Bash big firms very carefully

Barely eight years after Steve Jobs’ death elevated him to instant secular sainthood for having founded a worldchanging company, tech firms have become secular devils. Sen. Elizabeth Warren has made taxing them and curbing their power key planks in her plank-heavy election platform, while the U.S. Justice Department — possibly to blunt Sen. Warren’s appeal to the populists Donald Trump considers his own — has just opened what the Wall Street Journal calls a “broad, new antitrust review of big tech companies.”

The Journal says the review’s goal is to see whether they are “unlawfully stifling competition.” That’s fine. U.S. law proscribes anticompetitive behaviour of several kinds. If big tech firms are engaging in such behaviour, the Justice Department should go after them. That’s its job.

But populists hate these firms, not because of anything specific they do, but because they are big and successful and the people who run them are super-rich members of the upper slivers of the income distribution. Inequality has been rising in the U.S. and so has corporate concentration, ergo: if you go after the big firms, you make the U.S. fairer and more egalitarian.

Well, you might believe that, says “Markups and Inequality,” a new study by two economists at New York University, but it ain’t necessarily so. The two, Corina Boar and Virgiliu Midrigan, build a plausibly realistic model of corporate and entrepreneurial competition in the U.S., set their model up to mimic the behaviour of the actual economy in several important ways, and then demonstrate that when some of the more popular populist policies are introduced, funny things happen.

For instance, a 25 per cent tax on profits reduces the number of corporations. No surprise there. But because bigger firms are more efficient, the tax also reduces productivity by 3.6 per cent, which deadens the demand for labour. Wages end up falling 6.4 per cent. Inequality does abate, with the income share of the top one per cent falling from 17 per cent to 15 per cent. But “only 29.4 per cent of all households gain from the introduction of a profit tax. The median household experiences a … loss equivalent to a 0.5 per cent permanent drop in consumption.”

You might think that if you restricted new profits taxes to the largest firms, as Sen. Warren proposes to do, you’d get different effects. And you do. But they’re actually worse. Suppose you tax only the top 0.5 per cent of profit-earning firms, the real business stars, who despite being such a small percentage of firms nevertheless account for half of U.S. corporate profits. Once you include the effects on productivity and the demands for labour and capital, such a tax ends up making only five per cent of households better off. The median household experiences a 0.6 per cent loss.

Taxing just the biggest and best firms does help midsized entrepreneurs, “who can expand production due to the drop in factor prices (i.e., the decline in wages and interest rates because of reduced output) and decreased competition from corporate firms.” The authors continue: “The policy therefore has a redistributive component, transferring resources from workers to privately-held business owners, a third of whom benefit from the reform.” Make workers pay! Privilege the haute bourgeoisie! It’s not quite what the anti-corporate mob thinks it’s working for.

In general? “Policies that reduce firm concentration lead to large output and (productivity) losses and increase inequality.” Get that? The populist policies that are the flavour of this U.S. election cycle — and maybe our own, too — can actually increase inequality. It takes real courage in the current political, intellectual and social media climate to publish orthodoxy-challenging conclusions of that sort. I hope my mentioning the professors’ real names doesn’t lead to social-media trolling of them. (How can a medium that enables trolling really be social?)

People often object to economic models. But if you’re dealing with any reasonably complicated set of interactions, you need to spell out the links as carefully as you can and then alter the initial conditions to see what happens. How else could you proceed? Intuition? What if your intuition generates different results than mine? What do we do? Arm wrestle to establish who’s right?

It goes without saying that a model’s results are only as good as the assumptions underlying it. Professors Boar and Midrigan’s assumptions seem reasonable. They’ve got a corporate sector with relatively easy access to capital. They’ve got entrepreneurs of varying abilities whose access to capital is constrained, as most entrepreneurs complain it is. And they assume productivity patterns by firm size that look like those observed in the U.S. economy.

As good academics, they would be the first to admit their model could be wrong. Maybe they’ve assumed entrepreneurs are overly important to production, or overly sensitive to tax rates. Exactly how different assumptions could give rise to different results — results more like Sen. Warren and others would prefer — will be part of the normal reaction to their work.

But how much would you be willing to bet that they’re wrong? Six per cent of your wage? A half point permanent reduction in your consumption? A 3.6 per cent decline in your productivity? Some candidates don’t seem to understand that when you start remaking a real economy, not a model, real people can get hurt.