...when President Roosevelt (FDR) backed the wrong horse.  He rejected a 30-hour maximum workweek and went instead with a minimum wage and maximum makework - and anything else he could throw together to placate the shorter-hours forces without giving them (and American labor in general) the one thing they needed to survive = control over the supply of...themselves...which they could have by controlling worktime per person.
          Once Hoover's "No Controls" path was rightly rejected, the Great Fork in the Road faced us with two choices, One Control or Many Controls.  Unfortunately FDR careened the country from one extreme to the other, from No Controls to Many Controls, and missed the One Control option in the middle.  So FDR -

• rightly rejected the path on the right (No Controls is a mirage, because the most aggressive players rush into the vacuum & bend the "free" market their way),
• wrongly recoiled from the path in the middle (Art Dahlberg's one planned adjustment to make capitalism hum, the shorter-workweek bill that Sen. Hugo Black pushed through the US Senate on April 6th) mistaking it for socialism, and
• tragically lurched down the only path left (left-leaning Rex Tugwell's burgeoning bureaucratic micromanagement, as good an imitation of socialism as you can get),
• only to start regretting it two years later ("voiced regret that he had not got behind the 30 hour bill and pushed it through Congress"),
• and never solved even half the Depression before the War.  The Depression's unemployment rates (& sequel) were 1929:3.2%, 1930:8.7, 1931:15.9, 1932:23.6, 1933: 24.9%, 1934:21.7, 35:20.1, 36:16.9, 37:14.3, 1938: 19.0%, 39:17.2, 1940:14.6, 41:9.9, 42:4.7, 43:1.9 (from Ross Robertson's History of American Economy, 3rd Ed., p.682) & 44:1.2%, 45:1.9, 46: 3.9, 47:3.9, 48:3.8, 49:5.9, 1950:5.3, 51:3.3, 52:3.0, 53:2.9 (from Historical Statistics of the US, Table D85-86).  With a 1-year lag, the New Deal started in 1933 and got unemployment down to 14.3% by 1937, not even half the 1933 rate.  Worse, it slipped back up to 19% the next year.
  In 1938-40, weak timesizing (too high, too late) was finally enacted, yet it still worked to cut at least 2% unemployment for each 2-hour cut from the workweek, similar to the more recent 4-hour-cut results in France.  In the U.S., a nationwide 44-hr/wk cap was instituted on Oct. 24, 1938.  The 1938-39-40-41 figures in the U.S. are: undefined hrs 19.0%, 44 hrs 17.2%, 42 hrs 14.6%, 40 hrs 9.9%.  So workweek definition in late 1938 yielded a nearly 2% drop in 1939 to 17.2%.  The 42-hr/wk cap in late 1939 yielded at least another 2% drop the next year to 14.6% (the extra drop possibly ascribable to Britain's declaration of war in August/39).  The 40-hr/wk cap took hold on Oct. 24, 1940, and yielded at least another 2% drop in 1941 to 9.9% (the extra drop now ascribable to the formalization of Lend Lease on March 11 and to Pearl Harbor on December 7).  In France between 1997 and early 2001, the workweek was cut from 39 to 35 hours and unemployment went from a high of 12.6% when workweek reduction was voted in in '97, to a low of 8.6% in the spring and early summer of 2001 before the US-led recession hit France and dragged its unemployment rate back up.
  In the United States, World War II got "better" results than timesizing by drafting millions of surplus labor hours out of the job market - for war service - and thus gave the coup de grace to the Great Depression - which timsizing had already solved - by reducing unemployment to less than 2% by 1943 (1942: 4.7%, 1943: 1.9%).
  So the over-credited, "feel good" New Deal of FDR did not solve the Great Depression.  Timesizing did.  But then its role was obscured as the War finished the job in the worst possible way - by killing and maiming large portions of the workforce.  The US labor surplus was not restored until 1965-75 when the postwar babyboomers entered the job market.  The labor surplus's wage-flattening, consumption-damping, recession-inducing effects were augmented by the by-now rebuilt factories of Europe and Japan and the consequent reduction of US exports.
  The lesson - you get roughly a one-percent drop in the unemployment rate for every one-hour cut in the workweek - was overlooked, forgotten and by the 1980s, completely denied and even ridiculed as the "Lump of Labor Fallacy."  Why?  One reason - if we can automatically guarantee full employment by automatically adjusting the workweek so that unemployment maintains a predetermined target level, automatically reducing as long as unemployment is too high, etc.), then we don't need economic micromanagement and so many economists....

1. Hoover wanted to stay with only a political definition of "share," the one-person one-vote principle that hadn't prevented the Depression. This was the last time we could equalize on a point (one:one).

2. FDR wound up trying to get an economic definition of "share" by defining everything but an upper limit, hence his Many Controls. Hey, at least he got the focus moved over from the political to the economic sphere. He also got clear on the concept that we could equalize on a range (with two required definitions: an upper and a lower limit), not just a point (with only one required definition).  Definitions of lower limits popped up like mushrooms:  Minimum wage, minimum income (federal poverty line), under-time (for state unemployment insurance benefits), disability (for workmen's comp), under-age (for child labor prohibition).... One peripheral upper limit even popped up:  Over-age (for purposes of Social Security retirement benefits).  And at last, too little too late, a central upper limit appeared:  Over-time was defined as hours over 44/wk in 1938, 42 in '39, and 40 in '40 with 40¢/hr minimum wage - the famous 40-40-40 Plan, whose overtime provisions motivated overtime as much as they discouraged it, and whose enforcement quickly lapsed as World War II geared up.

3. Art Dahlberg and the AFL and Hugo Black got an economic definition of "share" by defining nothing but an upper limit, hence their One Control.  Kellogg's, Sears Roebuck, GM/Tarrytown, Standard Oil/NJ, Hudson Motors, several cotton mfrs and uncounted small firms were doing it successfully in practice at the 30-hour level to avoid layoffs (Hunnicutt, 148). And it's been independently reinvented by countless small firms in every recession since then, in the form of "cutting hours, not jobs", i.e., timesizing, not downsizing.