Book Review for Not Working: Where Have All the Good Jobs Gone?
Book Review for Not Working: Where Have All the Good Jobs Gone?
复制标题
《不工作：好工作都去哪儿了？》书评
作者：
David Wiczer
在从大衰退中复苏的整个过程中，人们普遍认为劳动力市场的复苏速度比一般措施所显示的要慢。即使失业率下降到按衰退前的标准看起来“正常”的范围内，金融媒体和政策制定机构的许多人仍持怀疑态度。一个常见的问题是，统计数据不再是对劳动力状况的可靠总结，复苏仍然不完全，需要进一步的扩张性政策。长期失业率、工作转换率和工资增长率。劳动力参与率（其周期性很少成为焦点）持续低于大衰退前的水平，并成为挥之不去的担忧来源，这可能表明劳动力市场表现不佳。不工作：所有的好工作都去了哪里？作者认为，失业率，这是我们对劳动力市场状况最有用的指南，不再那么有意义。正如本书第二章令人信服地指出的那样，部分原因是痛苦可以与低失业率共存。具体来说，尽管失业率很低，人们还是担心失业，减少了冒险。但主要的原因是，失业率不再像以前那样与工资增长相关，因此用处不大。如果有人认为劳动力市场的“疲软”和工资增长应该是负相关的，那么失业-工资增长关系的破裂表明失业不再是疲软的指标。虽然失业率的变动可能是整体劳动条件的一个有用的比喻，但工人进入和脱离失业状态的变动在机械上并不是工资增长的可能来源。这是因为失业工人倾向于离开相对低工资的工作，然后回到低工资的工作，因此中心时刻（平均值和中位数）几乎没有移动，正如Daly和Hobijn（2016）或Hahn等人令人信服地表明的那样。“就业不足”被认为是一个比失业更优越的上级指标。在美国，这本书的就业不足概念是指由于劳动力需求不足以让他们找到全职工作而从事兼职工作的工人比例。这一系列，“经济原因兼职”（PTFER），使用了当前人口调查中一个长期被问到的问题，经常被用来补充劳动力市场的其他指标。事实上，这个确切的系列是联邦经济学家在大衰退复苏期间推出的两个广泛劳动力市场指数中的一个指标。来自堪萨斯城联储的Hakkio和Willis（2013）以及来自联邦储备理事会的Chung et al.（2014）都将其作为输入，分别为24或19。这些指数通过因子分析使用数据的协方差结构来确定每个变量应该获得多大的权重，以衡量整体劳动力市场的健康状况。这些研究还得出结论，失业率比其他指标更快地从大衰退中恢复过来，因此可能描绘了劳动力市场过于乐观的景象。事实上，这些其他指标与兼职工作密切相关，布兰奇弗劳尔是兼职工作的支持者，他认为兼职工作是衡量懈怠的一个指标。图1显示了与PTFER相邻的Hakkio and Willis指数的“动量成分”，其中指数乘以了与PTFER标准差相反的-16.93，使其具有相同的规模，因为“改善”意味着指数上升，但就业不足率下降。 * 大卫威策david. stonybrook.edu
Throughout the recovery from the Great Recession, there was a common refrain that the labor market was recovering even more slowly than common measures would suggest. Even as the unemployment rate fell into ranges that would seem “normal” by pre-recession standards, many in the financial press and policy-making institutions were incredulous. A common refrain was that the statistic was no longer a reliable summary of labor conditions and the recovery was still incomplete and warranted further expansionary policy.1 Many labor market indicators did recover more slowly than the unemployment rate, e.g., the rate of long-term unemployment, the rate of job-to-job transitions, and wage growth. Labor force participation, whose cyclicality was rarely a focus, remained persistently below its pre-Great Recession levels and became a lingering source of worry, potentially indicating that the labor market was under-performing its unemployment rate. Not Working: Where Have All the Good Jobs Gone speaks to these sentiments. The author suggests that the unemployment rate, which was our most useful guide to the state of the labor market, is no longer so meaningful. This is partly because misery can coexist with low unemployment, as the book convincingly argues in its second chapter. Specifically, despite low unemployment people worry about job loss and take fewer risks. But principally, unemployment is less useful because it does not correlate with wage growth as it once did. If one believes that labor market “slack” and wage growth should be inversely correlated, then the breakdown of the unemployment-wage growth relationship suggests that unemployment is no longer an indicator of slack. While movements in the unemployment rate may be a useful synecdoche for overall labor conditions, movements of workers in and out of unemployment is mechanically not a likely source of wage growth. This is because unemployed workers tend to have left relatively low-wage jobs and then return to low wage jobs, and thus move the central moments (mean and median) little, as convincingly shown by Daly and Hobijn (2016) or Hahn et al. (2018). Instead of unemployment, “underemployment” is supposed to be a superior indicator. For the US context, the book’s notion of underemployment is the fraction of workers who are working part-time because labor demand is insufficient for them to find full-time work. This series, “part-time for economic reasons” (PTFER), uses a long-asked question in the Current Population Survey and is often used to complement other indicators of the labor market. In fact, this exact series is an indicator in two of the broad labor market indices introduced by Federal Reserve economists during the recovery from the Great Recession. Both Hakkio and Willis (2013) from the Kansas City Fed and Chung et al. (2014) from the Federal Reserve Board of Governors used it as an input, one of 24 or 19, respectively. Those indices used the data’s covariance structure through factor analysis to determine how much weight each variable should get as a measure of the overall labor market health. Those studies also concluded that unemployment recovered from the Great Recession more quickly than other indicators, and thus might have painted an overly rosy picture of the labor market. And indeed, those other indicators are strongly correlated with the part-time work, of which Blanchflower is a proponent as a measure of slack. Figure 1 shows the “momentum component” of the Hakkio and Willis Index next to PTFER, where the index has been multiplied by − 16.93, which is the opposite of the standard deviation of PTFER, to give it the same scale, and because “improvement” means an increasing index but a falling underemployment rate. * David Wiczer david.wiczer@stonybrook.edu