Featured Resources
Fiscal Brief
July 2019
Slacker City?
New York City’s Incredible Shrinking Workweek
PDF version available here.
Summary
From the perspective of job growth since the 2008-2009 recession, New York City’s economy has been booming. In 2018, the city had 715,200 more private-sector jobs than in 2008, a 22.0 percent increase. But the city’s economy looks less robust from the vantage point of growth in hours worked over that same period.
A new analysis from IBO finds that average weekly hours worked in the city has trended downward over the decade since the last recession—and that the decline has been steeper here than in other large and midsized metropolitan areas across the U.S. Among our key findings:
The decline in average hours worked in the city may partly explain why New York’s long expansion has been accompanied by relatively tepid wage growth. With the average number of hours worked decreasing, the local job market may not have tightened as much as the falling unemployment rate implies. Consequently, there would be less market pressure pushing up wages.
How Much Work?
One of the most basic questions we ask of an economy is how much work is it providing? The size of an economy and the income of its participants depend on the volume and productivity of labor. Typically, we measure volume by looking at employment—we count jobs. But a truer indicator of the volume of labor is total hours worked. This is a product of the number of jobs and average hours worked per job.
In New York City the number of jobs dropped during the 2008-2009 recession, but these losses were dwarfed by employment gains over the subsequent recovery and expansion. Average hours worked also slumped in the recession—but since then continued to bump downward, dropping the average workweek in the city’s private sector from 35.5 hours in 2008 to 34.1 hours in 2018. (Average hours have continued to fall through the first half of 2019.)
In this fiscal brief we the gauge the impact of the city’s shrinking workweek, put it in context (comparing hours worked trends across the United States), explore its underpinnings, and consider some of the consequences of our findings. We shall see that the shrinking workweek has cut deeply into the growth of the volume of labor in New York City, that the decline in average hours worked has been much more pronounced here than in most other metropolitan areas and regions in the U.S., and that the decline may have been a contributing factor—though certainly not the only factor—in the city’s weak wage growth during the current expansion.
A decrease of 1.5 hours worked per week does not look like a lot, but it has exerted a pronounced drag on the growth of the volume of labor in the city economy. While private employment stood 22.0 percent higher in 2018 than in 2008, average weekly hours were 4.1 percent lower, and total hours were up by a more modest 17.0 percent.
To put this in perspective, there were 715,200 more private-sector jobs in 2018 than there were in 2008, but had overall average hours not declined it would have taken an increase of only 551,100 jobs (17.0 percent) to yield the same gain in total hours. The decline in average hours worked thus offset the equivalent of 164,100 jobs, almost a quarter of the 2008-2018 employment gain.1
Down for Everyone, or Just for Us?
We tend to find longer workweeks in larger cities or metropolitan areas, and nine years ago New York City fell in line with expectations, with 35.5 average weekly hours worked in 2008 compared with 35.1 hours in other large and midsized metropolitan areas or divisions (those with employment of 500,000 or more in 2018) and 33.3 hours in the rest of the United States (small metro and non-metro areas). This put the city’s workweek well above the overall national average (34.3 hours).
Then the recession struck, and in 2009 average workweeks were down everywhere: down to 34.9 hours in New York City, 34.7 hours in the other large and midsized metropolitan areas, and 32.8 hours in the rest of the United States. New York City was in about the same position relative to the national average (33.8 hours) as a year earlier.
But by 2018 New York City was in a sharply different position compared with the rest of the country. Average weekly hours rebounded somewhat from their recession lows in other large and midsized metro areas (to 34.9 hours) and surged to new highs in the rest of the U.S. (34.1 hours), all of this bringing average hours worked back to just above the pre-recession level for the nation as a whole (34.5 hours).
Meanwhile New York City’s workweek continued to shrink, to the point that our 34.1 average weekly hours worked in 2018 put us well below the national average and indeed in a range now typical of a small metro/non-metro area.
Does our shrinking workweek tarnish New York City’s recent image as a post-recession pacesetter of U.S. economic growth? To some extent it does.
As a group, midsized and large U.S. metropolitan areas have sustained robust employment growth, tempered by moderate net declines in average hours worked over the past decade. But New York City has been at the extreme of both developments. Only one other large metropolitan area or division (Dallas-Ft. Worth) has enjoyed stronger private employment growth than New York City since 2008—and none has suffered a larger decline in average hours worked.
The result of all this is that New York City is closer to the middle of the pack of big metropolitan areas in terms of growth in aggregate hours worked. As can be seen in the table on page 3, Dallas-Ft. Worth (22.5 percent), New York City (22.0 percent), and San Francisco-Oakland (21.0 percent) were bunched close together and well ahead of the other big metros in terms of 2008-2018 employment growth. But in terms of growth in aggregate hours both Dallas-Ft. Worth (21.7 percent) and San Francisco-Oakland (20.7 percent) substantially outpaced New York City (17.0 percent). Houston and Atlanta lagged our city in employment growth but surpassed us in aggregate hours’ growth. Boston and Phoenix were far behind New York City in employment growth but only slightly behind in aggregate hours growth.
Why Us?
The decline in citywide average weekly hours could be due to a changing industry mix—an increasing share of employment in industries with typically shorter workweeks—or to shrinking workweeks within industries, or to a combination of the two.
The federal Bureau of Labor Statistics (BLS) does not provide average weekly hours worked by industry for local areas. But extrapolating from state-level data we can get a ballpark sense of average hours worked in major industry groups in New York City. (See the appendix for our methodology for estimating city-level industry hours worked.)
These estimates suggest that sectors with relatively long workweeks tended to grow more slowly over the past decade than sectors with relatively short workweeks. Manufacturing, construction, and financial activities had a combined average workweek of 38.6 hours in 2008 and employment growth ranging from -26.2 percent to 18.8 percent (1.6 percent combined) over the ensuing decade. Professional and business services, trade and transportation, and information had a combined average workweek of 36.7 hours in 2008 and employment growth ranging from 10.0 percent to 26.3 percent (18.6 percent combined). Education and health services, leisure and hospitality, and other services had a combined average workweek of 33.0 hours in 2008 and employment growth ranging from 20.1 percent to 48.7 percent (37.8 percent combined).
Had there been no change in average hours within industries, the concentration of city employment gains in sectors that had relatively short workweeks to begin with would by itself have shaved 0.3 hours from the citywide average workweek.
But average hours have changed in several broad industry groups, with notable increases in information and construction but steep declines in education and health services; trade, transportation, and utilities; professional and business services; and leisure and hospitality. On balance, these changes in industry group hours cut 1.2 hours from the citywide average workweek.
We have to keep in mind though that each major industry group is itself an evolving mix of many industries, sometimes with significant variations in hours worked among them. Education and health services deserves particular mention here, because much of the recent employment growth within this sector has been generated by home health care services, which has been adding jobs in New York City at a pace unmatched across the rest of the United States. Much of this growth, in turn, has been driven by the rapid expansion of the Consumer Directed Personal Assistance Program (CDPAP) in New York City.2 Average hours worked were relatively low in home health-related employment to begin with (under 30 hours a week according to national data) and probably lower still among CDPAP providers. And as home health employment has surged in the city, both CDPAP and (to a lesser degree) non-CDPAP average hours appear to have dropped.3 Thus, both an increase in the employment share of low-hours sectors and a decline in average hours within those sectors have been driving down the average workweek in education and health services.
Pressure Drop (On Wages)?
The declines in average hours worked in New York City might be part of an explanation of why the city’s long expansion has been accompanied for the most part by relatively tepid annual wage growth. Shorter workweeks by themselves translate into smaller annual wage gains for any given increases in hourly compensation. Moreover, with falling average hours offsetting a nearly a quarter of recent employment growth, the city’s labor market may not have actually tightened as much as the dramatic drop in the unemployment rate suggests. This implies weaker upward pressure on wages than “full employment” would otherwise impart.
New York City average weekly hours were roughly flat in 2018 but resumed their downward drift in 2019. Through June the average workweek dropped by another 0.2 hours compared with the same period the year before, and while year-over-year employment growth was 2.2 percent, aggregate hours increased by a less robust 1.8 percent. Thus the headline jobs numbers continue to give a misleading answer to the question we started out with: how much work is the city economy providing?
Prepared by David Belkin
Appendix
To impute a set of industry average hours for New York City that match up with the total private average and aggregate hours numbers provided by BLS, we use the following formula:
Hci = Hsi x {1 + [∑(Hsi x Eci) / (Hc x Ec) - 1] x [(Eci / Esi) / (Ec / s)]},
where H = average hours worked, E = employment, the superscripts s and c designate New York State and New York City, and the subscript i designates industry.
The term in the first set of square brackets sums the products of state industry average hours times city industry employment, ∑(Hsi x Eci), and divides this sum by total city aggregate hours (that is, by city average hours times total private city employment, Hc x Ec). This yields an adjustment factor that, if applied proportionally to state industry average hours, produces a set of city industry average hours that collectively generate a match to total hours worked in New York City as provided by BLS.
The term in the second set of square brackets weights the adjustment factor according to whether New York City accounts for a relatively large or relatively small component of statewide employment in an industry. The greater the city share of state employment in a given industry (Eci / Esi) relative to the city share of total private employment (Ec / Es), the smaller is the estimated difference between city and state average hours for that industry relative to the baseline difference obtained through the proportional adjustment factor (the first term in the equation) alone. These industry employment share weights have an aggregate impact of 1 on the baseline adjustment factor, so the final estimates of city industry hours worked again collectively generate a match to BLS-given total city aggregate hours:
∑(Hci x Eci) = Hc x Ec.
In 2018, New York City accounted for 49.6 percent of total private employment in New York State. The city’s share of industry employment was relatively high in information (76.1 percent), financial activities (67.6 percent), and professional and business services (58.2 percent); about average in education and health services (49.9 percent) and leisure and hospitality (49.5 percent), and other services (47.8 percent); and below average in trade, transportation and utilities (41.4 percent), construction (40.6 percent), and manufacturing (16.3 percent).
Endnotes
1Total annual hours worked in New York City private industry rose from 6,017.4 million in 2008 to 7,038.5 million in 2018, an increase of 1,021.1 million (17.0 percent). Dividing 1,021.1 million hours by the 2008 average of 35.5 weekly hours worked (and then by 52.143 weeks per year) yields “fixed hours equivalent” employment growth of 551,100. This is 164,100 less than the actual employment gain.
2This program enables eligible seniors and the disabled to choose their own home-care providers, including family members, who are ‘co-employed’ by entities called fiscal intermediaries that are reimbursed by Medicaid. For more detail, see Past as Prologue: Revised Histories and Extraordinary Trends in the New York City Economy (IBO, May 2019).
3Annual average wages for employees of fiscal intermediaries have fallen sharply over the past decade and are now almost a third lower than annual wages for others working in home health-related sectors. This is not due to declines in hourly compensation rates for home care providers, which are fixed (and rising) by law, and is thus circumstantial evidence that CDPAP providers work fewer hours than other home health care workers and that as CDPAP has expanded average hours worked has declined.
p class="basic-paragraph" style="font-family: 'Times New Roman', Times, serif; text-align: left;">PDF version available here.
Receive notification of free reports by e-mail