通常，美联储在谈到劳动力市场时，主要使用unemployment rate和nonfarm payroll这两个数据。不过最近，美联储主席Yellen在国会听证会上提到了一个新指标：Labor Market Condition Index，它从我们常用的一些数据中用动态因子模型把最有用的信息挖掘出来。这个做法类似于经济领先指标（Economic Leading Indicator）。
根据新指标，Yellen对目前（2014年7月）美国经济状况的解读是：“significant slack remains in the labor markets and wages are rising very slowly, all of which points to an economy which has not yet fully recovered from the Great Recession and still needs the sustenance of low interest rates.”
Assessing the Change in Labor Market Conditions
Hess Chung, Bruce Fallick, Christopher Nekarda, David Ratner1
The U.S. labor market is large and multifaceted. Often-cited indicators, such as the unemployment rate or payroll employment, measure a particular dimension of labor market activity, and it is not uncommon for different indicators to send conflicting signals about labor market conditions. Accordingly, analysts typically look at many indicators when attempting to gauge labor market improvement. However, it is often difficult to know how to weigh signals from various indicators. Statistical models can be useful to such efforts because they provide a way to summarize information from several indicators. This Note describes a dynamic factor model of labor market indicators that we have developed recently, which we call the labor market conditions index (LMCI).
Estimating the labor market conditions index
A factor model is a statistical tool intended to extract a small number of unobserved factors that summarize the comovement among a larger set of correlated time series.2 In our model, these factors are assumed to summarize overall labor market conditions.3 What we call the LMCI is the primary source of common variation among 19 labor market indicators. One essential feature of our factor model is that its inference about labor market conditions places greater weight on indicators whose movements are highly correlated with each other. And, when indicators provide disparate signals, the model's assessment of overall labor market conditions reflects primarily those indicators that are in broad agreement.
The included indicators are a large but certainly not exhaustive set of the available data on the labor market, covering the broad categories of unemployment and underemployment, employment, workweeks, wages, vacancies, hiring, layoffs, quits, and surveys of consumers and businesses. Table 1 lists the 19 indicators and provides some basic information about each indicator. All are measured at a monthly frequency and have been seasonally adjusted.
Because the various indicators are available for different time periods, and some are only available with a lag, we estimate the model on an unbalanced panel, a situation for which factor models are well suited. We begin our estimation sample in July 1976, when data are available for a majority of the indicators. The inputs to the factor model are detrended using a time-series filter.
To give some sense of which indicators are most influential for the LMCI, the last column of Table 1 reports the correlation of the 12-month change in each indicator with the 12-month change in the index. The unemployment rate and private payroll employment are most strongly related to the LMCI. The composite help-wanted index, the insured unemployment rate, the quit rate, and persons working part-time for economic reasons are also highly correlated with the index. In contrast, government employment, the participation rate, and average hourly earnings are relatively uncorrelated with the LMCI.
Changes in labor market conditions over the past 35 years
Chart 1 plots the average monthly change in the LMCI since 1977.4 Except for the final bar, which covers the first quarter of 2014, each of the bars represents the average over a six-month period. Changes in the LMCI align well with business cycles as defined by the National Bureau of Economic Research (NBER). The LMCI generally declines during recessions (the grey shaded areas) and typically rises during expansions.
Table 2 reports the cumulative and average monthly change in the LMCI during each of the NBER-defined contractions and expansions since 1980. Over that time period, the LMCI has fallen about an average of 20 points per month during a recession and risen about 4 points per month during an expansion. In terms of the average monthly changes, then, the labor market improvement seen in the current expansion has been roughly in line with its typical pace. That said, the cumulative increase in the index since July 2009 (290 index points) is still smaller in magnitude than the extraordinarily large decline during the Great Recession (over 350 points from January 2008 to June 2009).
Changes in labor market conditions since 2007
Chart 2 zooms in on the Great Recession and ongoing recovery, reporting the average monthly change for each quarter since 2007. The LMCI began falling in the second quarter of 2007 and deteriorated sharply in late 2008 and early 2009, as the financial crisis reached its height. The LMCI started improving in the second half of 2009, and the uneven pace of the ongoing labor market recovery is apparent at this time scale. Indeed, the LMCI captures several periods of sluggish improvement in the early parts of 2010, 2011, and 2012, which some dubbed the "spring swoons."5 The particularly marked slowdown in labor market improvement in the second quarter of 2012 stands out, as does the subsequent pick-up. Gains in the LMCI slowed a touch in the beginning of this year, likely reflecting unseasonably cold and snowy weather this winter, but the pace has picked up in recent months.
Over the recovery, as is typical, a few indicators account for the bulk of the increase in the LMCI. We can see this by decomposing the change in the LMCI into contributions from each indicator, holding the remaining indicators constant. Chart 3 shows the annual change in the LMCI and the contributions of 5 key indicators plus the sum of all other indicators. In the first two years of the recovery, the insured unemployment rate (the green portion of the bars) made a large contribution to the improvement in the LMCI, reflecting a substantial slowing in layoffs; this contribution has since diminished. Gains in private payroll employment (the blue portion of the bars) and declines in the unemployment rate (the pink portion of the bars) have been consistent contributors to the improvement, although more in some years than in others. So far in 2014, private employment and the unemployment rate have been the primary sources of improvement in the LMCI.
Closing remarks
Overall, the LMCI appears to be a useful tool for assessing the change in labor market conditions based on a broad array of labor market indicators. Of course, any purely statistical procedure may be sensitive to the many choices one must make in specifying the model, and will not take into account idiosyncratic events, at the one extreme, or changes in economic structures, on the other. A single model is, therefore, no substitute for judicious consideration of the various indicators. Nevertheless, such a model provides a summary that can usefully inform those deliberations.
Assessing the Change in Labor Market Conditions
Hess Chung, Bruce Fallick, Christopher Nekarda, David Ratner1
The U.S. labor market is large and multifaceted. Often-cited indicators, such as the unemployment rate or payroll employment, measure a particular dimension of labor market activity, and it is not uncommon for different indicators to send conflicting signals about labor market conditions. Accordingly, analysts typically look at many indicators when attempting to gauge labor market improvement. However, it is often difficult to know how to weigh signals from various indicators. Statistical models can be useful to such efforts because they provide a way to summarize information from several indicators. This Note describes a dynamic factor model of labor market indicators that we have developed recently, which we call the labor market conditions index (LMCI).
发帖
雷达卡
京公网安备 11010802022788号
