This morning, the Federal Housing Finance Agency (FHFA) released the latest data on house prices for states and metropolitan areas. The data for Arkansas show a rather sharp decline in in the first quarter of 2010: The purchase-only index was down 2.5 percent from a year earlier, while the all-transactions index was down 1.4 percent over the same period. (For a description of these indexes, see the previous post, Home Prices in 2009.)
Source: Federal Housing Finance Agency
In the first quarter of 2010, house prices were down rather sharply in four of the state’s metro areas. [For Metropolitan Statistical Areas (MSAs), only the all-transactions indexes are available.] Prices were down 4.8 percent in Fort Smith, 2.7 percent in Northwest Arkansas, 1.5 percent in Jonesboro and 1.1 percent in the Little Rock area. Hot Springs, Pine Bluff and Texarkana all recorded small increases for the quarter.
Source: Federal Housing Finance Agency
Compared to a year earlier, house prices are lower in every Arkansas MSA except Pine Bluff. Over the longer run, however, prices in each of the states MSAs are higher than five years ago (2005:Q1).
The latest information on Arkansas employment and unemployment was released by the U.S. Bureau of Labor Statistics and the Arkansas Department of Workforce Services this morning. The unemployment rate held steady in April at 7.8 percent, while the number of employed declined slightly from March.
Source: Bureau of Labor Statistics
Although the unemployment rate itself was unchanged, the household survey showed that the number of unemployed people declined from 197.1 thousand to 106.3 thousand. This was the first monthly drop in the number of unemployed since March of 2008. This drop failed to have an impact on the unemployment rate, however, since the number of employed reported in the household survey also declined.
The payroll survey also indicated a slight decline in the number of jobs. Total nonfarm payrolls were down by approximately 1600 jobs in April (seasonally adjusted).* The April decline follows a sharp increase we saw in March (up 8900 jobs, revised). In terms of longer-term performance, therefore, the April decline does not appear to signal any change in the trend of employment growth. For the first four months of the year, nonfarm payrolls have increased by 3600. Employment is stabilizing, but there is no clear indication (yet) of a resumption of long-term job growth.
Source: Bureau of Labor Statistics
The decline in nonfarm payroll employment in April was largely attributable to service-providing sectors. Goods-producing sectors had been very hard-hit during the recession so recent increases in manufacturing employment, in particular, have represented welcome news.
Source: Bureau of Labor Statistics
# # #
*Seasonally adjusted data for nonfarm payroll employment, reported in a format compatible with the monthly press release from the Arkansas Department of Workforce Services, are available here: Table – Seasonally Adjusted NFPE.
“Sales rose by 2.4 percent in the first quarter of 2010 following a 1.0 percent increase in the fourth quarter of 2009…”
As described in a previous article, data on state sales tax collections can be used to measure overall sales activity in the state, producing a data series we called Arkansas Taxable Sales. The Institute for Economic Advancement has developed this measure to help fill an existing gap in the availability of timely, accurate data on the retail sales sector of the Arkansas economy.
The latest report on state revenues from the Department of Finance and Administration showed that April gross receipts collections (primarily sales taxes) were up from the previous year, and were 0.7 percent above DFA’s forecast. With this information, preliminary* data on Arkansas Taxable Sales for the first quarter of 2010 can be estimated.
After adjustments for the timing of tax collections relative to underlying sales, changes in tax laws, and seasonal patterns in the data, the Arkansas Taxable Sales series serves as a proxy measure for recent conditions in the retail sales sector.
The new reading on this statistic shows that sales rose by 2.4 percent in the first quarter of 2010 following a 1.0% increase in the fourth quarter of 2009, lifting sales to a level 3.4 percent higher than in the trough that is now evident in third quarter of 2009.
Arkansas Taxable Sales statistics are constructed by the Institute for Economic Advancement using tax collection data from the Arkansas Department of Finance and Administration.
The first quarter data indicate a pattern of recovery in Arkansas retail sales. With two consecutive quarterly increases, this is a clear indication that positive growth has taken hold in the Arkansas retail sales sector. Mirroring similar gains that we’ve seen in statistics for retail sales nationwide, this information confirms recent trends suggesting that Arkansas is firmly back in the expansion phase of the business cycle.
* Data are preliminary until the release of the DFA report, Arkansas Fiscal Notes, for April 2010. Data will be updated when information becomes available.
Metro area unemployment rates for March 2010 were released last week by the Bureau of Labor Statistics and the Arkansas Department of Workforce Services. Over the first three months of 2010, seasonally-adjusted unemployment rates rose in all of Arkansas’ Metropolitan Statistical Areas (MSAs). The Pine Bluff MSA continued to register the highest unemployment rate in the state: 9.9% in March–slightly above the national average of 9.7%. All of the state’s other metro areas were below the national average, with the lowest rate being 6.6% in the Fayetteville-Springdale-Rogers MSA.
Sources: Bureau of Labor Statistics, Institute for Economic Advancement
Nonfarm payroll employment was up sharply in the statewide data for March (see previous post). For the state’s metro areas, employment was unchanged or up slightly in six of the seven MSAs. Only Texarkana showed a decline for the month. However, Texarkana had shown increases earlier in the year and it is the only MSA in Arkansas with positive employment growth, on net, over the most recent three months (from December).
Source: Bureau of Labor Statistics
Compared to March of last year, employment in March 2010 was down in all of Arkansas metro areas. In percentage terms, Jonesboro showed the smallest decline: down by only -0.6 percent. Having shown a sharp decline in the first three months of 2010, employment in Hot Springs is down by 3.2 percent from a year earlier.
The March figures for Arkansas employment and unemployment came out on Friday. The news was mixed: The household survey showed higher unemployment while the payroll survey indicated strong employment growth, particularly in the goods-producing sectors.
The unemployment rate ticked up again, from 7.7 percent in February to 7.8 percent in March. The household survey indicated an increase in the number of unemployed persons by nearly 900, and a decline in the number of employed by approximately 7,800. Hence the unemployment rate reflected both an increase in unemployment and a decline in the labor force.
In contrast, the payroll survey showed a sharp increase in employment, up by nearly ten-thousand jobs (seasonally adjusted).
Source: Bureau of Labor Statistics
Goods-producing sectors showed notable gains: Construction employment was up 3,100 and manufacturing employment rose by 1,600. Month-over-month increases were also reported in information services, professional and business services, other services, and government.
Source: Bureau of Labor Statistics
In addition to the strong payroll growth in March, the figures for February were revised upward by 1,000 jobs — it turns out that February was slightly less “dismal” than previously reported. Over the past six months (from September 2009 to March 2010) employment has increased in construction, manufacturing, and several service-providing sectors.
It is not unusual for the household data and the payroll data to give conflicting signals, particularly on a month-to-month basis. In addition to having different sources, the two surveys differ in the way that they measure agricultural employment, self-employment, and multiple job-holders (among other things).
The payroll figures show signs of recovery in employment growth for some sectors. Nevertheless, the increases are not (yet) sufficient to reduce a stubbornly-high unemployment rate.
The February report on state-level employment and unemployment was released by the U.S. Bureau of Labor Statistics and the Arkansas Department of Workforce Services this morning. Set against the backdrop of expectations for a stabilizing employment situation, the Arkansas report is disappointing at best.
Household survey
The headline news was an increase in the unemployment rate from 7.6 percent in January to 7.7 in February. This is not a substantial change, but the upward creep is unwelcome news. Changes in the underlying components of the unemployment rate highlight the weakness: The number of employed was down by 1,142 and the number of unemployed was up by 1,285.
Source: Bureau of Labor Statistics
Payroll Survey
Not seasonally adjusted, nonfarm payroll employment in Arkansas fell only slightly in February (down by 300 jobs from January). However, as shown in the chart below, January is usually a low point in the seasonal cycle of employment, with February tending to show a rebound as we head for the spring months. After seasonal adjustment, the decline in February was substantial: down 7,300 jobs. In percentage terms, this represents a decline of 0.6%–the third largest drop among the 50 states in February. This is a larger one-month loss than any we observed for Arkansas during the depths of the 2008-09 recession.
Source: Bureau of Labor Statistics
The table below summarizes the change in employment by sector. Employment was down in nearly every sector in February. The largest declines were in Trade, Transportation and Utilities; Professional and Business Services; and Other Services. Construction employment continues to fall. In February, construction employment fell below 50,000 (seasonally adjusted) for the first time since January 1999.
Source: Bureau of Labor Statistics
There were two relatively positive aspects of the report. First, the data from January were revised to show a larger employment increase than previously reported. As a result, some service-sector categories have shown positive growth, on net, over the first two months of the year. The second positive element of the February report was manufacturing employment, which was up by 1,500 jobs. Manufacturing employment has been experiencing a protracted period of decline. The February increase provides further evidence that the sector is stabilizing.
One should never put too much emphasis on a single month’s observations—and the numbers released this morning are subject to future revision. Nevertheless, taken at face value, today’s data-release represents a dismal report.
This morning, the Bureau of Economic analysis released their report on State Personal Income for the fourth quarter of 2009. Because the fourth quarter figures complete the data for 2009, the BEA press release focused on annual income in 2009 compared to 2008. For the year, 44 states experienced declines in personal income, with an average of -1.7%. In Arkansas, personal income was down by only 0.2%, placing our state in the top ten in terms of income growth for the year.
Per capita income averaged $31,946 for Arkansas in 2009, down by 1.0% (compared to a 2.6% decline for the U.S. average). In terms of per capita income growth for the year, Arkansas ranked 12th in the nation.
The quarterly pattern of income growth over the year is also of interest. Following two quarters of modest declines, Arkansas personal income was up by 1.1% in the fourth quarter. This was the 9th largest increase in the country.
The chart below shows the pattern of personal income for Arkansas compared to the U.S. Both series peaked in the second quarter of 2008, so the data are normalized to equal 1.0 in that quarter.
Sourse: Bureau of Economic Analysis
For the U.S., personal income fell by to 2.7 percent of its 2008Q2 value by the first quarter of 2009. By the end of the year, U.S. income was down 1.4% from its peak. The decline in Arkansas was more protracted and gradual. Arkansas personal income hit a low point in the third quarter of 2009 (1.7% below peak), but the strong performance in the fourth quarter raised incme to a level only 0.6% below the 2008Q2 peak.
The latest information on employment and unemployment for Arkansas Metropolitan Statistical Areas (MSAs) was released last week. The data for January were mixed. There were no clear signs of recovery in regional labor markets, but no further deterioration was evident either. This is consistent with the statewide data reported in a previous post–showing the state’s unemployment rate unchanged with only a small increase in payroll employment.
Household Survey
As displayed in the table below, the household survey showed that changes in unemployment rates around the state were mixed. The unemployment rate ticked up in the Fayetteville, Fort Smith, Hot Springs and Texarkana. It was unchanged in Little Rock and declined in Jonesboro and Pine Bluff.
Sources: Bureau of Labor Statistics, Institute for Economic Advancement
The drop in Pine Bluff’s unemployment rate was particularly sharp. In recent years, Pine Bluff has experienced large swings in employment and unemployment from December to January. The changes were smaller this year, resulting in a measured decline in the seasonally-adjusted unemployment rate. We’ll have to wait and see whether this is just a statistical fluke or a significant change.
Payroll Survey
The January payroll survey also showed mixed changes in employment. Seasonally adjusted, January payroll employment was up slightly from the previous month in Texarkana, unchanged in Jonesboro, and down a bit in the state’s other MSAs.
Source: Bureau of Labor Statistics (Note: Although we have seen no formal annoucement of the change, the BLS public database now provides seasonally adjusted payroll employment data for Arkansas MSAs. Assuming that this practice continues, the Arkansas Economist will henceforth rely on these official seasonally adjusted series, rather than using data adjusted by the Institute for Economic Advancement.)
Benchmark Revisions
This data release also included the annual benchmark revisions that have been discussed extensively in recent posts on the Arkansas Economist (MSA revisions were discussed here). As expected, the revised data show sharply greater job losses during the recession than were previously reported — for most of the state’s MSAs. The revised data show that Jonesboro was the only metro area to show positive growth in 2008, and all of the states MSAs suffered employment losses during 2009. The decline in Fort Smith was particularly steep.
Source: Bureau of Labor Statistics
Although the revisions showed larger employment declines for most of Arkansas’ MSAs, the revised data for Pine Bluff show that job losses were not as large as previously reported. In this case, it’s no fluke. The updated statistics are based on very accurate data from the Quarterly Census of Wages and Employment (QCEW) for 2008 and early 2009.
The anomolous upward revision to the payroll data for Pine Bluff was anticipated in a previous post on the Arkansas Economist. A more complete evaluation of our exercise in “forecasting a revision of history” is forthcoming.
A new indicator of economic activity for the Arkansas economy
Retail sales statistics for the U.S. economy are regularly compiled by the Census Bureau. However, there is no readily-available, timely source of data for retail sales on the state level. One approach to filling this gap is to use state sales tax collections as a measure of taxable sales. In principle, the approach is quite simple: Taxable sales = Sales tax collections/tax rate. In order to extrapolate sales statistics from tax data, however, it is necessary to consider the coverage of the tax, changes in tax law, and the timing of state revenue collection relative to the underlying sales.
Arkansas has four separate sales taxes imposed on a common base: a general sales and use tax of 4.5 percent, a tax for the educational adequacy fund of 0.875 percent, a property tax relief tax of 0.5 percent, and a conservation tax of 0.125 percent (for a total tax rate of 6.0 percent). Each of these taxes could conceivably be used to identify overall sales activity. However, recent changes in tax structure complicate matters. In particular, the tax rate on grocery purchases was lowered from 6.0 percent to 3.0 percent in July 2007, then lowered again to 2.0 percent in July 2009. When these tax cuts were adopted by the state legislature, the associated revenue reductions were specified to be subtracted from three of the four specific taxes, in proportion to their relative rates. As a result, these three taxes (general sales and use, educational adequacy, and property tax relief) no longer represent a constant fraction of total sales–rather, they represent a weighted average of taxes on non-grocery items (at the statutory rate) and taxes on groceries (at a lower rate). Only the 1/8-percent conservation tax remains at its statutory rate. (The conservation tax is not affected by recent tax law changes because it is written into the state constitution.) The conservation tax is therefore a unique candidate for use in estimating the underlying tax base.
A second important factor for deriving sales information from tax collection data is the timing of tax receipts relative to the sales being taxed. Tax laws in Arkansas impose different requirements on different sized businesses. Some larger firms can choose to pay estimated taxes for the current month, while some smaller firms are required to file and pay only quarterly or annually. For most business, however, the key deadline comes on the 24th day of the month. On that date, final statements and payments for taxes owed in the previous month are due.
The tax payment deadlines suggest that revenues received by the state during a month primarily reflect taxable sales from the previous month. An examination of the raw data supports this timing convention. For example, when the special conservation tax was implemented in 2001—and again when the earmarked educational adequacy tax was introduced in 2004—revenues received in the first month of implementation were a very small proportion of subsequent months. The seasonal pattern of sales tax receipts is also indicative: The most prominent recurrent peak in tax receipts occurs in January, one month after the December retail sales peak. As the payment regulations indicate, tax receipts reflect a mix of taxable sales from current and previous months. Nevertheless, the best practical estimate seems to be a simple one-month lag.
So, using receipts from the conservation tax and adjusting the timing to reflect the one-month collection lag, we can calculate a measure of taxable sales. Monthly data for this series are illustrated in the figure below:
Although there is a great deal of month-to-month variability in this measure, recent observations suggest that sales activity reached a low point around mid-2009, with signs of recovery since then.
In interpreting the data, a number of caveats should be considered:
The series is labeled “taxable sales” rather than “retail sales” because Arkansas sales and use taxes apply not only to retail transactions, but also to some business-to-business transactions.
Some of the monthly variability in the series can be attributed to specific institutional factors in the tax collection process. For example, the rather sharp upturn observed for December 2009 (reflecting tax collections in January 2010) was partly attributable to “one-time gains from audit payments” [see General Revenue Report for January (FY 2010)].
Another previous change in the tax structure should also be mentioned: in 2004, with the introduction of the educational adequacy tax, the sales tax base was broadened. Although there is no evident break in the path of total tax revenues, this consideration warns against making direct comparisons between post-2004 and pre-2004 data. (The expansion of the tax base accounts for a revenue increase of only about 0.5 percent.)
The most recent observations in the series should be considered preliminary. Data for the conservation tax are not explicitly reported in the General Revenue Reports, but are available in the publication Arkansas Fiscal Notes, which is available a few weeks after the initial report. Until the publication of Arkansas Fiscal Notes, the change in general gross receipts is used as a preliminary approximation.
Given some of these considerations, it is best not to focus on very short-term changes in the data. The chart below shows the results of averaging the data over calendar quarters. The chart juxtaposes Arkansas Taxable Sales with a quarterly measure of U.S. retail sales.
Sources: U.S. Census Bureau, Arkansas Department of Finance and Administration, Institute for Economic Advancement.
A comparison of the measures for Arkansas and the U.S. reveals an interesting pattern: Sales in Arkansas reached a peak about one quarter later than the U.S., and experienced sharp declines with a similar lag. Moreover, the percent change from peak to trough is smaller for the Arkansas data than for the U.S. This pattern is consistent with the notion that Arkansas’ economy was dragged into recession by economic weakness in the rest of the country. Both measure show increase toward the end of 2009, confirming the general observation that both the national and Arkansas economies are beginning to recover from the recession of 2008-09.
The latest state-level data on employment and unemployment were released today, covering January 2010. We’re getting this monthly information a little bit later than usual because the Bureau of Labor Statistics took some additional time to go through the annual process of revising previously-published data using more complete information.
The unemployment rate in Arkansas held steady at 7.6 percent in January, remaining well below the U.S. average of 9.7 percent. Today’s data release included a new smoothing technique for seasonally-adjusted unemployment data that eliminates some of the month-to-month variability. By this new measure, the unemployment rate in Arkansas has been holding in the range of 7.5 to 7.6 percent for the past 6 months. There is no indication that the unemployment rate is poised to go any higher, but neither is there any indication (yet) of an imminent decline in unemployment.
Source: Bureau of Labor Statistics
The payroll employment figures showed some signs of recovery in service-sector employment. Total nonfarm payrolls increased by 1,700 jobs in January (seasonally adjusted) with increases in the categories of Financial Services, Professional and Business Services, Education and Health Services, Leisure and Hospitality, and Other Services. We continue to see weakness in the goods-producing sectors. Manufacturing employment had been up slightly toward the end of 2009, but was down by 1,900 jobs in January.
Source: Bureau of Labor Statistics
The payroll employment figures that were released this morning incorporated new benchmark revisions. As reported on previous pages of the Arkansas Economist (here, here and here), the annual exercise of reconciling data from the Current Employment Statistics (CES) program with the Quarterly Census of Wages and Employment (QCEW) was expected to show sharply larger job-losses during 2009 than were previously recorded in the data. This did, in fact, prove to be the case. As shown in the figure below, the revised data show that payroll employment declined more sharply during 2008 and 2009 than was reported at the time, with previously unrecorded job losses accumulating to 18,200 by the end of the year (as of December).
Source: Bureau of Labor Statistics
The fact that job losses during 2009 were greater than previously measured is–to some extent–old news. However, it means that it will take longer to see employment increase to the levels we saw prior to the onset of recession.
