A little more light is shed on the implications of the US Census Bureau’s data if one considers the US Internal Revenue Service data. To its credit, the IRS acknowledges the adjusted gross income (AGI) it uses is defined as that from IRS Form 1040, Line 41. Therefore, we have a standard definition of income. The IRS also uses a more informative income bracket break-down that categorizes incomes greater than the US Census Bureau limit of $100,000.
Unfortunately, the IRS will only release its income data as a comparison of “Number of
Returns”, which includes returns filed singly and jointly; without stating how many of each were filed. And while the IRS uses income ranges of $5,000 for incomes below $30,000 or $60,000 (depending on which version of the data one has); the income ranges increase as incomes increase, thus giving a visually misleading inference of what the middle income might be.
Instead of thirty-five income brackets up to “$100,000 and over”, as used by the US
Census Bureau, the IRS uses twenty-two or nineteen income brackets (depending on the version of the data) up to “$10,000,000 or more”. But, like the US Census Bureau, there are more income brackets at the low end of the income spectrum than at the high income end. While the middle bracket on the US Census Bureau table is “$40,000 to $42,499” the middle bracket on the IRS table is “$50,000 under $55,000” or “$50,000 under $75,000” (depending on version).
In both cases, one is led to the visual impression that the middle income bracket may be the middle of the middle class. And, one can be led to believe the following idea:
“Since the US Census Bureau median income is $26,197, any income over $26,197 is ‘above average’ and, since the US Census Bureau mean is $49,445, any income around $49,445 is middle class and the middle class is doing better than most”.
Both the US Census Bureau and the IRS tables support this inference. It’s almost as if the US Census Bureau data establishes expectations and the IRS data shows those expectations are surpassed.
But why did the IRS revise its earlier table?
A possible answer might lie in what the ” # returns” column seemed to indicate in the earlier version. The first three AGI ranges show about 12,000,000 returns. Then the number of returns declined from the $15,000 to $60,000 AGI and then jumped up significantly for $60,000 to $200,000 AGI.
This makes it appear the “middle class” AGI returns are fewer in number while “low” and “high” income AGIs are greater in number. Therefore this table might imply a dramatic income inequality with a shrinking middle class.
The revised table divides the lowest AGI range, so it no longer has the second greatest number of returns, while combining several “middle income” brackets, so that nearly all AGI ranges have no less than 10,000,000 returns. Also, the visually middle AGI range (“$50,000 under $75,000”) has the greatest number of returns.
While this revision somewhat implies “middle class” is from $50,000 to $75,000 (contradicting the US Census Bureau by 250-300%) it also resolved the implication that there’s a great income disparity and a shrinking “middle class”.
It appears the IRS made a value choice of the ‘lesser of two evils’; in that, “yes, the middle class makes more than we’d prefer you to believe but, no, the middle class is not shrinking…in fact, they’re the largest group and they have better than expected incomes.”