Shlaes makes a different argument about numbers, because she uses different numbers. She starts each chapter with a rat-a-tat of just-the-facts, but instead of GDP, which represents the overall economy, she quotes the Dow Jones Industrial Average, which represents the maybe 10 percent of Americans who owned stock. And though she quotes an unemployment number, she doesn't quote the figures I've just mentioned. Instead she chooses different estimates of unemployment that (she acknowledges) show a much larger share of Americans out of work during the New Deal.
If you want to know how the New Deal treated ordinary Americans, this choice really matters. Let's look at a figure Shlaes gives twice in her book and again in her Wall Street Journal editorial: She has unemployment at 20 percent in the 1937-38 recession. That's appalling—almost as bad as 23 percent in 1932. Based on such a statistic, you could think the New Deal wasn't alleviating the Great Depression. But that number hides something: A third of the people Shlaes counts as unemployed had a job that the New Deal gave them through its relief programs.
Now, you may say, wait: Those people really shouldn't count as employed—we're not interested in government make-work, we're interested in the real economy. Fair enough—and if you look again at Historical Statistics of the United States, you'll see another measure of unemployment—private, nonfarm unemployment—measuring the real, industrial economy. And on that measure, unemployment again runs markedly lower under Roosevelt than under Hoover. John Maynard Keynes might have explained that the New Deal wasn't just offering make-work, it was stimulating the economy—and Shlaes in fact at one point says the same: "[I]t functioned as Keynes ... hoped it would." Yet of all the possible ways to measure unemployment, Shlaes chooses the only way that hides the effect of New Deal relief programs and makes it look as though the economy performed as poorly under Roosevelt as under Hoover.
Roosevelt did make mistakes, the National Recovery Administration—which let industry cartels set noncompetitive, price-raising codes—chief among them. Roosevelt admitted NRA was "pretty wrong," and historians generally agree. But the NRA lasted less than two years and did not typify the New Deal, which mostly protected Americans' rights to strike bargains for themselves in a marketplace it left largely intact. Laws like these—for example, the Wagner Act protecting the right to unionize—kept the state small, as Sen. Robert Wagner explained: "[W]e intend to rely upon democratic self-help by industry and labor instead of courting the pitfalls of an arbitrary or totalitarian state." That was Roosevelt's legacy.
Shlaes prefers Calvin Coolidge, whom she calls a "minimalist president." This might surprise Coolidge, who declared, "The nature of man requires sovereignty. Government must govern. To obey is life. To disobey is death." Coolidge backed an increase of federal power to restrict immigration, and under him the federal payroll grew by 40,000 employees (after shrinking under Warren Harding).
But for Shlaes, as for the Liberty Leaguers, government isn't big unless it restricts big business; then big government is bad. She makes this point plainly: "[T]o force business to go on spending when it did not want to was to hurt business," she says of wage-raising policies. Yes, and to force labor to go on starving when it did not want to was to hurt labor. Sometimes government, as Coolidge said, must govern, which means balancing one interest against another. Roosevelt made this homely insight the hallmark of his administration. And it's why the American people kept him in office and why so much of his New Deal remains.