A Response to Wynne’s “Sticky Prices: What is the Evidence?”

Mark Wynne attempts to look at the evidence behind changes of the stock of money and whether this has implications for the overall economy (i.e. employment, growth rates) in the short run.  This has obvious implications for effectiveness of monetary policy because through open market operations, the Federal Reserve controls the money supply.  For over two hundred years, this issue has been “debated” and the only conclusions that seem to be able to be drawn from this is that “prices are ‘sticky’ at nonmarket-clearing levels.”  This will directly effect the real factors of the economy.  Suppose that people were magically inundated with more money than they had before.  Also, suppose that this increase in the money supply was a one-time, unexpected policy.  Since it can be assumed that each person was holding the optimal amount of cash previous to the increase in the money supply, this excess cash would be spent on stuff.  However, if everybody spent their excess cash holdings to return to their optimal holding of cash (which existed before the increase in the money supply), nothing will motivate the producers to put out more output.  Thus, the long-run result would be an increase to the price level by the same proportion that the money supply was increased.  New Keynesians, however, are interested in the “transition stage” that occurs between the event that got the economy out of disequilibrium until the time that the economy has restored equilibrium.  This transition stage, according to Wynne, could see one of two scenarios–either an instantaneous increase in the price level, which would end the story, or a rigidity of prices.  The rigidity of prices is the more interesting situation.  If some producers are slow to raising their prices, due to the menu costs and other situations that were discussed in class (even though nominal demand has increased with this excess amount of money in the economy), then output in the short run may increase without an increase to prices.  This increased output would show up as a real increase in the short run until all firms had a chance to raise prices equal to the growth rate of the initial increase in the money supply.  Wynne mentions in the introduction that his article will focus on sticky prices rather than wages because many analysts view the failure of wages to adjust to changes in the economy as wage stickiness.  Rather, this “rigidity” is due to labor contracts that are spelled out, which requires the wage to be paid out in installments in the form of paychecks.  Therefore, it is because of this locked-in labor contract that wages don’t adjust as often as prices.

Wynne’s earlier study with Sigalla in 1993 concluded that raw data that are used to compile the producer price index and the consumer price index are often list prices instead of transaction prices.  There are two answers for this practice.  One is so firms protect themselves against potential antitrust litigation and the other reason is so these spelled out prices don’t fall into the hands of competitors.  In order to get around this dilemma, the BLS will take the average sticker price of various stores and the average discount or coupons associated with the purchase of this product.  This averaging of raw price data makes for a difficult time in assessing the flexibility of the prices.  Since some average prices fluctuate more than their “constituent price series,” this, too, will make for an unreliable estimate as to the flexibility of prices.  Wynne points to the earliest studies of the frequency of price changes, which was conducted by Mills (1927).  He developed a wholesale price index (WPI) in which he recorded 206 commodities.  The WPI ranged from 0 to 1 with an index value equal to zero if the price never changed over the period monitored and a value of one if the price changed every period recorded.  The shapes of these graphs were U-shaped, that is, there were a lot of commodities that didn’t exhibit price changes over the recorded time frame and a lot of commodities that exhibited price changes almost every period.  There were fewer commodities that fell in the middle range.  The products that exhibited the most price changes were farm products.  (An interesting note is that during WWI, the WPI graph didn’t show a U-shaped distribution, but rather an even distribution of commodities exhibiting ratios in the middle of the graph and a lot of commodities at the right-hand side of the graph.  The two criticisms of Mills’ work are that he used averages and used list prices rather than transaction prices.

Cecchetti’s (1986) study of magazines is a good example of price stickiness.  He was able to get away from the criticisms that plagued Mills (1927) because magazine prices are transaction costs and there are few discounts associated with magazines.  His sample period from 1953-1979 suggested high price stickiness because real costs were decreasing as nominal prices were increasing during high periods of inflation in the 1970s.  Therefore, he concluded that menu costs–that is, these fixed costs–were very high.  Nevertheless, his study had other shortcomings that Mill didn’t experience.  He looked at newsstand prices of magazines, but it is recognized that many people buy a subscription for a magazine, which is similar to the criticism of labor contracts.  That is, these individuals enter into a contract with the magazine company for a year and many times, subscriptions allow for customers to receive discounts.  This commonality, unfortunately, isn’t reflected in Cecchetti’s 1986 study.

Koelln and Rush (1993) look at whether controlling for quality of a product affects price rigidity, something for which Cecchetti couldn’t control.  Looking at magazine data from 1950-1989, the two conclude that Cecchetti’s study of price rigidity was overstated.  Looking at the number of pages of text, they were able to conclude that as inflation “erodes the real price of the magazine,” the number of pages of text will decline.  Therefore, this “price rigidity” can be confused with the declining quality of the product.  Carlton (1986) revisits Stigler and Kindahl (1970) in which the two looked at transaction costs rather than list prices of various industrial commodities.  Stigler and Kindahl (1970) collected data from buyers and not from sellers because buyers have less of an incentive to report list prices.  Thus, Carlton concludes that industrial commodities, especially industries dealing with steel, chemicals, and cement kept prices unchanged for a period of at least one year.  Other studies that have been conducted were those in the retail business.  Kashyap (1991) looked at retail catalogues and concluded that nominal prices remain unchanged for periods of at least one year and when prices do change, both the magnitude and number of changes is irregular.  Blinder (1991) conducted interviews with firms and found out that fifty-five percent of the firms interviewed claimed to have changed their prices no more than once a year, with only ten percent claiming to change prices monthly.  An interesting note from Blinder’s study is that three-fourths of the firms will change something other than price (i.e. delivery lags, quality of products) when demand is tight.

There have been some overall assessments of price stickiness.  Many studies mentioned by Wynne only have to do with a small fraction of the country’s GDP (i.e. magazines).  Other studies deal with intermediate products rather than finished products (i.e. industrial companies).  Lastly, of most importance is the price rigidity studies that actually deal with transactions involving money.  Since many products are bought via credit, it doesn’t represent the demand for money, and consequently, these studies will not determine whether money plays an important role in price rigidities.  Wynne also brings it to the attention of the reader that studies, such as Cecchetti and Stigler-Kindahl reaffirmed their theories of price rigidities rather than looking for price rigidities.  What I mean by this is that these two studies picked areas of the economy where it was already hinted at that prices were already inflexible and thus, the studies produced biased results that reaffirmed, rather than proved, that prices in these markets were rigid.  Carlton (1983) also criticizes the studies done on price rigidities.  For instance, it was known that there were price controls during WWII that held nominal prices at a constant level.  However, to get around this, the quality of the products being offered were decreased.  Thus, in a sense, the products were no longer homogeneous because of the varying qualities of the products being assessed.

Wynne concludes that there is little evidence to suggest that prices are sticky in the overall economy.  With all the thinking and assumptions of price stickiness, he was shocked that only three studies were able to be produced that showed actual price stickiness.  There are ways to deceive the idea of price stickiness by either withholding delivery during a heightened demand or by lowering the quality of the product.  In essence, just because markets take longer to clear than in a Walrasian auction, doesn’t mean that the evidence points to price rigidities.  To go back to the original question regarding the effectiveness of monetary policy and its effects on the real side of the economy–only a small degree of price rigidity needs to be in place for those external, monetary shocks to be able to trace out the observed business cycle.  Even if all prices were deemed flexible, monetary policy could still affect the real side of the economy–the shocks would simply then come from macro market failures or market incompleteness.

Source: Wynne, Mark A.  1995.  Sticky prices: What is the evidence?  Federal Reserve Bank of Dallas Economic Review (1st Quarter): 1-12.

Comments are closed.