Federal Reserve Economic Data

The FRED® Blog

Don’t be surprised: Employment data get revised

BLS revisions to metro area employment data can be substantial but predictable

The Bureau of Labor Statistics (BLS) released the latest state and local employment data on March 13, 2017. The BLS initially estimated that St. Louis added 38,300 jobs in 2016, but the revised data show that St. Louis actually added 17,100 jobs. This isn’t the first time large revisions have occurred in St. Louis: Last year, BLS revisions to St. Louis employment added 6,800 jobs. Each March, these revisions account for data through the third quarter of the previous year. They can have a significant impact on the story of job growth in the region, so it’s important to be cautious about these preliminary estimates.

The top graph shows year-over-year growth in total nonfarm employment in the St. Louis MSA before and after the revision. Before the revision, St. Louis was estimated to have added 38,300 jobs and grown at 2.8% in 2016. St. Louis has not seen growth that strong since the 1990s. However, the BLS revised those numbers down to around 17,100 jobs added and growth of only 1.3%. While these revised numbers still show moderate growth, it is below the national rate of 1.6% and represents a significant decline from 2015’s 27,800 jobs added and 2.1% growth. Overall, the MSA has 16,600 fewer jobs than previously thought.

Data revisions occur because counting new jobs is a difficult process that relies on samples and advanced statistical techniques. As more information becomes available, data are revised. The BLS uses the monthly Current Employment Statistics (CES) survey to estimate local employment for nonagricultural industries, but the best source of local employment statistics comes from their Quarterly Census of Employment and Wages (QCEW). The QCEW includes data derived from establishments’ reports to the various unemployment insurance programs that are released with about a 6-month lag. Every March, the BLS reconciles the CES estimates with the data from the QCEW, which can result in significant revisions, as we’ve seen repeatedly here in St. Louis.

This year’s revisions underscore the importance of a cautious approach, despite the temptation to take unrevised data at face value: The BLS often revises employment data significantly, and the average absolute revision to 2016 metro area employment growth this year was 1.1 percentage points. The good news is that the revisions are generally predictable. The QCEW data had been growing at a much slower pace than the CES for much of the year for both nonfarm and leisure and hospitality employment. In addition, employment in St. Louis tends to follow the national cycle, so any large deviations in growth from the national rate should be corroborated with other sources of information. It can also be worthwhile to look at other sources such as the Fed’s Beige Book, to gain an understanding of the labor market beyond the latest BLS estimates. While the unrevised estimates were reporting strong growth in St. Louis, the Beige Book reported only modest or moderate growth for the region.

The BLS revised employment numbers in many of the MSAs in the Eighth District. In Memphis, the unrevised data reported employment gains of 2,700 and growth of 0.4% in 2016, but the revision brought those numbers up to 7,900 and 1.2%. In Louisville, jobs for 2016 increased from 4,400 to 5,400 and growth increased from 1.7% to 2.7%. In Little Rock, numbers were revised down only slightly, beginning in July 2015, with minimal effects on jobs added and growth in 2016.

How these graphs were created: The St. Louis Fed maintains records of all data revisions in its ALFRED® database, which allows you to retrieve vintage versions of data that were available on specific dates in history. On the “All Employees: Total Nonfarm in St. Louis, MO-IL (MSA)” page on FRED, click on “ALFRED Vintage Series” in the “Related Content” section underneath the chart to retrieve the two most recent releases, which currently include the revision. Under the “Edit Graph” button, click on “Format” and change the graph type from bar to line. Click on “Edit Lines” and select “Percent Change from Year Ago” for the units and copy to all. The three other graphs are built in a similar fashion.

Suggested by Charles Gascon and Paul Morris.

View on FRED, series used in this post: LOINA, LRSNA, MPHNA, STLNA

Regional income

In our previous blog post, we reported that inflation was higher in the New York City area than in the Cleveland area. Today, we look at their local incomes to see if the same conditions apply. The top graph shows personal income per capita for these areas. It’s not possible to get a perfect geographic match, but we use Cuyahoga County to represent the Cleveland consolidated metropolitan statistical area and a broader area around New York City (but without the western Connecticut towns used in the previous post). Without forgetting these mismatches, we see that overall income growth seems to be similar in both areas, except for the bubble in NYC just before the previous recession. So it doesn’t look like New Yorkers are compensated for their more-quickly-increasing living expenses.

The bottom graph reveals a similar exercise but with median household income, which is collected by county. Again, we use Cuyahoga County as a proxy for the Cleveland MSA. We use New York County to represent the NYC area.* The picture looks quite different, with Manhattan residents showing impressive income gains recently that probably aren’t matched in the wider New York metropolitan area.

*Admittedly, this is a restrictive choice; but piecing together data from the many surrounding counties is beyond the scope of this post. Readers are always encouraged to browse through FRED’s data aisles to find the best data to meet their needs.

How these graphs were created: Search for “personal income per capita Cleveland” and Cuyahoga County should be near the top. From the “Edit Graph” button, add a line searching for “personal income per capita New York.” For both lines, change units to “100 for selected date” using 1984-01-01 to match the data of the previous blog post. For the second graph, search for “median income Cleveland” then add the second line by searching for “median income New York county.”

Suggested by Christian Zimmermann.

View on FRED, series used in this post: MHINY36061A052NCEN, MHIOH39035A052NCEN, NEWY636PCPI, PCPI39035

Regional inflation

If everyone uses the same currency in the United States, shouldn’t prices and inflation rates be nearly uniform across the nation? To help answer this question, the U.S. Bureau of Labor Statistics computes a limited set of consumer price indices for consolidated metropolitan statistical areas (that is, the agglomeration of neighboring MSAs). They don’t provide as much detail as the nationwide CPI and they’re not necessarily available at monthly intervals, due to data sampling issues. But they can be revealing.

The graph above compares Cleveland and New York City: NYC prices seem to be climbing more than those in northern Ohio. Note that this doesn’t say anything about the levels, only the evolutions. But is this inflation differential uniform across goods? The graph below eliminates shelter from the mix and the gap between the two is noticeably smaller. In other words, differences in inflation for a refrigerator or a gallon of milk are much smaller across the country than differences in inflation for housing.

How these graphs were created: Search for “CPI CMSA Cleveland” and click on the monthly series. From the “Edit Graph” section / “Add Line” tab, search for “CPI CMSA NY” and select the monthly series. In the “Format Graph” section, change the mark type to “square” (there are marks for Cleveland because data are not available for every month). Finally, change the sample period to start on 1984-01-01. For the second graph, repeat the procedure by adding “shelter” to the search terms.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: CUURA101SA0L2, CUURA210SA0, CUURA210SA0L2, CUUSA101SA0


Back to Top