Last week on this page, I wrote: “If facts could kill the supply-side myth, it would be dead already.” This week, I’d like to explore how, despite the lack of empirical evidence, this myth not only lives but flourishes.

My last post was based on testimony before the Joint Economic Committee, part of which drilled down on this false notion that tax cuts generate anywhere near enough growth to offset the revenue they lose. While there are certainly conditions under which tax cuts can generate some degree of growth, the empirical record fails to find economically large, lasting, offsetting growth effects that occur through the supply-side channels of greater labor supply and capital investment.

Art Laffer, a progenitor of supply-side theory, was one of the Republicans’ witnesses, and as you’d expect, he disagreed strongly with my attack on his life’s work.

Now, before I go any further, let me stress that I’ve been good friends with Art — an important and highly influential economist, not to mention a great guy — for decades. We’ve long disagreed, but we’ve done so, including last week, without being disagreeable. I hope the facts eventually move him, but if they don’t — and they haven’t yet — I hope to keep arguing with him for years to come.

But let’s take this moment of disagreement in a joint Senate/House hearing as a microcosm of how a bad idea with so little evidence survives. I do this not to play “he said, she said” but to try to get to the bottom of the important problem of destructive ideas that won’t die.

In the course of our argument at the hearing, Art said:

“If you take the nine states that have no earned income tax in the U.S., if you take the nine states with the highest tax rates, if you look at their growth rates, over the last 50 years, every single year the nine states without earned income taxes have grown much faster in every single metric than have the nine states with the highest income tax rates.”

This, thanks to Art’s characteristic clarity, is an easily testable statement. And it is, of course, completely false. In 2015 alone, the combined performance of the “nine states with the highest tax rates” was similar to and sometimes better than the combined performance of the “states that have no earned income tax” on a number of economic variables, as shown in the graph below. In most cases, the bars between the two groups of states are the same heights, though real wages grew more in 2015 in high-tax states (EPOPs are state employment-to-population ratios).

The next figure shows that there is substantial variation among states in each category. Some no-income-tax states, such as Wyoming, lost employment in 2015, while some high-income-tax states, such as California, saw relatively strong growth. Go ahead and check the past 50 years if you like, but Art’s outlandish claim is wrong in the most recent year for which we have data.

This should not surprise anyone: Taxes are but one factor, and often a minor one, driving growth. And, to be clear, these one-year changes are far from dispositive evidence of that point; they simply show it takes but a few mouse clicks to easily belie Art’s claim (see my blog for similar figures over longer time periods). The best way to do a more complete analysis is to test the statistical relationship between state tax changes and various growth measures over the course of many years, as Gale et al do here, yielding similar findings in recent years to those I’ve shown above for 2015.

Yet conservatives continue to argue otherwise, invariably without evidence.

For example, each year, Art collaborates with the conservative state-level policy organization ALEC on the “ALEC-Laffer State Economic Outlook Rankings.” With a heavy emphasis on state tax policy, their rankings are supposed to forecast economic performance “based on a state’s current standing in 15 equally weighted policy areas,” including personal, corporate, property, sales and estate taxes, as well as states’ willingness to deprive workers of the minimum wage and the right to collectively bargain. Art and ALEC predict that states that do “better” on this metric (i.e., have lower rankings) will “tend to experience higher rates of economic growth than states that tax and spend more.”

Except they don’t. I plotted each state’s 2015 ranking against its subsequent economic performance that year in scatter plots, shown below.  If Art and ALEC were right, the dots would decline from the upper left to the lower right, indicating that lower-tax states had better economic outcomes and higher-tax states had worse outcomes.  Instead, there is no correlation between the rankings and GDP and jobs growth (see my blog for similar charts on wages and EPOPs).

I’m not the first person to challenge these assertions. Figures from the Institute on Taxation and Economic Policy and economist Peter Fisher, respectively, show the same story as that above in previous years. No-income-tax states do not outperform high-income-tax states, and the ALEC-Laffer rankings have been consistently terrible at predicting state economic performance.

This argument is not theoretical. Politicians at both the national and state level listen to supply-siders such as Art, and numerous states, most notably Kansas, have cut taxes sharply in response. The results, as you’d guess based on the evidence I’ve been trotting out, are predictable. From my testimony:

“The cuts in Kansas that took effect in 2013, for example, have now blown a $400 million hole in the state’s budget.  When [Art] helped design these cuts, he predicted (along with Stephen Moore of the Heritage Foundation) that they would provide an ‘immediate and lasting boost’ to the Kansas economy. Yet not only have the cuts caused serious underfunding of the state’s education system, they’ve also coincided with weak job and GDP growth. The Kansas Legislative Research Department’s projections suggest that the economy will remain weaker than the overall US economy for the foreseeable future.”

How can it be that these ideas remain as strong as ever, showing up, for example, in the tax plan of almost every Republican running for president in recent years, including George W. and Jeb Bush (the former pushed through aggressive tax cuts early in his term, yet GDP, job and wage growth in the 2000s was uniquely weak), Mitt Romney, Ted Cruz and Donald Trump?

Of course, a big piece of what’s going on here is that Art and others are telling politicians exactly what they long to hear: that they can cut taxes and not worry about the budget. That’s like telling them they can eat all they want and not gain weight, especially if they ignore the scale.

But the other problem is that there’s no mechanism to evaluate such claims in political settings. Art and I disagreed, shook hands, went our separate ways, and no one was the wiser.

Considering the stakes, we must do better. Here’s a thought: When, in a congressional hearing, there’s a clear, testable, factual disagreement about the economic record, the chair or ranking member should be able to refer the matter to one of our statistical agencies, such as the Congressional Budget Office or the Bureau of Labor Statistics. They would then adjudicate the factual dispute and, if a witness were found to be wrong, after giving her a chance to defend her claim (with the possibility of overturning the agency’s ruling), the public record would reflect the correction.

I don’t know whether that is practicable, but it’s all I can think of. If anyone’s got a better idea, let me know. Somehow, we’ve got to find the way back to Factville.