I'm referring to the long term policy of tax cuts, starting in the 1980s. The data set you link to has real median income (after inflation) going up by about 40% since 1975, under 1% per year.

That's not zero, but compared to GDP, the stock market, and other measures of overall economic health -- which have gone up by an order of magnitude or more -- the median is practically immobile.

And for much of that time, it was even more immobile. The median was $31k in 1997, and was $31k in 2014. It bottomed out at $30k in 2012; it has been rising since but that was almost two decades of complete stagnation -- while GDP rose and stock markets climbed.

Here is an image that plots the effective capital gains rate from the 1950s to recent years: https://en.m.wikipedia.org/wiki/Capital_gains_tax_in_the_Uni...

Let's analyze three periods:

1953-1985: no significant trend in effective capital gains tax

1985-1996: upward trend and higher capital gains tax

1996-2020: downward trend

Now let's compare those periods to real median family income (which tracks to 1953). https://fred.stlouisfed.org/series/MEFAINUSA672N

Annualized growth in real median income:

1953-1985: 1.8%

1985-1996: 0.8%

1996-2020: 2.0%

Real median income growth during lower capital gains tax policy was twice that of the period of higher capital gains tax policy.

Your choice of a thirty-two year period, and 11 year period, and a 24 year period, when the unequal periods aren't supported by some kind of consistent in-period patterns in the conjectured independent variable, plus using fuzzy subjective descriptors for the independent variable, suggest a breathtaking degree of cherry picking and desired-results-driving-methodology, here.

They are unequal periods, and that's why we annualize the growth.

The periods aren't "cherry-picked." If you'd actually look at the plotted effective capital gains rate, you'd see a clear difference in tax policy in those three periods.

The "fuzzy subjective descriptors" are totally in line with the parent poster's: "long term tax policy."

> The periods aren't "cherry-picked." If you'd actually look at the plotted effective capital gains rate, you'd see a clear difference in tax policy in those three periods.

The first period includes within it a perfectly flat period longer than shortest of the periods you chose followed by a period of equal length that is more consistently increasing than the (mostly flat) period you've held up as an increasing trend, and then a period nearly as long as your shortest that is as consistently and more sharply decreasing than the one you hold up as showing a decreasing trend.

So, no, I don't see natural breakpoints in the data.

(You've kind of muddied whether you are more concerned with the actual cap gains tax rate or the trend in changes to the rate, seeming to lean a little more heavily on the former, so I focussed on that in my criticism; but if you care about the rates instead of the deltas the periods you chose still make no sense, for similar reasons.)

> The first period includes within it a perfectly flat period longer than shortest of the periods you chose followed by a period of equal length that is more consistently increasing than the (mostly flat) period you've held up as an increasing trend, and then a period nearly as long as your shortest that is as consistently and more sharply decreasing than the one you hold up as showing a decreasing trend.

It sounds like you're referring to the "maximum tax rate on long term gains." As I stated, I used the "effective capital gains tax rate," which is the average rate that is being taxed (long term cap gains tax is different than short term).