals in the 80s experienced dropping incomes. And this has nothing to do with falling productivity. Individual worker productivity grew about 1 percent a year in the 80s. The only individuals who experienced growing incomes were those in the top quintile -- and those in the top 1 percent saw sky-rocketing incomes.

Households: The U.S. Census defines a household as any number of individuals in a single housing unit or private living quarter. This includes both families and non-families. Between 1980 and 1990, the number of households grew from 80 million to 93 million. In that same time, the size of the average household fell from 2.76 to 2.63 persons, and this was part of a much longer trend.

Families: The U.S. Census defines a family as a group of two or more persons related by birth, marriage or adoption and residing together in the same household. Between 1980 and 1990, the number of families grew from 60 million to 66 million. Family size during that time fell from 3.29 to 3.17 persons. This, too, was part of a much longer trend. (Interestingly, the Census does not consider an unmarried couple with a child to be a family. It has a separate entry for unmarried couples. Between 1980 and 1990, the total number of unmarried couples grew from 1.6 million to 2.9 million, and the sub-category of those with children grew from 431,000 to 891,000. This is usually too small to noticeably affect income statistics.)

The Business Cycle: At least since the dawn of capitalism, our economy has followed an irregular cycle of recessions and recoveries. This does indeed affect incomes; incomes generally fall during recessions and climb during recoveries. In the long run, however, our economy grows as our population expands and our productive efficiency improves. Which means that the deeper the recession, the steeper the recovery. (History bears this out.) One of the fun things about debating the 80s is that supply-siders attempt to blame the recession of 1982 on Carter, so they can credit Reagan for the expansion. Liberals believe that the business cycle is natural and inevitable, and is influenced far more by the Chairman of the Federal Reserve Board than the President. We'll dig into this in more detail later, but for now be aware that it is a central issue.

Now let's apply these concepts to some frequently-used charts in this debate:

Other Income Charts

The following is one of the simplest as well as most frequent supply-side charts:
Median Household Income (1994 Dollars)

1978  $33,074
1979   32,966
1980   31,891 << recession year 
1981   31,374
1982   31,269 << recession year
1983   31,274 
1984   31,972
1985   32,530
1986   33,665
1987   33,999 
1988   34,106 
1989   34,547
1990   33,952 << recession year 
1991   32,780 << recession year 
1992   32,361
1993   32,041
1994   32,264

As you can see, incomes dipped in recession years and started climbing in-between. (The last recovery has been uncharacteristically flat, but that's because income inequality is becoming extraordinarily great. But more on that later.) Notice that the 1978 median income was not recovered until 1986. Because economies grow in the long run, income continued to rise after 1986 to $34,547, before sinking into the next recession.

But ultimately this chart is not useful in learning anything about improving incomes in the 80s. There are three reasons. First, this chart does not show how much overtime people were working to compensate for falling wages. Second, the workforce as a percentage of the population has been growing for decades now. In 1980, 63.8 percent of the population was working. In 1990, this rose to 66.4 percent. What this means is that there were more paychecks in all those households, which would drive up household income, even if paychecks themselves were falling. Third, median income doesn't show the broadening gap between the very richest and the very poorest. But there is a way to get an idea of what this gap is.

Compare the chart above on median income with the chart below on average income. Recall that the greater the gap between the two, the more income is concentrated at the very top. As you can see, the gap has been growing:

Average Income (1994 dollars)

1978  $38,927
1979   39,161
1980   37,929
1981   37,481
1982   37,684
1983   37,795
1984   39,174
1985   40,033
1986   41,592
1987   42,281
1988   42,615
1989   43,647
1990   42,411
1991   41,263
1992   41,027
1993   42,489
1994   43,133

Therefore, yet another criticism of the first chart on median income is that it does not show how the poor were actually losing ground.

Below is another chart often used in the debate. It is the median individual income for full-time, year-round workers in 1994 Dollars:

Year    Men     Women
1972  $35,174  $20,204
1973   36,008   20,371
1974   34,728   20,485 << recession year
1975   34,107   20,355 << recession year
1976   34,577   20,738
1977   35,338   20,668
1978   35,265   21,167
1979   35,005   21,090
1980   34,525   20,872 << recession year
1981   34,035   20,490
1982   33,570   21,181 << recession year
1983   33,491   21,544
1984   34,239   21,998
1985   34,432   22,384
1986   35,014   22,775
1987   34,807   22,914
1988   34,253   23,232
1989   33,965   23,471
1990   32,859   23,348 << recession year
1991   33,003   23,117 << recession year
1992   32,568   23,337
1993   31,873   23,044
1994   31,612   23,265

Very generally, you can see that male median income has been falling since 1973, and that female income has been rising. The men's loss is roughly the same as the women's gain, but that doesn't mean they cancel each other out. The workforce was 61 percent male in 1970, 57 percent in 1980 and 55 percent in 1990. So overall individual median income has been declining slightly over the decades.

If you look closely at the men's falling and women's rising trends, you can see evidence of the business cycle. Incomes tends to dip in recession years, and recover in-between. But these cycles did not overcome the larger trends.

As before, this chart does not indicate how much overtime people worked to compensate for falling wages.

Let's again compare median individual income with a chart on average income. Once more, remember that the larger the gap between average and median income, the more income is concentrated near the top:

Average Income of year-round, full-time workers (1994 Dollars) 

Year   Men      Women
1972  $39,376  $21,783
1973   39,622   21,703
1974   39,283   21,942
1975   38,801   21,787
1976   39,173   22,345
1977   39,697   22,359
1978   40,190   22,840
1979   39,924   22,927
1980   38,600   22,925
1981   38,192   22,814
1982   38,459   23,473
1983   38,318   23,798
1984   38,852   24,345
1985   39,594   24,913
1986   40,739   25,581
1987   41,035   26,124
1988   40,886   26,398
1989   41,527   26,687
1990   39,772   26,524
1991   39,235   26,449
1992   39,293   26,662
1993   40,883   27,402
1994   41,099   27,816

Average male income has remained steady at about $39,000 a year -- but we've just seen that male median income is falling. This tells us that men's inequality is growing dramatically. Women inequality is growing too, but nothing quite like men's.

A chart that is much more informative in this debate is income by quintile. (The top 5 percent has been added because that is where most of the income growth occurred.)

Mean income of households by quintile (1994 Dollars)

       Quintiles:
Year   First   Second   Third    Fourth   Fifth    Top 5%
1967  $6,638  $18,098  $28,897  $40,430  $73,267  $116,784 
1968   7,202   19,034   30,186   42,113   73,754   114,189
1969   7,361   19,620   31,351   43,911   77,184   118,808
1970   7,281   19,359   31,176   43,947   77,810   119,432
1971   7,310   19,012   30,826   43,824   77,652   119,100
1972   7,730   19,687   32,127   46,119   82,798   128,330
1973   8,063   19,988   32,661   46,953   83,271   126,903
1974   8,312   19,911   31,999   46,205   80,693   118,985
1975   8,001   18,997   31,082   45,138   78,607   115,870
1976   8,178   19,411   31,841   46,209   80,637   119,271
1977   8,238   19,442   32,058   46,941   82,286   121,449
1978   8,358   20,006   32,955   48,258   85,168   126,519
1979   8,239   20,069   33,035   48,451   86,096   128,568
1980   8,073   19,482   32,066   47,213   82,929   119,959
1981   7,954   19,062   31,484   46,898   82,147   116,940
1982   7,756   18,971   31,306   46,547   83,966   122,381
1983   7,795   19,040   31,403   47,119   85,264   124,903
1984   7,996   19,447   32,160   48,417   87,933   129,271
1985   7,984   19,737   32,691   49,162   90,684   136,281
1986   8,037   20,230   33,776   50,872   95,113   145,285
1987   8,045   20,331   33,991   51,378   97,709   153,940
1988   8,148   20,441   34,189   51,681   98,665   155,610
1989   8,391   20,797   34,570   52,292  102,221   165,153
1990   8,158   20,444   33,768   50,913   98,804   157,335
1991   7,903   19,748   32,803   50,006   95,895   149,649
1992   7,698   19,205   32,356   49,669   96,240   152,751
1993   7,602   19,134   32,073   49,843  103,846   178,234
1994   7,762   19,224   32,385   50,395  105,945   183,044

The first thing that stands out about this chart is that income stagnated in the first three quintiles, grew somewhat in the fourth and fifth quintiles, and skyrocketed in the top 5 percent. In fact, if this chart had shown the top 1 percent, you would have seen exploding incomes.

Supply-siders attempt to use this chart to show that income increased for all quintiles during the Reagan years. But this chart suffers from all the usual drawbacks: it doesn't reflect increased overtime, it doesn't reflect a growing percentage of workers in those households, and how a president was doing depended on which part of the business cycle he was in.

And if you really want to bug the living daylights out of a supply-sider, just point out that the poorest quintile actually did better under Carter than under Reagan. The average income of the poorest quintile was $8,076 from 1978 to 1982 (Carter's four budget years plus the recession year of 1982, which supply-siders incorrectly blame on Carter). But the poor averaged slightly less -- $8,056 -- during Reagan's so-called "Seven Fat Years" from 1983 to 1989. So, incredibly, the poor did even better during the worst recession since World War II than under Reaganomics!

In conclusion, here are some important questions to ask before accepting any income comparison:

To cut through much of the confusion that surrounds this debate, Congress published a study in 1992 about eroding family incomes during the Reagan years. Here is what they found:

"This report takes a close look at one important type of family -- two-parent families with children -- in two years, 1979 and 1989, and concludes that such families have a legitimate basis for their growing economic concern:

[Congressional Study: Families on a Treadmill: Work and Income in the 80's, January, 17, 1992.]

Return to Overview