Inflexible wages and prices? evidence in the current recession.
Keywords: small business, wage, price flexibility, sticky wages and prices
There are two basic schools of economic thought when it comes to the debate over economic policy. One argues that markets provide the best adjustment to "shocks" and that business cycles are part of the process of markets "healing themselves," correcting prices and wages (disequilibria in factor and product markets) and disciplining institutions that manage resources poorly. With flexible wages and prices, there is little need for government to intervene to restore the economy to "equilibrium" (assuming that it wasn't a policy shock that caused the economic disturbance initially).
Another school (Keynesian in its foundations) argues that markets are unable to do this job well due to inflexibilities in wages and prices, which do not send the needed signals to reallocate resources. The "stickier" wages and prices are, the longer it takes for the market to clean up a mess, raising the specter of prolonged economic slack as in the Great Depression. "As wage and price rigidity is the basis for Keynesian theory and policy recommendations, understanding the potential causes of rigidity is important" [Abel et al 2008, p. 398].
The fundamental proposition is that the more flexible wages and prices are, the faster the economy adjusts to a shock and restores full employment. To the extent that government prevents or impedes these adjustments (for example, saving firms from bankruptcy, preventing needed adjustments in compensation, or establishing regulations that prevent adjustments to competitive pressures--such as import restrictions or imposing a minimum wage), the economy is compelled to spend more time in recession or experiencing slower growth while trying to make the needed adjustments. A number of developments have enhanced the potential for more flexibility in wages and prices. Technology has reduced the importance of "menu costs," reducing the cost of actually producing a price list (catalogue or "menus"), and has made it much easier to change prices--especially those posted in catalogues available on the internet. In the private sector, the importance of union contracts has diminished, increasing potential wage flexibility (but not in the public sector where union membership is ubiquitous). Labor legislation (including minimum wage laws) has probably reduced wage flexibility for small businesses. For small businesses, that produce half of private GDP and employ well over half of the private sector workforce, unionization (1) is less of an issue, and relatively few have expensive "price lists" to deal with.
It is "sticky wages and prices" that are alleged to be the transmitters of financial market disruptions to the real economy. If business owners are unable to adjust nominal prices and wages in ways that insulate real output and employment from financial market turbulence, a real business cycle results. Rigidity is a "core belief in the Keynesian explanation of recessions since it prevents real wages and prices from adjusting fast enough to equate supply and demand.
Classical economists argue that a major cause of unemployment is a mismatch between labor supply and demand. For example, many auto workers are available for work but the need for more workers is not in Michigan, and no matter how low wages go, these displaced workers cannot find jobs there. Since there is also hope that their jobs might be saved by government and their unions, workers are not inclined to move to other labor markets. This kind of mismatch takes time to resolve and is exacerbated by government involvement providing (false?) hope that jobs will be restored. These rigidities in labor mobility (exacerbated by the mortgage crisis which made changing houses more difficult in this recession) slow the adjustment process as does the government "rescue" of firms that should have failed (locking in labor and compensation at levels that contributed to the firms' failures).
Without arguing the merits of these competing views, this paper presents some evidence on the speed with which wages and prices are adjusting in the latest period of serious economic weakness compared to the 1980-82 recessionary period. The data collected by the National Federation of Independent Business (NFIB) (2) for the past 35 years suggest that small business owners may be adjusting wages and prices and inventory stocks more quickly than they have in the past. However, the nature of the distress is different in every recession, so the rapidity of the response might be due more to the nature of the recession than to a fundamental change in the structure of the small business community.
The next three sections of this paper provide evidence from the periodic surveys conducted by NFIB that indicates that small business owners have adjusted prices, compensation and inventories and capital spending much more quickly in response to economic stress in this recession than in past episodes of weakness. The last section speculates as to why this might have occurred using the 1980-82 recession period for comparison and explores the implications of more flexibility in prices and wages for appropriate government policy.
1. Adjustments in Labor Costs
There are two ways to adjust labor costs: (1) reduce employment (hours) and (2) reduce compensation for those employed by the firm. Table 1 shows the adjustments made in private-sector employment for each recession since 1973 from peak to trough (based on quarters, not months). In simple terms, the adjustment in private-sector employment in the current recession period (especially on a quarterly basis) has been substantially larger than in any other recession (and the bottom for employment, as of December 2009, has yet to be reached). The average reduction in private employment per quarter during the recession period has been much larger than in any past recession since 1973.
Table 1. Private Sector Employment: Peak to Trough (Qtr) Peak Employment Trough Employment Decline Qts Avg (%) (%) 1973:2 64,294 1975:2 61,733 4.0 8 0.5 1980:1 74,649 1980:3 73,657 1.3 2 0.6 1981:3 75,435.7 1982:4 72,835.3 2.4 4 0.6 1990:2 91,245.3 1992:1 89,587.3 1.8 7 0.25 2001:1 111,670.3 2003:2 108,253 3.0 9 0.33 2007:4 115,670.3 2009.3 108,765.7 6.0 8 0.75
Figure 1 shows the reported average change in employment per NFIB firm since 1976:4, when the data were first collected. Owners were asked if they had increased or decreased the total number of people working for them and if so, the number of workers affected. The average change is for all firms, including those that did not change employment. It is clear that the reductions in employment in the fourth quarter of 2008 and in 2009 were the largest in the 32-year history of this data series. The adjustment in labor costs (head count) was rapid and deep.
Figure 2 shows the incidence of compensation cuts (the seasonally adjusted net (3) percent of firms reporting higher employee compensation), clearly sensitive to the business cycle. When the economy weakens, the incidence of compensation reductions rises. Unfortunately, no compensation data are available for the 1980-82 period. Interestingly, the incidence of compensation reductions is much higher in the milder 2001 recession period than in the 1990-91 period (more flexibility?). Figure 3 shows that the incidence of reported reductions soared in the first quarter of 2009, rising to levels more than double any in the history of the NFIB survey. This suggests that there might be more downward flexibility in nominal compensation today than in the past. (4) The percent of owners reporting increases in worker compensation also dropped to record low levels. This process may have been made easier by the behavior of inflation in this period, which favorably impacted real wages. Failing to raise compensation in an inflationary environment would represent a reduction in real compensation. However, the CPI inflation rate was negligible in 2008, when the incidence of compensation increases reached a survey low, so the real reductions in compensation were muted by falling prices.
The NFIB data suggest that labor costs (employment and compensation) have adjusted to the economic "shock" of the subprime crisis much more quickly than in response to past shocks such as the 2000-01 stock market crash. This would generally suggest less need for government intervention to aid the economy, as prices and wages are adjusting more quickly than in past recessions. Of relevance, of course, is the degree to which the adjustment falls on "wages" rather than employment (for example, everyone takes a 10 percent "cut" rather than a 10 percent reduction in employment).
2. Adjustments in Prices
Figure 4 shows the net percent of firms raising average selling prices quarterly since 1973. From a cyclical peak of 61 percent raising prices (net of those cutting prices) in the fourth quarter of 1979, the net percent of owners raising prices fell to negative 1 percent in January of 1983, a decline of 60 points. But inflation was as high as 16 percent in the first quarter of 1980 and fell to 0.3 percent in the first quarter of 1983. In July of 2008, a net 32 percent reported raising average selling prices, and CPI inflation peaked at 6.6 percent in the third quarter. The inflation rate was -8 percent in the fourth quarter. Through April 2009, the net percent of owners raising prices declined 56 points to a negative 24 percent of all owners. By October, there was a small improvement to negative 17 percent. All this in only four quarters, compared to the 13 quarters it took to accomplish a reduction in price raising activity of about the same magnitude in the early 1980s.
Figure 5 illustrates the time path of the net percent of owners raising average selling prices and the net percent raising worker compensation. Historically, firm profitability has been highest (as a share of GDP) during periods in which the percent of owners raising compensation has been reasonably matched by the percent of owners passing these costs on through higher prices.
Blinder  sampled 200 firms and concluded that most firms change prices infrequently. This would logically seem to depend on such factors as current economic conditions, the current rate of inflation, and the industry in which the firms operates: a barber may rarely change his price, but gas stations report frequent price changes. In this cycle, prices and compensation have been cut with unprecedented frequency. One interpretation of this is that we are getting the job done more quickly, shortening the time until a "bottom" is reached.
3. Inventories and Capital Spending: Part of the Adjustment Process
The objective of "price cutting" is to bring inventory stocks into balance with sales (or expected sales) as well as to raise cash and reduce the cost of financing inventories. Inventory reduction occurs in every recession, but Figure 6 shows that the frequency of the reductions reached record rates in the current recession. The pervasiveness of the reductions should, once again, adjust the existing stock of inventories to match desired stocks at a much faster pace than in past recessions. Unfortunately, data on inventory change are not available for the 1973/4 recession period or for the peak quarters before fourth quarter of 1980, so a comparison to these recession periods is difficult. But for the period for which data are available, it is clear that the most pervasive periods of inventory reductions occurred in the fourth and first quarters of the most recent recession (reported in NFIB's January and April 2009 surveys).
Capital spending showed a similar pattern, the percentage of firms making any capital outlays declining far more sharply than in the 1980-82 period, as shown in Figure 7. Comparisons are complicated by the addition of leasing to the question in 1994. Parallel surveys indicated that adding leasing to the question added on average six points to the percent of firms reporting outlays. No data are available before 1979. The percent of owners reporting outlays was 64 percent in the first two quarters of 1979 and fell to 47 percent in the third quarter of 1982, a decline of 17 percentage points. In the most recent expansion, the peak of 66 percent (60 percent after subtracting the leasing adjustment of 6 percentage points) reporting outlays was reached in the first quarter of 2004. The low point (to date) was 45 percent (39 percent adjusted) in October, 2009, a decline of 21 points to the lowest levels of those reporting capital outlays in the history of the survey.
4. Structural Change or Business as Usual?
The NFIB data clearly demonstrate that the incidence of adjustments to employment, prices and inventories was substantially more widespread in the current recession than during the 1980-82 period (which included two NBER recession periods). Data on the magnitudes of the changes are not available for most measures, only the direction of change. Over 20 years, could the structure of the small business community have changed so much as to explain the highly attenuated responses of owners? Certainly the computer and the internet have revolutionized large firms' management information systems, but not so much for small firms, 90 percent of which have fewer than 20 employees and have not changed their physical structure substantially over the period. The metrics of the two recession periods may provide some insights. The 2007-09 recession was longer (25 months) than that of 1980-82 (22 months), and the economy did have a "relief rally" in between the two NBER 1980 = 1982 recession periods (which were six and 16 months respectively), a pause not enjoyed in the 2008-09 recession.
Leading into the 1980-82 recession period, the NFIB Optimism Index (5) peaked in 1977:2 and declined fairly steadily to 90.2 in January, 1980, a clear signal that the economy was slowing and entering a period of negative growth. (6) The Index declined 10.7 points over 11 quarters (a decline of about one point per quarter). Over the next three months, the Index fell to its record low level of 80.1, a 10.1 point decline in one quarter, a clear signal of a severe decline in economic activity to follow. This squares with the NBER dating of the peak of the expansion as January 1980.
In the current cycle, the Index peaked in 2005:4 at 103.7 (the peak is defined at the highest value reached before a clear downward trend in the Index was established, based on quarterly survey data). The Index usually attains its highest value early in each expansion. By 2007:4, the Index had declined to 96.2, a decline of 7.5 points, about one point per quarter from the peak. In January 2008, the Index fell to 91.8, a loss of an additional 4.4 points, a clear sign of a sharp break in economic activity.
The behavior of the Optimism Index in each recession period is quite similar, but actual cuts in employment, prices, and inventories were strikingly different in each period, in both frequency and speed. These three variables are not "forward looking" and are not included in the Index. There are some possible explanations for the apparent differences in the responses of small firm owners in the two periods:
1. Inventories were larger relative to sales in 2007, triggering a much broader reduction in stocks. The need to reduce stocks produced the more rapid change in prices.
2. Employment was too high relative to output (perhaps based on overly optimistic expectations); and when output fell, cuts in employment were larger than usual.
3. Owners were more uncertain about future economic performance than in other recession periods, due to the unusual financial market crisis and the change in political leadership.
Table 2 provides a "high level" picture of the economies in the two periods. Perhaps the most important differences between the two recession periods are those in real consumer spending growth and the level of inflation and interest rates. Inflation was substantial in the 1980-82 period, increasing the value of spending in current periods, especially for durable goods, to avoid paying even higher prices later. If anything, price behavior (and psychology?) in the current recession favored saving. This unexpected saving created an excess of actual inventory over that which was desired based on previous and expected sales. Desired stocks of inventory were premised on years that were characterized by a very low consumer saving rate and an unprecedented expansion of debt. The sudden shift to positive savings in 2008:Q4 produced a surplus in inventories that had to be liquidated for survival. The "credit crunch" did not appear to be a major concern to small businesses in this recession, as Figure 8 indicates that there was no spike in concerns during the "crunch" period. However, the 1980-82 recession period was characterized by large percentages of owners citing financing and interest rates as their top business problem. (7) Clearly, the weakness in spending was the top concern in this recession, which motivated inventory liquidation as well as price cutting to attract customers and reduce inventory stocks.
Table 2. Recession Statistics Change Peak to Trough (NBER dates) (1) 1980-82 (2) 2007-09 Consumption + 3.9% -1.1% Residential investment -30.4% -31.4% Consumer Sentiment + 4.1 points -9.1 points Private employment -2.4% -6.0% Total employment -2.3% -4.9% Unemployment rate +4.4 points +4.7 points 10-year treasury -1.3 points -0.8 points CPI inflation rate -14.6 points -2 points NFIB optimism + 5.8 points -8 points Change in consumer credit + 13.5% -3% S&P 500 + 23.9% -33.3% "Financing" the top problem 37% (4) 5% (3) Credit harder to get (5) 28% (4) 14% (1) Manufacturing output bottomed in the third quarter, but employment probably did not bottom until the end of the fourth quarter. For this analysis, the third quarter is treated as the "bottom" of the NBER recession since fourth quarter NIP A data were not available when the article was written. (2) NBER identifies two recessions in this period of time, 1980:1 (January peak) to 1980:3 (July) and 1981:3 (July peak) to 1982:4 (November). This is typically treated as one 16-month recessionary period. The current recession started 2008:1 (the peak was identified as December 2007) and probably ended in 2009:4, although as of January 2010 the NBER has yet to make a formal declaration. (3) The question asked: "What is the most important business problem facing your business today?" Respondents have 10 choices and a "fill in the blank". (4) Highest value in the period, percent of all owners. (5) Asked of owners who report that they borrow at least once a quarter: "Are these loans easier or harder to get than they were three months ago?"
As Figure 9 shows, inventories were not excessive relative to sales going into this recession compared to previous recession periods. At this "macro" level, it does not appear that an inventory overhang in physical terms was responsible for the record incidence of price cuts or inventory reductions. But, relative to the desired stock of inventories, stocks quickly became excessive and survival demanded a quick liquidation. (8)
The "stock" of employees was too large, but not unusually so (see discussion above). But, with profits severely squeezed by the sudden turn in spending and the price cutting needed to reduce inventories, firms had no choice but to rapidly reduce costs. Employment, about 80 percent of operating costs, had to be rapidly reduced as profits plunged along with inventories.
If employment was too high (exuberance from the housing boom, and so on), the need to reduce employment dramatically would produce the record declines in employment per firm observed in this recession. To test this, a model predicting expected private employment was produced by regressing private sector employment on private real GDP. The residuals from this equation provide a measure of "excess employment", employment not expected based on the level of real private GDP. The model is estimated from 1974 to 1993 and then from 1993 to 2008 to account for structural change before and after "Y2K" and the "dot-com" boom and bust. (9) The residuals from the two equations are spliced together as a measure of excess employment--private sector employment that was larger or smaller than what might be expected based on private sector GDP. These are shown in Figure 10. The hiring surge in the late 1990s reflects the expectations for "new economy" growth to continue, and the gap in part reflects the fact that many people were employed by companies that produced few or no sales (indicated by private GDP). Employment was not unusually high as the economy entered the most recent recession suggesting that over-hiring was not the main cause of the rapid decline in employment. Rather, it was a sudden shift in the desired level of employment that was triggered by the decline in consumer spending. With a higher savings rate, permanently fewer employees will be needed going forward.
There is no doubt that psychology played a very important role in the development of the recession decline. But, as noted above, the behavior of the Small Business Optimism Index was not much different from that observed in the 1980-82 period. It should be noted, however, that announcements of policy changes can have a major impact on owner expectations. For example, the announcement of the first cut in the Federal Funds Rate took a heavy toll on the economic outlook of owners. Just before the Federal Reserve Board announced the first of a series of rate cuts and warned of a recession on September 18, 2007, a net 10 percent of the owners in the NFIB September 2007 survey expected business conditions to be better in six months. Among the last 100 interviews, a negative 11 percent expected better business conditions, a huge loss of confidence (a 21-point decline in the net percent of optimists). Hiring plans experienced a similar reduction--going from 16 percent planning to add employees before the rate cut to 2 percent after. There were no intervening important economic events that might have triggered this decline. (10) No doubt, the rhetoric surrounding the "financial crisis" produced similar responses from business owners, reducing the level of economic activity in the small business sector.
A key mechanism transmitting exogenous shocks such as the bursting of the dot-com or subprime bubbles to the "real" sector (output and employment) is alleged to be inflexible wages and prices. In moving from one equilibrium to another, this makes sense, inasmuch as people work for "stuff" rather than money, and the value of money is understood in terms of the quantity of stuff a monetary unit will buy. It does appear that firms adjusted prices much more quickly in the current recession than in any past decline, but employment (and output) bore much more of the adjustment than compensation, although compensation displayed much more responsiveness than at any time in the recent past. This suggests that the economy could recover from a shock of a given magnitude much more quickly on its own than in the past and that less government intervention might be called for.
The inventory adjustment process also appeared to be much deeper and faster, consistent with a much more rapid response to the economic shock from financial markets. Government policies that retard these adjustments, of course, blunt the markets ability to make adjustments and weed out enterprises that inefficiently deploy resources. This lowers subsequent growth and growth potential. Every recession is different, of course, and unfortunately we have relatively few data points to analyze for each one. Whether prices and compensation and inventory adjustments can be counted on to exhibit the same responsiveness in future recession events is unclear. But, absent any apparent change in the structure of small firms, it appears that the most likely reason for the rapid and pervasive changes in prices, inventory and employment may have been the pace of events, amplified by the psychology of the event, the media coverage, the alarmists, and the clear focus of events on the banking system. The stock market and home-value declines were hard to take, and consumers responded by cutting spending dramatically. That said, the ultimate issue is whether or not massive government intervention helps the economy or retards the adjustment process and misallocates resources. This paper does not answer that question with clarity, but does suggest that wage and price flexibility may have been enhanced over the past 20 years, and it clearly shows that small-business owners can act quickly to adjust wages and prices, as well as employment and inventories.
Abel, A., Bernanke, B., and Croushore, D. 2008. Macroeconomics, 6th ed. Pearson/Addison Wesley.
Blinder, Alan. 1994. "On Sticky Prices: Academic Theories Meet the Real World," in Monetary Policy, edited by N.G. Mankiw, University of Chicago Press.
Dunkelberg, William, and Scott, Jonathan. 2009. "The Response of Small Business Owners to Changes in Monetary Policy." Business Economics, 44(1): 23-37.
Kashyap, Anil. 1995. "Sticky Prices: New Evidence from Retail Catalogues." Quarterly Journal of Economics. 110(1): 245-74.
(1) Kashyap  looked at price changes on identical items in major catalogue sellers and found that many prices change little if at all in successive new catalogue issues. However, this depends primarily on the competitive and inflationary environment. These sellers do not have the market power to simply increase prices each time a catalogue is printed, and some prices arc reduced as well. But this analysis says little about the price behavior of 6 million small employers and does not deal with pricing in the service sector, which now dominates the U.S. economy.
(2) Economic surveys were initiated in the fourth quarter of 1973, conducted in the first month of each quarter through 1985, monthly after that, with NFIB's member first, varying in number from roughly 300,000 to 600.000 over the years. "Quarterly" observations are based on the survey taken in the first month of each quarter
(3) The net percent is the percent of owners giving a positive or favorable response less the percent of owners providing an unfavorable response.
(4) It is possible that some owners interpret the question in the context of overall compensation costs, which are reduced when employees are laid off as well as when compensation (wages or benefits) is reduced. But, the question is unchanged over time and thus, whatever "interpretational bias" might be present, it can be presumed to be constant over time.
(5) The Small Business Optimism Index is based on the net percent of favorable responses to 10 forward looking questions. These can be viewed in detail at www.smallbus.org or at www.nfib.org/research.
(6) The Index peaked at 102.9 in April 1977. There were no monthly surveys prior to 1986. "Quarterly" surveys are conducted in the first month of each quarter.
(7) Nominal interest rates were very different in these two periods. If owners were unable to distinguish between real and nominal rates, this could explain the heightened concern in the 1980-82 period. Interest rates were very high, but marginal tax rates were higher, and the value of inventory appreciated with the high rate of inflation that prevailed in that period. In the 2007-09 recession, interest rates and inflation rates were historically very low.
(8) A stock-adjustment model of inventory holding would have actual and expected sales as major determinants of desired stocks. Both of these variables declined dramatically starting in 2008:Q4.
(9) The split at 1993 was arbitrary, but the models were quite different in the two periods, with a much larger coefficient on private GDP in the pre-1993 period. Using a model estimated over the entire period yields the same picture over time, but with wider numerical fluctuations.
(10) Job creation plans, capital spending plans, and inventory investment plans also fell [Dunkelberg and Scott 2009).
WILLIAM C. DUNKELBERG, JONATHAN A. SCOTT, MICHAEL J. CHOW*
*William C. Dunkelberg is the Chief Economist of the National Federation of Independent Business and professor of economics and former Dean. School of Business and Management, Temple University. Dr. Dunkelberg is also Chair of the Global Interdependence Center and Chairman of the Board of Liberty Bell Bank. He received BA. MA, and Ph.D. degrees in economics from the University of Michigan and has served on the faculties of the University of Michigan, Stanford University, and Purdue University. He also served as Study Director at the Survey Research Center, with responsibility for the Survey of Consumer Finances.
Jonathan A. Scott is an associate professor of finance at the Fox School of Business at Temple University and an adjunct scholar with the National Federation of Independent Business Research Foundation. He is currently the interim director of the Fox School Honors program and managing director of the Fox School's student managed investment fund. His research focuses on small business access to credit markets, with publications in a number of financial and small business journals. Prior to joining Temple, he was a senior financial executive at the Federal Home Loan Bank of Dallas during the thrift crisis in the mid- to late-1980s. He received his BA in economics from the University of Cincinnati and his MS and Ph.D. in economics from Purdue University.
Michael J. Chow is a Senior Data Analyst at the National Federation of Independent Business Research Foundation. He has previously worked for the Office of Economic Policy and Analysis at the Department of Labor and for the President's Council of Economic Advisers. His research interests are in finance, macroeconomics, and applied econometrics. He received an M.Sc. in Econometrics and Mathematical Economics from the London School of Economics and a B.S.E. in Electrical Engineering from Princeton University.
Business Economics (2010) 45, 94-101.
|Printer friendly Cite/link Email Feedback|
|Comment:||Inflexible wages and prices? evidence in the current recession.|
|Author:||Dunkelberg, William C.; Scott, Jonathan A.; Chow, Michael J.|
|Date:||Apr 1, 2010|
|Previous Article:||The impact of the housing market boom and bust on consumption spending.|
|Next Article:||Victorian financial crises and their implications for the future.|