半世紀以上に亘る勤務日の推移:週4日労働の台頭

というNBER論文をハマーメッシュらが書いているungated版)。原題は「Days of Work Over a Half Century: The Rise of the Four-day Week」で、著者はDaniel S. Hamermesh(テキサス大オースティン校)、Jeff Biddle(ミシガン州立大)。
以下はその要旨。

We examine patterns of work in the U.S. from 1973-2018 with the novel focus on days per week, using intermittent CPS samples and one ATUS sample. Among full-time workers the incidence of four-day work tripled during this period, with over 8 million more full-time workers on four-day weeks. The same growth occurred in the Netherlands, Germany, and South Korea. The rise was not due to changes in demographics or industrial structure. Four-day full-time work is more common among less educated, younger, and white non-Hispanic workers, among men, natives, and people with young children; and among police and firefighters, health-care workers, and in eating/drinking places. Based on an equilibrium model of its prevalence, we show that it results more from workers’ preferences and/or daily fixed costs of working than from employers' production costs. We verify the implication that the wage penalty for four-day work is greater where such work is more prevalent, and we show that the penalty has diminished over time.
(拙訳)
我々は、不定期なCPSのサンプル*1と一つのATUSサンプル*2とを用いて、週何日働くかに新たに焦点を当てて、1973-2018年の米国における労働パターンを調べた。フルタイム労働者の中で週4日労働の割合はこの期間に3倍となり、800万以上のフルタイム労働者が週4日に加わった*3。同様の伸びはオランダ、ドイツ、韓国でも生じた。この伸びは人口動態や産業構造の変化によるものではなかった。週4日のフルタイム労働は、学歴が低く、若く、非ヒスパニックの白人労働者で、男性で、米国出身であり、幼い子供を抱えている人の間でより一般的であった。また、警察、消防、医療従事者、飲食業でより一般的であった。その普及の均衡モデルに基づき、雇用者の生産コストではなく、労働者の選好ないし日々の労働の固定コストによってそうした結果が生じていることを我々は示す。週4日労働が普及しているところほどその賃金ペナルティが大きいということの含意を我々は確認するとともに、そのペナルティが時間とともに減少していることを我々は示す*4

*1:(ungated版の)本文には、「In May of selected years, the U.S. Bureau of Labor Statistics included a Work Schedules Supplement to the May Current Population Survey (CPS). These data are available for 1973, 1985, 1991, 1997, 2001, and 2004 (with 48,356; 25,858; 48,409; 45,375; 38,430; and 45,494 observations respectively). 」と記述されている。

*2:本文には、「The Supplement has not been fielded since 2004. In 2017 and 2018, however, the only times in its history, the American Time Use Survey (ATUS) asked respondents who stated that they worked for pay, 8,780 workers in the two years, specifically whether they worked on Sunday, Monday, ... , and/or Saturday. (For convenience we refer to the ATUS sample as being for 2018.) Since the ATUS is an extension of the CPS, it provides the same demographic and other information as in the May Supplements and information on hours usually worked on the main job (Hofferth et al., 2018).」と記述されている。

*3:本文には「They represent an additional 8.2 million 30+ hour four-day workers in 2017-18 compared to the number in 1973, and an increase of 6.0 million 40+ hour four-day workers (and implicitly 2.2 million additional workers putting in 30-39 hours per week).」と記述されている。

*4:本文の結論部では「Within this framework, a wage penalty for four-day work is consistent with an excess supply of workers to such job schedules. The secular decline in this penalty, however, suggests that employers' demand for such schedules has been rising relative to workers' supply.」と記述されている。