The Democrats now control the House of Representatives, and their more radical members have wasted little time in putting forward a slew of the most progressive policy proposals. At present, the idea receiving the most coverage from talking heads on television and the Twitter accounts at Vox has been freshman Representative Alexandria Ocasio-Cortez’s proposal to raise the top marginal income tax bracket for individual earners to 70%.

Conservatives, naturally, were quick to come out against the proposal, while progressives pointed to a history they believed worked in her defense: “In the 1950s we had 90% tax rates and the economy was fantastic,” has been the oft-repeated talking point. This statement is true, but only semantically. The U.S. did have a nominal top marginal rate of 90% throughout much of the 1950s, but practically no one paid that rate, and because of myriad exemptions, deductions, carve-outs, and loopholes, the real, effective rate of taxation of top earners in that decade was actually lower than it is today. In this and many other ways, the U.S. tax system is more progressive now than it ever was in the supposed leftist nirvana of sky-high marginal tax rates. If we raised top rates back to even 70%, but kept the post-1986 reform structure in place, it would be disastrous for the U.S. economy.

Historically, there have been two interrelated problems with high marginal tax rates on top earners: 1) They depress economic activity, specifically, spending on capital investment and maintenance, and 2) They fail to raise much more income than lower, less confiscatory rates. But even if we take those problems out of the conversation, defenders of a 70% top marginal rate cannot point to history as their ally. Yes, it is true that the top rate in the 1950s was 90%, but it is also true that the average tax rate paid in the 1960s (inclusive of federal, state, and local income taxes, and social security taxes) was far lower than it is now — 25% in the 1950s versus 37% today.

Furthermore, the effective tax rate for top earners was far below the nominal 90% written into the Internal Revenue Code. Throughout the “golden age” of the 1950s, the average marginal rate paid by those making over $1 million fluctuated between 40% and 50%. The average rate paid by those making $500,000 to $1 million did the same. And the average rate paid by those making $50,000-$100,000 started the decade of the 1950s at 40% and dropped to 30% by 1960. Note that the nominal rate assessed against top earners did not change from 1945 to 1964; the top marginal rate was 91% for the entirety of that 19-year period, but the average effective rate paid never came close to that.

Representative Ocasio-Cortez has chided opponents on the right for their supposed failure to understand the marginal nature of the tax system. She has argued that it is not such an odd result to see average effective rates so far below the top marginal rate found in the code, because the top rate only kicks in at the highest levels of income. This is true, but it again ignores the facts of the 1950s tax structure. Sky-high nominal rates of taxation were not reserved for only the wealthiest in the United States. In 1954, for example, a rate of 50% kicked in at an adjusted gross income (AGI) of $16,000 (around $125,000 in 2018 dollars) and a rate of 30% kicked in at only $6,000 AGI ($56,788 in 2018 dollars). Therefore, the average effective rate paid by the wealthiest Americans did not exceed the nominal rates quoted for the upper middle class.

The truth is that no one paid 90% in taxes because there were so many loopholes and carve-outs of which to take advantage. The 1986 tax reform bill signed by President Reagan changed all that. The reform drastically cut rates for top earners, but also eliminated many of the exemptions and deductions that the highest earners had for so long enjoyed. The reduction in loopholes was so thorough that when the proverbial dust settled, the effective marginal rates paid by those in the top brackets was essentially unchanged. The effective rate for the top 1% of income earners in 2010 was only six points lower than the rate paid by the 1% in the 1950s; the top rate was 39.6% in 2010, it was 91% in the 1950s. Surely a less complicated system with a broader base and lower rates is superior to a complicated structure with dozens of brackets and sky-high rates that no one pays, especially if the former yields the same proportion of taxes paid as the latter.

One final data point to drive the argument home: no matter the marginal rate structure, in the post-war era the United States government has brought in tax revenues equal to between 18-20% of GDP. It does not matter whether the top marginal rate is close to 100%, or if it is in the low 30s. Whatever the ordering of brackets and rates, 20% of GDP seems to be an upper limit for tax collection by the federal government, at least as far as the income tax is concerned. Other developed countries bring in significantly more than 20% of GDP in tax revenue, but they do so by way of a value-added tax (VAT). These countries may also have high income taxes on top earners, but it is only a VAT that sufficiently disguises itself such that the substitution effect is mitigated long with dead-weight loss. A VAT is politically unpalatable in the United States, at present, and given its inevitable failure to bring in the revenue sought, so too should be a return to 1950s top marginal rates.