The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Federal Reserve Bank of Chicago
The Minimum Wage, Restaurant
Prices, and Labor Market Structure
Daniel Aaronson, Eric French, and
James MacDonald
REVISED
August 3, 2007
WP 2004-21
The Minimum Wage, Restaurant Prices, and
Labor Market Structure
Daniel Aaronson, Eric French, and James MacDonald∗
August 3, 2007
Abstract
Using store-level and aggregated Consumer Price Index data, we show that restaurant prices rise in response to minimum wage increases under several sources of identifying
variation. We introduce a general model of employment determination that implies minimum wage hikes cause prices to rise in competitive labor markets but potentially fall in
monopsonistic environments. Furthermore, the model implies employment and prices are
always negatively related. Therefore, our empirical results provide evidence against the
importance of monopsony power for understanding small observed employment responses
to minimum wage changes. Our estimated price responses challenge other explanations
of the small employment response too.
Comments welcome at efrench@frbchi.org or daaronson@frbchi.org. Author affiliations are Federal
Reserve Bank of Chicago, Federal Reserve Bank of Chicago, and Economic Research Service, U.S. Department of Agriculture respectively. Work on the store-level price data was performed under a memorandum of
understanding between the Economic Research Service at the Department of Agriculture and the Bureau of
Labor Statistics (BLS), which permitted onsite access to the confidential BLS data used in this paper. We
thank Bill Cook, Mark Bowman, and Scott Pinkerton of the BLS for their advice and help. We also thank
Gadi Barlevy, Jeff Campbell, William Evans (the editor), Bob LaLonde, Derek Neal, Dan Sullivan, the referees, and seminar participants at the Federal Reserve Bank of Chicago, the University of Illinois-Urbana, the
Econometric Society, and SOLE meetings for helpful comments and Tina Lam for excellent research assistance.
The views expressed herein are not necessarily those of the Federal Reserve Bank of Chicago, Federal Reserve
System, Bureau of Labor Statistics, or U.S. Department of Agriculture. Author correspondence to Daniel
Aaronson or Eric French, Federal Reserve Bank of Chicago, 230 S. LaSalle St., Chicago, IL 60604. Telephone
(312)322-6831, Fax (312)322-2357.
∗
1
1
Introduction
This paper utilizes unique data to test whether restaurant prices respond to minimum
wage changes. We find that restaurant prices unambiguously rise after minimum wage increases are enacted.1 Furthermore, these price increases are larger for establishments that are
more likely to pay the minimum wage. These results are derived from a panel of store-level
restaurant prices that are the basis for the food away from home component of the Consumer
Price Index (CPI) during a three-year period with two Federal minimum wage increases, and
are corroborated using a longer panel of city-level food away from home pricing from the
CPI.
Because of the breadth of our price data, we can take advantage of several sources of
variation. First, some states set their minimum wage above the Federal level. Second, we can
distinguish restaurants that tend to pay the minimum wage from those that do not. Third,
the fraction of workers paid at or near the minimum wage varies across geographic areas. All
three sources of variation indicate that most, if not all, of the higher labor costs faced by
employers are pushed onto customers in the form of higher prices.
As suggested by Brown (1999), the size and sign of these price responses can be used
to infer whether monopsony power is important for understanding the employment response
to minimum wage hikes. The minimum wage literature has become contentious since Card
and Krueger’s (e.g. 1995, 2000) research found that an increase in the minimum wage has
no, or even a small positive, effect on employment. Therefore, their research contradicts
standard models of competitive labor markets, which, prior to their work, most researchers
suspected was relevant for industries which primarily employed minimum wage workers.2
1
We are not the first to estimate price pass-through in this context. See, e.g., Converse et al (1981),
Card and Krueger (1995), and Aaronson (2001). Card and Krueger use Consumer Price Indices for Food
Away from Home in 27 large metropolitan areas over a three year period, finding larger price increases in
those cities with higher proportions of low-wage workers. Although their estimates are consistent with full
pass-through, their standard errors are extremely large. They cannot reject zero price pass-through in many
of their specifications. Moreover, additional evidence from specific state increases in Texas and New Jersey
suggests close to no price response. As a result, they conclude that their estimates are “too imprecise to reach
a more confident assessment about the effects of the minimum wage on restaurant prices.” The size of the
panel that we use in this study allows us to estimate price effects much more precisely.
2
See, for example, Brown et al (1982). The standard competitive model’s predictions are generally consistent with recent views reported in a survey of leading labor economists (Fuchs, Krueger, and Poterba 1998)
as well. However, a full quarter of respondents believe there is no teenage disemployment effect from a 10
percent increase in the minimum wage.
2
However, their results are consistent with monopsony power in the labor market, as Stigler
(1946) discussed many years ago. The diverse findings reported in the flurry of replies to
their work (e.g. Neumark and Wascher (1996, 2000), Deere et al. (1995), Kim and Taylor
(1995), Burkhauser et al. (2000), Dickens et al (1999)) led one observer to note that “[Card
and Krueger’s] lasting contribution may well be to show that we just don’t know how many
jobs would be lost if the minimum wage were increased...and that we are unlikely to find
out by using more sophisticated methods of inference on the existing body of data. What is
needed is more sophisticated data” (Kennan 1995).
Restaurant prices complement the existing evidence on employment responses because,
as we show below, output prices and employment are unambiguously negatively related in response to an exogenous change in wage rates. In order to show this relationship, we introduce
a general model of employment determination that allows for a range of output and input
market structures. Part of the reasoning behind the negative relationship between output
prices and employment is based on the negative relationship between prices and output. We
also add some weak assumptions to the model to show that output and labor input are positively related. Therefore, if the output price rises in response to a minimum wage hike, both
output and labor input have fallen. This will be the case if labor markets are competitive.
Conversely, if the output price falls in response to a minimum wage hike, total output and
labor input have increased. This will potentially be the case if firms have monopsony power
in the labor market.
Research on monopsony power has recently been revitalized by the empirical and theoretical work of Card and Krueger (1995), Burdett and Mortenson (1998), Bashkar and To
(1999), Ahn and Arcidiacono (2003), Flinn (2006), Manning (1995), and Rebitzer and Taylor
(1995).3 But our results suggest that competition is likely more relevant than monopsony.
Moreover, in Aaronson and French (2007), we show that a computational model of labor
demand with a competitive labor market structure predicts price responses that are comparable to those found in this paper. The employment elasticities that are derived from that
3
Although few believe that low wage labor markets are characterized by pure monopsony, as in Stigler
(1946), many models give rise to monopsony-like behavior that corroborate Card and Krueger’s findings of
small or even positive employment movements after a minimum wage increase. These include models where
transportation (Bashkar and To 1999) or employee search (Burdett and Mortenson 1998; Ahn and Arcidiacono
2003; Flinn 2006) is costly and employers are not able to discriminate high and low reservation wage workers.
Efficiency wage models such as Manning (1995) and Rebitzer and Taylor (1995) can also cause monopsony-like
employment effects. See Boal and Ransom (1997) and Manning (2003) for broader reviews.
3
calibrated model are within the bounds set by the empirical literature.
To be clear, as Boal and Ransom (1997), among others, point out, our results do not
necessarily prove labor markets are competitive. Although the results are clearly consistent
with this conclusion, if the minimum wage is set high enough, positive comovement between
the minimum wage and prices may be consistent with the monopsony model as well. We
discuss this point more formally below.
However, our results provide evidence against the hypothesis that monopsony power is
important for understanding the observed small employment responses found in some minimum wage studies. Indeed, our estimated price responses provide evidence against other
explanations of the small employment response as well, including the potential substitution
of nonwage for wage compensation and the importance of endogenous work effort or efficiency
wages. They do, however, provide support for a model of ”hungry teenagers,” whereby higher
income resulting from a minimum wage increase causes low wage workers to buy more minimum wage products, attenuating the disemployment effect of the minimum wage. Although
our test answers a fairly narrow question, we believe that the answer to this question is of
broad interest. Given that the low observed employment responses to minimum wage changes
sparked particular interest in the importance of monopsony power in the labor market, our
results should temper this interest.
Finally, it is important to emphasize that our estimates are for the restaurant industry
only. This industry is a major employer of low-wage labor and therefore a particularly relevant
one to study.4 However, as a result of different intensities of use of minimum wage labor,
substitution possibilities, market structure, or demand for their products, other industries
might face different employment responses. See Aaronson and French (2007) for further
details.
4
Eating and drinking places (SIC 641) is the largest employer of workers at or near the minimum, accounting
for roughly a fifth of such employees in 1994 and 1995. The next largest employer, retail grocery stores, employs
less than 7 percent of minimum or near minimum wage workers. Moreover, the intensity of use of minimum
wage workers in the eating and drinking industry is amongst the highest of all sectors, with approximately 23
percent of all workers, encompassing 11 percent of the industry wage bill, within 10 percent of the minimum
wage. All calculations in this footnote are based on the Current Population Survey’s outgoing rotation groups.
Other prominent examples of studies that concentrate on the restaurant industry include Katz and Krueger
(1992), Card and Krueger (1995,2000) and Neumark and Wascher (2000).
4
2
Data
Under an agreement with the Bureau of Labor Statistics (BLS), we were granted access
to the store-level data employed to construct the food away from home component of the
CPI during 1995 to 1997.5 While the time frame is short, this three-year period contains
an unusual amount of minimum wage activity. A bill signed on August 20, 1996 raised the
federal minimum from $4.25 to $5.15 per hour, with the increase phased in gradually. An
initial increase to $4.75 (11.8 percent) occurred on October 1, and the final installment (8.4
percent) took effect on September 1, 1997. Moreover, additional variation can be exploited
since price responses will vary geographically. This occurs for two reasons. First, market
wages may exceed minimum wages in some areas but not in others. Second, some states set
minimum wages above the federal level. We capture the latter source of heterogeneity by
allowing the effective minimum wage to be the maximum of the state and federal level.6
The sample itself is based on nearly 7,500 food items at over 1,000 different establishments
in 88 Primary Sampling Units (PSUs).7 Because restaurants in some geographic areas are
surveyed every other month, all numbers are reported as bimonthly (every other month)
price changes.8 Within an establishment, specific items, usually 7 or 8, are selected for
pricing with probability proportional to sales. During our time frame, an “item” usually was
a meal, as the BLS aimed to price complete meals as typically purchased (for example, a
meal item might consist of a hamburger, french fries, and a soft drink). Our dataset codes
items broadly, such as breakfast, lunch, dinner, or snacks. Unfortunately, because there are no
specific item descriptions, we cannot tie price changes to item-specific measures of input price
changes (such as ground beef or chicken price indexes). Nevertheless, the BLS strives to price
5
Because the BLS introduced a complete outlet and item resampling in January 1998, we only use data
through December 1997. Data prior to 1995 are no longer available. Bils and Klenow (2004) use the same
1995 to 1997 period.
6
This source of variation is especially useful in section 3.3, when we look at city level variation between
1979 and 1995. During 1995 to 1997, 10 states (not including Alaska which is always $0.50 above the federal
level) had minimum wages above the federal level for some part of the period. Six states (HI, MA, NJ, OR,
VT, and WA) were at or above the federal minimum of $4.75 prior to October of 1996. Three states (HI, MA,
and OR) were at or above the federal minimum prior to September 1997. Further variation is available from
states (CA, CT, NJ, WA) that were between the old $4.75 minimum wage but below the new $5.15 minimum
wage.
7
The 88 PSUs cover 76 metropolitan statistical areas and 12 other areas representing the urban non-metro
U.S.
8
PSUs were assigned to one of three reporting cycles: outlets in the five largest were surveyed each month,
while others were surveyed in two bimonthly cycles of odd and even numbered months. For sample size and
consistency, we randomly assigned outlets in the five largest PSUs to odd or even two month cycles.
5
identical items over time, and codes in our database describe temporal item substitutions due
to discontinuance and alteration. Our analysis focuses on price changes for identical items,
and we do not compare prices where the BLS has made an item substitution.9
A particular advantage of this data is its depiction of the type of business. Limited service
(LS) outlets, which account for roughly 30 percent of the sample, are those stores where meals
are served for on- or off-premises consumption and patrons typically place orders and pay at
the counter before they eat. Roughly half the sample is comprised of full service (FS) outlets,
establishments that provide wait-service, sell food primarily for on-premises consumption,
take orders while patrons are seated at a table, booth or counter, and typically ask for
payment from patrons after they eat.10 The minimum wage is likely to increase wages at
LS restaurants more than at FS restaurants, for two reasons. First, wages for cashiers and
crew members are higher, perhaps by 60%, at FS restaurants than LS restaurants.11 Thus,
a higher fraction of workers are paid the minimum wage at LS restaurants. Second, many
FS employees are paid through tips and the federal tipped minimum wage remained $2.13
throughout our sample period. Thus, minimum wage changes have smaller effects on wages
9
Firms could respond to a minimum wage increase by reducing quality, instead of raising price. While we
do not have direct measures of quality, the dataset notes whether an item is the same as the item priced in
the previous month. There is no evidence of any increased incidence of item changes or substitutions following
minimum wage increases, suggesting that quality changes or item substitution are not a standard means of
dealing with a cost shock. There also might be concern that a minimum wage increase changes the composition
of items sampled. If revenues are negatively correlated with prices and sampling probability is a function of
sales, a change in the minimum wage could result in a shift in the distribution of sampled products towards
high priced items (or stores with fewer minimum wage workers). To minimize this concern, we ran everything
with sampling weights and found identical results.
10
The BLS replaced an old ordering with these types of business codes in July 1996, and began to report
price indexes for type of business groupings after our data period in January 1998. Businesses surveyed
early in our sample period were retroactively assigned the new codes. The remaining fifth of non-LS and
non-FS outlets, which we usually exclude from this analysis, include meals consumed at department stores,
supermarkets, convenience stores, gas stations, vending machines, and many other outlets.
11
Assuming that LS wage rates are identical to U.S. McDonald’s wage rates collected by McKinsey Global
Institute and reported in Ashenfelter and Jurajda (2001), then wage rates among cashier and crew members
in FS establishments in the outgoing rotation files of the CPS are about 60 percent higher than wage rates in
LS establishments. Ashenfelter and Jurajda report the average U.S. McDonald’s wage for crew and cashier
workers was $6.00 and $6.50 in December 1998 and August 2000, respectively. We compared these figures
to the average wage of $7.81 and $8.52 for CPS workers that report their industry as eating and drinking
places and their occupation as food preparation and service occupation, janitors and cleaners, or sales counter
clerks during the fourth quarter of 1998 or the third quarter of 2000. Assuming all LS establishments pay the
same wage as McDonalds and noting that the 1997 Economic Census of Accommodations and Food Services
reports that 48 percent of all employees in FS and LS establishments are employed in the LS sector, we can
back out that FS establishments pay roughly 60 percent higher hourly wages than LS establishments within
these occupation codes. The Economic Census also reports average weekly wages that are approximately 20
percent higher in FS establishments. But this calculation does not correct for differences in hours worked per
week and cannot refine the sample by occupational class.
6
in FS outlets than LS outlets because tip earnings usually exceed effective minimum wages.12
3
Estimates of Price Pass-Through
One problem with price (as well as employment) data is that they are potentially measured
with error.13 In other words, pijkt = p∗ijkt + ǫijkt , where pijkt is the measured price of item
k at outlet j in state i during month t, p∗ijkt is the model predicted value, where the model
is described in section 4, and ǫijkt is measurement or model misspecification error.14 We
approach the data in three ways.
3.1
Store-level Descriptions of Price Increases and Decreases
Our first approach ignores errors in the price data and simply tabulates price increases and
decreases after a minimum wage change. In the model described in section 4, we formally show
that price data can be used to infer labor market structure. In particular, price cuts tend to
be an outcome unique to monopsonistic labor markets. In the absence of measurement error,
observed price cuts allow us to identify individual firms that potentially have monopsony
power. On the other hand, if variability in ǫijkt is significant, measurement or misspecification
error may cause us to erroneously infer monopsony power when in fact none exists.
12
Federal law sets a separate cash minimum for tipped employees (which is $2.13 throughout our sample
period), but requires that tips plus cash wages must at least equal the nontipped employee minimum. For
example, in September 1996, $2.62 ($4.75-2.13) in tips are allowed to be applied to a tipped employee’s wage
to reach the minimum wage. In 1996, only Rhode Island and Vermont changed their state-specific tipped
minimum. In 1997, Maryland, Michigan, North Dakota, and Vermont did as well. Of these states, only
Maryland and Michigan are included in our CPI sample.
13
Measurement error is unlikely to be important in this data (Bils and Klenow 2004). The BLS has procedures in place to flag and investigate unusual observations. Nevertheless, some measurement error may exist
for the following three reasons. First, the price on the menu may differ from the price paid because of discounts and coupons distributed outside of outlets. Prices are collected net of sales and promotions but some,
particularly those not run by the outlet itself may be missed. Second, high frequency but short-term price
changes may not be captured by our monthly data (Chevalier et al 2003). Third, surveyors may falsely report
last month’s price instead of going to the restaurant to record the price, a practice known as “curbstoning”.
There is no reason to think that any of these sources of error are correlated with the presence of a minimum
wage change.
14
There are other interpretations for ǫijkt , such as menu costs of switching prices. Moreover, many firms in
our sample offer short term sales for reasons that are unrelated to changes in input prices.
7
Outlet type
Limited Service
Two month period with minimum wage change No
A. Share of price changes, observation is an item
Limited Service
Yes
Full Service
No
Full Service
Yes
Percent increases
Percent decreases
Item Observations
22.6**
2.5
3,853
10.7
1.8
44,632
12.0**
1.6
7,045
11.5
2.9
25,815
B. Share of price changes, observation is the average of all items at a store
8
Percent increases
Percent decreases
Store Observations
24.1
8.0
3,799
38.3**
6.7
551
19.5
5.1
6,809
22.4
4.8
1,036
5.3
8.4
4.8**
8.2
4.8
7.5
4.9
9.3*
C. Size of price changes (in percent)
Mean item price change—increase
Mean item price change—decrease
* (**) = Statistically different from months without a minimum wage increase at the 5(1) percent level.
Table 1: The Frequency and Magnitude of Store-Level Price Changes
Table 1 reports descriptive statistics on the frequency and size of price changes for Limited
Service (column 2) and Full Service (column 4) outlets in the two months immediately after a
minimum wage change. For comparison, all other two month periods are reported in columns
1 and 3. In panel A, the observational unit is a food item. Because multiple items are surveyed
for each store, individual stores are in these computations up to 8 times each period. Panel
B computes price changes by store. That is, an average price is calculated from each store’s
sampled items and consequently a store is included, at most, once every two months.
There are several notable features of the data. First and foremost, prices increase in response to a minimum wage change. During the two months after a minimum wage increase,
22.6 percent of LS items increased in price. This is almost double the 11.5 percent share of
LS price quotes that are increased in months without a minimum wage increase.15 Moreover,
as expected, the minimum wage effect is substantially smaller, although still statistically significant, for full service outlets. In such stores, the share of quotes that are higher than the
previous two months is 12 percent, exceeding months that do not follow a minimum wage
increase by 1.3 percentage points. Excluding small price changes (say, those less than 2 percent) that might be driven by measurement error reinforces these differences. Price increases
over 2 percent are more than twice as likely in minimum wage months in LS establishments
(17.6 versus 8.6 percent) and 22 percent more likely in FS establishments (8.9 versus 7.2
percent).
Conversely, there is little evidence that minimum wage increases cause price declines.
The share of prices that decline is stable throughout the three years regardless of whether the
minimum wage has been altered. The results are identical when small changes are excluded.
Therefore, based on incidence alone, the data suggest that many firms raise their price, but
few reduce it, in response to a change in the minimum wage.
These results are robust to looking at price movements at the store-level, as reported in
panel B. Here, prices are computed by averaging the price of all items in a store. Again, there
is a notable acceleration of LS outlet price increases following a minimum wage increase (38.3
percent versus 24.1 percent) but no unusual increase in price declines during these periods.16
15
The unconditional probability of a price change matches Bils and Klenow (2004), who also use CPI
microdata.
16
We have explored looking at longer intervals but are quite limited by our data. The September 1997
increase is only 4 months before the end of our sample, so we are forced to rely on a single comparison: preand post- the 1996 increase. The results are a bit more muted but similar inferences can be drawn. LS price
9
Finally, increasing the frequency of price changes is not the only avenue for firms to raise
or lower prices. The size of price changes could be altered as well.17 However, panel C shows
that, if anything, price increases tend to be slightly smaller in size after minimum wage
increases. The size of LS price cuts are unaffected by minimum wage changes. There is some
evidence of larger price cuts in FS establishments after minimum wage hikes, but this is a
rare event (only 1.6 percent of all FS item observations). It is worth noting that, while price
cuts (in FS and LS stores) are rare, they can be large. Over a quarter of all food away from
home price cuts exceed 10 percent, and over a tenth exceed 20 percent. By comparison, price
increases are more tightly concentrated, with about half under 4 percent and less than a tenth
above 10 percent.18 However, there is little evidence that the size of price cuts change in any
meaningful way after a minimum wage increase. Kolmogorov-Smirnov D-tests for differences
in price distributions finds no significant shift in the size of price cuts following minimum
wage increases.19 The fact that these large declines exist in roughly the same fashion in nonminimum wage change periods suggests to us that the largest cuts typically reflect temporary
sales.
Assuming that all markets are competitive (or, as we point out in Section 4.3, factor
markets are competitive and product markets are monopolistically competitive) and firms
have a constant returns to scale production function, it is straightforward to show that all
cost increases will be passed onto consumers in the form of higher prices. If minimum wage
labor’s share of total firm costs is smin , then a 10% increase in the minimum wage should
increase the product price smin × 10%.
To get a sense of whether the observed price responses are consistent with competition,
we note that minimum wage labor’s share of total costs is equal to labor’s share of total
hikes in the 10 months after October 1996 are 6 percentage points (53 vs. 47 percent) more common than in
the average 10 month period in the year and a half prior to October 1996. Price declines occur slightly more
frequently (but not statistically so) in the 10 months after the increase: 6 percent versus 5 percent.
17
See MacDonald and Aaronson (2006) for a more extensive discussion of the various ways restaurants
construct price changes. The price change distributions are available upon request from the authors.
18
Price changes cluster near the mean, with excess kurtosis of 62.0 and 80.8 for price changes among LS and
FS outlets, respectively. Distributions of increases and decreases are also quite peaked compared to normal
distributions, with excess kurtosis of 14.2 (LS) and 8.6 (FS) for increases, and 1.6 (LS) and 6.8 (FS) for
decreases. Kashyap (1995) also reports positive excess kurtosis in his sample of catalog prices.
19
In particular, the K-S D-test suggests no significant difference in the distribution of increases among LS
outlets. Small changes (less than 2 percent) occur among 22 percent of LS price changes in minimum wage
bimonths, compared to 25 percent in all other bimonths. Large changes (greater than 10 percent) occur
among 12 percent of LS price changes observed in minimum wage bimonths, compared to 13 percent in all
other bimonths. There is a statistically significant difference, driven by the higher incidence of very small
increases, among FS outlets.
10
costs multiplied by minimum wage labor’s share of labor costs. 10-K company reports, the
Economic Census for Accommodations and Foodservices, and the IRS’ Statistics on Income
Bulletin all provide an estimate of labor’s share of total costs, and in each, the sample median
and mean are around 30 to 35 percent.20 Unfortunately, we are less certain of minimum wage
labor’s share of total labor costs for the average firm. Using household level data, we know
that about a third of all restaurant workers are paid near the minimum wage over this
time period, constituting 17% of all payments to labor.21 Using these values, we make two
calculations that bound the competitive response. If there is only one type of labor, all firms
have the same employment level, and all firms either pay 0% of their workers or 100% of
their workers the minimum wage, depending on the labor market, then 33% of all firms pay
the minimum wage. Given this, and the fact that about 33% of total costs are in the form of
labor costs, then a 10% increase in the minimum wage raises prices by 10% × 33% × 33% =
1.09%. Alternatively, if all firms hire above minimum wage labor in equal proportions, then
each restaurant must have 17% of its labor costs going to minimum wage labor. Thus, a 10%
increase in the minimum wage should raise prices by 10% × 33% × 17% = 0.56%.
Aaronson and French (2007) use a calibrated model of labor demand that accounts for
both firm and worker heterogeneity to show that when these factors are explicitly accounted
for, the competitive model predicts prices will increase by roughly 0.7%. Moreover, because
20
Of the 17 restaurant companies that appear in a search of 1995 reports using the U.S. Security and
Exchange Commissions (SEC) Edgar database, the unconditional mean and median of the payroll to total
expense ratio from 10-K reports is 30 percent. This search uses five keywords: restaurant, steak, seafood,
hamburger, and chicken. Limited service establishments, like McDonalds and Burger King, are at or below
the mean. Full service restaurant companies, like Bob Evans and California Pizza Kitchen, lie above. Similarly,
the 1997 Economic Census for Accommodations and Foodservices reports payroll as 31 and 25 percent of sales
at full and limited service restaurants, respectively. Since 10-Ks from food away from home companies generally
show that wages account for 85 percent of compensation, the Economic Census’ estimate of labor share based
on compensation is roughly 36 and 29 percent at full and limited service restaurants. Another method of
calculating lfabor’s share is through a sampling of 1995 corporate income tax forms from the Internal Revenue
Service’s Statistics on Income Bulletin. Because operating costs are broken down by category, it is possible
to estimate labor’s share. We restrict the sample to partnerships because of IRS concern that labor costs
are notoriously difficult to decompose for corporations. Despite the quite different sampling of firms relative
to the Edgar Database, labor cost as a share of operating costs for eating place partnerships is of a similar
magnitude to the other estimates, roughly 33 percent. Finally, these figures correspond well to a 2002 survey
of restaurants by Deloitte and Touche (2003). Among limited service establishments, Deloitte and Touche find
that wages and salaries make up 31 percent of total expenses. Benefits add another 2 percent. The payroll to
expense ratio is roughly 2 percent higher for full service establishments.
21
See Aaronson and French (2007) for a description of this calculation. Because wage distributions are not
available in company reports, we estimate the share of employees that are paid at or near the minimum wage
from the outgoing rotation files of the CPS for the two years prior to the 1996 legislation. We use a survey in
Card and Krueger (1995, p. 162) to account for the share of workers paid slightly above the minimum wage
that are impacted by new legislation.
11
limited service restaurants are more likely to pay the minimum wage than full service restaurants, competition would imply larger price increases at limited service restaurants. Given
that some restaurants do not increase prices after minimum wage hikes, but restaurants that
do raise their prices usually do so by more than 0.7%, it is difficult to compare the observed
price response to the competitive prediction. Section 3.2 presents a statistical model to better
make this comparison.
3.2
Estimates of Price Pass-Through
The next approach provides a more complete statistical model of the price response to
a minimum wage change. In our basic model we regress the log change in prices for item k
at outlet j in period t on the percentage change in the minimum wage in state i over the
contemporaneous, lag, and lead periods, and a set of controls:
∆ ln pijkt =
H
X
h=1
αh ∆ ln pijkt−h +
2
X
βh ∆ ln P P It−h +
1
X
ωh ∆ ln wmin,it−h + uijkt
(1)
h=−1
h=0
We include wmin,it−1 and wmin,it+1 to allow a more flexible response to the legislation, as
price responses can play out over time. Consider the timing of the 1996-97 federal increase.
When the law was passed, firms knew that minimum wages would be increased on October
1, 1996, and again on September 1, 1997. It is conceivable that firms could react to the
expectation of an increase (i.e. at the bill discussion or passage), rather than the enactment
dates. However, empirically, we found no evidence of longer leads or lags.22
The vector of controls include contemporaneous and lags of changes in the producer price
index for processed foods to account for material input price shocks faced by sample outlets
(ln P P It ).23 To allow for mean-reverting price movements that typically occur after sales or
22
Businesses knew of the 1996 increase just 2 to 4 months prior to implementation. They knew of the 1997
increase, specified in the 1996 bill, 13 to 15 months before implementation. The 1996 increase could not have
been predicted until shortly before the House of Representatives vote on May 23, after a week of legislative
maneuvering that almost consigned the bill to defeat (Weisman 1996 and Rubin 1996). Even then, the final
timing of the minimum wage increase did not become clear until adoption of the conference report on August
2. Aaronson (2001) shows that longer-run price pass-through estimates are roughly the same size as short-run
estimates using aggregated U.S. and Canadian price data.
23
Aaronson (2001) accounts for the costs of particular food items, such as chicken, beef, bread, cheese,
lettuce, tomatoes, and potatoes, and finds similar aggregated results to those reported here. We also controlled
for broader labor market pressures using changes in CPS median wages and fixed chain and PSU (i.e. city)
effects (not shown). Minimum wage point estimates and standard errors are quite robust to the inclusion of
these variables. Furthermore, the fixed effects themselves added almost nothing to the model’s fit. Controls
for mealtype (breakfast, lunch, or dinner) are also included but have no impact on the results.
12
price hikes, we also experimented with controls for lags in ln pijkt . However, the minimum
wage estimates barely change whether lagged dependent variables are accounted for or not
(the version reported in table 2 includes them).24
Table 2 presents the basic results. Since quotes from the same outlet are unlikely to be
statistically independent, all standard errors account for quote (i.e. menu items within an
establishment) clustering, using Huber-White robust estimation techniques. We also checked
for error clustering by city, chain, and outlet. While within-outlet effects were important,
within-city and within-chain effects were not. All results are reported as elasticities and
multiplied by 10 to gauge the impact of a 10 percent minimum wage increase.
24
To be clear, the inclusion of a lagged dependent variable potentially leads to inconsistent parameter
estimates. In practice, this bias appears to be negligible. But for completeness, we also ran a specification that
included one lag in ln pijkt that is instrumented by thrice lagged prices and found statistically indistinguishable
results. Moreover, to capture any asymmetry in the mean-reverting process, we separately measure percentage
increases and decreases in an item’s price in the previous periods. Parameter estimates are of the expected
sign: past price cuts lead to current period price increases, and past price increases lead to current price cuts,
presumably because the original increase reflected temporary cost increases or because rivals didn’t match the
price increase. However, these responses are dampened substantially. Full reversion to prior prices implies
absolute coefficient values of 1, while the estimated effects fall well below 1 and usually below 0.1.
13
Variable
Column
∆ ln wmin,it−1
All
1
0.229
(0.064)
Limited service
2
0.334
(0.117)
Full service
3
0.234
(0.082)
All
4
0.225
(0.067)
Limited service
5
0.295
(0.120)
Limited service
6
0.202
(0.326)
Full service
7
0.249
(0.087)
Full service
8
0.246
(0.157)
∆ ln wmin,it
0.407
(0.070)
0.940
(0.135)
0.190
(0.086)
1.444
(0.531)
2.695
(0.883)
2.392
(1.005)
1.039
(0.692)
1.245
(0.706)
∆ ln wmin,it+1
0.077
(0.063)
0.275
(0.136)
-0.102
(0.073)
0.078
(0.067)
0.243
(0.149)
0.451
(0.431)
-0.082
(0.079)
-0.228
(0.164)
-0.161
(0.078)
-0.278
(0.133)
-0.242
(0.134)
-0.128
(0.100)
-0.133
(0.100)
0.94
0.62
0.30
1.84
1.29
0.73
1.84
1.35
0.87
0.56
0.31
0.05
0.60
0.33
0.07
∆ ln wmin,it *wage20
Total effect
0.713
(0.140)
1.549
(0.275)
0.322
(0.168)
14
At wage20= $5.00
At wage20= $7.00
At wage20= $9.00
month dummies?
include PPI?
R2
no
yes
0.070
no
yes
0.167
no
yes
0.017
no
yes
0.070
no
yes
0.171
yes
no
0.175
no
yes
0.019
yes
no
0.021
N
71,077
21,883
36,928
63,630
18,691
18,691
33,875
33,875
See text for detail. Controls not shown in table include three lags in ln pijkt and mealtype (breakfast, lunch, or dinner).
Huber-White standard errors corrected for clustering at the item and establishment level are in parentheses.
Sample sizes in columns 2 and 3 do not add up to column 1 because some establishments are not categorized
as full or limited service restaurants.
Wage20 is the 20th percentile of the MSA’s hourly wage distribution, calculated from the 1996 CPS.
Table 2: The Price Response to a 10 Percent Minimum Wage Increase
Column (1) reports the minimum wage effect for the full sample of food away from home
establishments. We find that a 10 percent increase in the minimum wage increases prices by
roughly 0.7 percent (with a standard error of 0.14), of which over half the response occurs
within the first two months after the minimum wage change. However, because this result
combines outlets where the minimum wage is binding with those where it might be less
important, two particular sources of variation can be used to identify the price response to a
minimum wage increase.
First, as in table 1, we can take advantage of variation in the intensity of minimum wage
worker use between limited (column 2) and full (column 3) service restaurants. The price
increase generated by a 10 percent minimum wage hike is roughly 1.55 percent (standard error
of 0.28 percent) for limited service outlets but a fifth that size for full service enterprises.25
Second, market wages vary across local labor markets. Where prevailing low-skill wages
far exceed minimum wages, minimum wage increases will have little impact on market wages
and consequently costs. Where the minimum binds for low-skill workers, changes in the
minimum wage will have strong effects on wages.26 Therefore, we test whether the price
response varies with respect to the pay of low-skilled workers. We are able to perform this
test because our data include precise outlet locations (addresses and telephone numbers) that
we link to MSA hourly wage distributions estimated from the 1996 Current Population Survey
(CPS). Columns (4), (5), and (7) interact one version of these measures, the 20th percentile
from the MSA’s hourly wage (wage20) distribution, with the contemporaneous minimum
wage change using the full sample and subsample of limited and full service outlets.27 We
find that minimum wage increases have larger effects on prices in low wage areas, among both
limited and full service outlets. An MSA where the 20th percentile of the 1996 hourly wage is
$5 leads to a 0.56 percent price increase among full service outlets and a 1.84 percent increase
25
Aaronson (2001) finds an elasticity of around 1.5 for fast food restaurants in the American Chamber of
Commerce price survey, consistent with the findings on limited service establishments.
26
We can look at this directly using the outgoing rotation files of the CPS. Using a state panel developed
from the 1979-2002 files, we regressed log hourly earnings in the restaurant industry on the prevailing minimum
wage and state and year fixed effects. The data were too noisy and replete with missing observations at the
monthly level. Nevertheless, we find that wages rise by 4.4 percent in the restaurant industry following a 10
percent increase in the minimum wage. Although we cannot distinguish full and limited service establishments,
we note that wages rise by 10.7 percent among teens and 7.1 percent among high school dropouts. The results
are similar when we restrict the sample to the CPI cities.
27
The results are also comparable when changing the year used to calculate wage20, interacting wage20
with the lag and lead minimum wage change. Wage data for the 12 non-metro PSU’s are drawn from the
non-metro parts of the outlet’s state. CPS codes are unavailable for 9 MSAs, so sample sizes decline when
area wage data are included in the analysis.
15
among limited service outlets. At a wage of $7 ($9), this effect drops to 0.31 (0.05) and 1.29
(0.73) percent for full and limited service firms. We have also used the share of minimum
wage workers, prob(wit = wmin,it ), as a measure of variation in minimum wage bindiness
across local labor market. The LS results are similar to those reported here, although less
precisely estimated.28 The coefficient on ∆ ln wit × prob(wit = wmin,it ) is 0.49 (0.39) for
LS establishments and 0.17 (0.30) for FS establishments. At the mean value of the share of
workers paid at or near the minimum wage (6 percent), a 10 percent increase in the minimum
wage increases LS prices by 1.15 percent and FS prices by 0.38 percent, very similar to columns
(5) and (7) in table 2.
As a robustness check, columns (6) and (8) report results of a regression with the wage20
interaction that also includes a full set of month dummies.29 The month dummies eliminate
the possibility that the minimum wage changes are confounding other contemporaneous national inflation or economic trends or seasonal factors (since the two federal changes occur
in September and October). But as can be seen, this does not appear to be an economically or statistically important concern, either for LS or FS establishments. This is also true
when we use the share of minimum wage workers rather than the 20th percentile of the wage
distribution.
As a final alternative, we estimated logit models that explore the relationship between
minimum wage changes and the probability of a price increase or decrease by outlet type.
These regressions use a very similar specification to equation (1), but substitute ∆ ln pijkt
with an indicator of whether there is a price increase (pijkt > 0) in one specification and a
price decrease (pijkt < 0) in another.30 The first four columns of table 3 report a specification
that includes the lag, contemporaneous, and lead minimum wage change measure, along with
the controls described in the table. We find that the likelihood of a price increase in LS outlets
jumps from 12 to 28 percent if a 10 percent increase in the minimum wage is introduced in a
28
There is some debate on the extent to which minimum wage increases impact workers paid above the
minimum. See Card and Krueger (1995), Abowd et al (2000), and Lee (1999). We have experimented with
using those at or below the new minimum wage, as well as allowing for spillovers up to 20 percent above the
new minimum. This has little impact on the results.
29
For identification, we must drop the PPI controls in this specification.
30
The main deviation from equation (1) is that we include a dummy for prices ending in 99 cents, to account
for the extra stickiness apparent at such price points. See MacDonald and Aaronson (2006) or Kashyap (1995)
for further discussion. In the version reported here, we also include three lags in changes in item prices, an
indicator of whether any sampled price was changed by the outlet in the previous period, and indicators for
meal type. Inferences are not contigent on the inclusion of these additional covariates.
16
period with otherwise stable prices. The probability of a price increase in FS establishments
increases from 10 to 13 percent following a similar sized minimum wage change. However,
a 10 percent increase in the minimum wage has no statistically significant impact on the
probability of a price decline in either type of establishment. The final two columns show
that MSAs with a lower 20th percentile wage are much more likely to see price increases
in both LS and FS establishments after a minimum wage increase. There is no such effect
among price declines (not shown).
17
Establishment type
Price change
Column
∆ ln wmin,it−1
Limited service
Increase
1
0.010
(0.014)
Full service
Increase
2
0.019
(0.011)
Limited service
Decrease
3
-0.007
(0.022)
Full service
Decrease
4
0.004
(0.018)
Limited Service
Increase
5
Full Service
Increase
6
∆ ln wmin,it
0.104
(0.012)
0.027
(0.011)
-0.020
(0.021)
0.008
(0.022)
0.253
(0.095)
0.233
(0.092)
∆ ln wmin,it+1
0.017
(0.015)
0.010
(0.012)
-0.029
(0.029)
0.014
(0.021)
-0.024
(0.015)
-0.032
(0.014)
wage20 ∗ ∆ ln wmin,it
18
Constant
-1.974
(0.074)
-2.150
(0.063)
-3.732
(0.137)
-4.403
(0.136)
-1.842
(0.444)
-2.485
(0.381)
Base Probability of price change
After 10% minimum wage increase
at wage20=$5
at wage20=$7
at wage20=$9
0.122
0.282
0.104
0.132
0.023
0.019
0.012
0.013
0.137
0.077
0.349
0.241
0.158
0.189
0.121
0.076
Coefficients and standard errors are derived from a logit model.
See text for detail. Controls not shown in table include whether the price ended in 99 cents,
three lags in ln pijkt , an indicator for whether any sampled price item was changed in the previous period,
and mealtype (breakfast, lunch, or dinner).
Table 3: The Probability of a Price Increase or Decrease in Response to a 10 Percent Minimum Wage Increase
3.3
City-level Price Responses
The micro data suggest that prices move higher in response to minimum wage changes
that occurred between 1995 and 1997. In this section, we show that the results are robust to
looking at a longer earlier period. Here, we use the publicly available city-level price indices
of the CPI between 1979 and 1995 to test whether cities with higher fractions of restaurant
workers impacted by the minimum wage laws are more likely to change their food prices.
Hence, identification is based on the intensity of minimum wage worker usage. Our results
are based on a slightly modified form of equation (1):
∆ ln pit
= γprob(wit = wmin,it ) + β ′ xit + uit
∆ ln wmin,it
(2)
∆ ln pit
where i denotes city and the coefficient γ = E[ ∆ ln
wmin,it |wit = wmin,it , xit ] is the price
response to increases in the minimum wage. If producers have a constant returns to scale
production function, competitive theory implies that 100 percent of the higher labor costs
are passed on to the consumer in the form of higher prices. As we pointed out earlier, 100%
pass through implies that the percent increase in product price equals the percent increase
in the minimum wage multiplied by labor’s share. Therefore, γ should equal labor’s share
under perfect competition.
Figure 1 maps each city’s price response to a minimum wage hike against the share of
minimum wage workers in the city. Each observation, of which there are 82, represents a
city around the time of a minimum wage change. The data cover 4 federal minimum wage
hikes - in 1980, 1981, 1990, and 1991 – and a small number of state increases between 1979
and 1995.31 The horizontal axis plots prob(wit = wmin,it ), the share of workers in a city’s
restaurant industry that are paid the minimum wage. This is computed from the outgoing
rotation files of the Current Population Survey (CPS). Because employees paid just above
the minimum wage are also affected by the law, we include anyone paid within 120 percent of
31
The federal minimum wage increased from $2.90 to $3.10 per hour in January 1980, to $3.35 in January
1981, to $3.80 in April 1990, and to $4.25 in April 1991. State increases tend not to occur in states represented
by CPI survey cities. The 27 CPI cities are New York City, Philadelphia, Boston, Pittsburgh, Buffalo,
Chicago, Detroit, St Louis, Cleveland, Minneapolis-St. Paul, Milwaukee, Cincinnati, Kansas City, DC, Dallas,
Baltimore, Houston, Atlanta, Miami, Los Angeles, San Francisco, Seattle, San Diego, Portland, Honolulu,
Anchorage, and Denver. After 1986, prices for 12 of these cities – Buffalo, Minneapolis-St. Paul, Milwaukee,
Cincinnati, Kansas City, Atlanta, San Diego, and Seattle, Portland, Honolulu, Anchorage, and Denver – are
reported semiannually. Therefore, we only include pre-1986 observations for these cities.
19
Change in log city food away from home price / change in log
minimum wage
1.2
1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Share of city restaurant workers that are within 20% of the old minimum wage during previous 9
months
Figure 1: City level price increases
20
1
the old minimum wage during the nine months prior to the minimum wage enactment date.
However, the results are not sensitive to picking reasonable thresholds other than 120 percent
or time frames other than nine months. The vertical axis displays
∆ ln pit
∆ ln wmin,it ,
the ratio of
the log change in city food away from home prices to the log change in the city’s minimum
wage. The price data is the CPI for food away from home. The price changes are measured
from two months before to two months after the minimum wage is enacted.
∆ ln pit
∆ ln wmin,it
is
adjusted for year fixed effects to account for inflation and other secular changes in national
labor market conditions.
The most noteworthy aspect of figure 1 is the positive correlation between the two series.
The regression coefficient γ is 0.36 with a robust, city clustered-corrected standard error of
0.24.32 Not only is the sign of this coefficient consistent with competition but the magnitude
is as well. Assuming perfect competition in the labor market, the regression coefficient should
equal labor’s share. Recall from section 3, labor’s share is approximately 30 to 35 percent.
Note also the abundance of observations on
∆ ln pit
∆ ln wmin,it
that are positive. Of the 82 city-
year observations, 19 are negative, including only 2 of the largest 30 price responses, defined
∆ ln pit
as when | ∆ ln
wmin,it | > 0.20. These two are interesting in that they come from the same city,
Denver, over consecutive years, 1980 and 1981. Unfortunately, we have little information as
to what was happening in Denver during this time but we can highlight it for being the main
example where city-level prices fall quickly in response to a minimum wage change.33
The most plausible alternative explanation for these price responses is that they are driven
by shocks to demand that happen to be correlated with changes to the minimum wage. We
tried two ways to test this possibility. First, we estimated equation 2 without year fixed effects
but included changes in the city CPI in the xit vector. The intercept from this specification
is not statistically different from zero, suggesting that prices do not rise after minimum wage
hikes in areas where the minimum wage does not bind. This finding is consistent with the
32
Using the share of minimum wage workers within 110 percent of the old minimum, rather than 120 percent,
the point estimate (and adjusted standard error) is 0.42 (0.24). A Huber biweight regression procedure implies
a point estimate of 0.28 (0.16) and 0.42 (0.15) using the 110 and 120 percent minimum wage share thresholds.
Finally, out of concern that inflation, even over this short period, are driving our results, we tried two things.
First, when we include city price deflators as controls on the right hand side, the point estimates are roughly
0.50 with t-statistics of roughly 2 to 2.5. Alternatively, we look at price changes only over the two months after
the minimum wage change. In this case, the point estimate is between 0.20 and 0.30, again with t-statistics
of roughly 2 to 2.5.
33
Since Denver is one of the 12 cities surveyed semiannually starting in 1986, we do not include the 1990
and 1991 Denver data points in the figure. However, they are both positive, albeit small: 0.19 for 1990 and
0.09 for 1991.
21
view that demand shocks are not confounding our estimates because if they were, we would
expect that prices would rise in areas where the minimum wage does not bind.
As a second check, we searched for alternative measures of pit that vary by local demand
conditions, are available for our city panel, and, most importantly, are relatively unaffected
by low wage labor costs. By far, the two best candidates are housing and medical care.
Therefore, we reran the regression described above, but substituted food away from home
prices with these two indices. As expected, we find no evidence that minimum wage hikes
are associated with price hikes for housing and medical care.34
Finally, we can compare the estimates in table 2 to predicted pass-through under competition and constant returns to scale technology using metro variation in restaurant wage
distributions from the outgoing rotation files of the CPS. Under competition, the relationship between these predicted price responses and the share of restaurant workers impacted
by minimum wage laws should correspond to labor share.
To conduct this test, we define two groups of cities. High wage cities are those with an
average hourly wage among the top fifth of all metropolitan areas in 1997-98. Low wage cities
are those with average wages among the bottom fifth of all metropolitan areas in 1997-98.
Among high (low) wage cities, 34 (59) percent of all restaurant workers are paid within 120
percent of the minimum wage, our rough measure of the share of workers impacted by such
laws. Based on the estimates in column (4) of table 2, the average predicted price response
for low and high wage cities is 0.097 and 0.090, respectively.35 The relationship between these
variables (slope of the line connecting high and low wage cities) is
.97−.90
.59−.34
= 0.28, slightly
lower than, and statistically indistinguishable from, observed labor share in the restaurant
industry. Furthermore, we get a similar labor share prediction when we estimate equation
(2) using the store-level data.36
34
For housing, γ (and its adjusted standard error) is -0.31 (0.33). For medical care, it is -0.06 (0.33).
The 20th percentile restaurant wage in the high wage cities is $5.25, compared to $4.82 in the low wage
cities.
36
To derive γ from the micro data, we use the regression results from column 4 of table 2, which gives the
∆ ln pit
relationship between ∆ ln
and the 20th percentile of city market wage. Next, we regress the share of
wmin,it
workers in a city’s restaurant industry that are paid the minimum wage, prob(wit = wmin,it ), on the 20th
percentile of that city’s market wage. Using the 27 major CPI cities during 1995 to 1997, the point estimate
from this latter regression is prob(wit = wmin,it ) = 0.839 − 0.068 × wage20it + νit , where wage20it is the 20th
percentile of city i’s wage distribution at time t. From these two regression equations, we can solve γ = 0.24.
Ideally we could precisely estimate equation (2) using the micro data. Unfortunately, as we note in footnote 25,
the number of observations in the CPS for individual cities can be small. Consequently, prob(wit = wmin,it )
cannot be precisely estimated. But wage20it can.
35
22
4
Theory
In this section, we show how our results contribute to the debate on the employment effects
of the minimum wage. Assorted models offer differing explanations for why the estimated
employment responses to minimum wage hikes are small. As we point out below, however,
most of these models imply that prices either do not change or fall in response to a minimum
wage hike. Therefore, our results provide evidence against models that have been used to
explain the small employment responses found in the minimum wage literature.
Throughout the discussion we assume that all firms are profit maximizers and thus set
the level of employment, L, at the point where the marginal cost of the last worker hired
M C(L) is equal to the extra revenue she produces (her marginal revenue product of labor,
or M RP (L)). Appendices A and B contain the formal details of the model.
4.1
The Competitive Model
We begin by briefly considering the textbook competitive model. If a minimum wage is
introduced (or increased) beyond the market-clearing wage in a competitive labor market,
the marginal cost of hiring a worker increases. In response, holding all else equal, firms will
move along their downward sloping marginal revenue product of labor curve until they reach
the point where M RP (L) is again equated to marginal cost. Higher M RP (L) can only be
obtained by reducing the workforce. Why? One important reason is that fewer workers imply
less output. Even if an additional worker produced the same amount as the previous worker,
reduced output increases output prices, marginal revenue, and thus the marginal revenue
product of labor. Therefore, minimum wage hikes cause prices to rise and employment to fall
in a competitive labor market environment.
4.2
The Textbook Monopsony Model
However, under monopsony, increasing the minimum wage can cause employment to rise.
The fundamental reason is tied to the link between the wage, the marginal cost of labor,
and the product price. In this section we describe the textbook model of monopsony, where
firms are monopolists in the product market and monopsonists in the labor market. In
Section 4.3 we show that the key results hold under the more realistic scenario of monopolistic
competition in the product market and monopsonistic competition in the labor market.
23
Wage
Marginal cost (M C(L))
A
Supply (w(L))
w∗∗
w∗
Marginal revenue product (M RP (L))
L∗
L∗∗
Employment
Figure 2: Illustration of monopsony equilibrium
Unlike the competitive firm, which pays the prevailing market wage regardless of how
much labor it demands, if the monopsonist wants to expand its labor force, it has to raise
the wage of its current workers as well.37 Therefore, the marginal cost of hiring a worker
is greater than the new worker’s wage. Figure 2 shows the wage the firm would have to
pay in order to attract an additional worker, w(L), and the marginal cost of hiring that
worker M C(L). When the monopsonist maximizes profits by setting marginal costs equal
to marginal product, shown at point A, total market employment, L∗ , is lower than the
“competitive case” (L∗∗ ) where employment is set based on the prevailing market wage w∗∗ .
A properly placed minimum wage, set somewhere between the wage
37
This is true if the monopsonist cannot perfectly discriminate between workers with high reservation wages
and low reservation wages.
24
Wage
Marginal cost (M C(L))
F
E
A
Supply (w(L))
wmin
B
C
D
w∗
Marginal revenue product (M RP (L))
L∗
Lmin
Employment
Figure 3: Illustration of monopsony equilibrium with minimum wage (Bold line
denotes M C(L) curve)
paid by a monopsonist (w∗ ) and the wage paid by a perfect competitor (w∗∗ ), will increase
employment. The intuition for this result, displayed in figure 3, is that although the minimum
wage increases the firm’s average cost of labor, it reduces the marginal cost of labor. Recall
that, absent a minimum wage, the marginal cost of hiring that last worker (at point A) lies
above the wage paid by the monopsonist (because everyone’s wage has to be raised in order to
induce a marginal worker to join the firm). If the minimum wage is set above the monopsony
equilibrium wage but below the marginal cost of hiring a worker, the new marginal cost
of hiring a worker falls from point A to point C (the new marginal cost of labor curve is
BCDEF ); the marginal cost of hiring additional workers is now just the minimum wage.
Because the firm must pay all workers at least the minimum wage, regardless of employment
25
level, the firm does not have to increase the pay of its existing workforce to attract more
employees (so long as employment is below Lmin ). This reduction in the marginal cost of
hiring additional labor causes firms to expand output and employment in response to the
minimum wage increase.
Moreover, employment and prices are negatively related because the fall in the marginal
cost of labor causes the marginal cost of producing an extra unit to fall. Consequently, the
product price falls as well.
It is important to note that if firms are monopsonists in the labor market but the minimum wage is set sufficiently high (above w∗∗ in Figure 2), employment is determined by the
intersection of the minimum wage and the marginal product of revenue curve. In this case, an
increase in the minimum wage increases prices and reduces employment, just like in a competitive labor market. Thus our empirical results cannot necessarily disprove the existence of
monopsony labor markets in cases where the minimum wage is set high (above competitive
market-clearing levels). However, we believe our empirical results should temper enthusiasm
for monopsony power being the explanation for the negligible employment responses found
in the literature.
4.3
Price and Employment Responses when there is Monopsonistic Competition in the Labor Market and Monopolistic Competition in the
Product Market
The results in the previous section were based on a very stylized model. In this section,
we show that those results are quite general under weak assumptions about technology,
the product market, and the labor market. Specifically, the results hold when firms have
a production function with substitutability between labor and other inputs, monopolistic
competition in the product market, and monopsonistic competition in the labor market.
Several researchers have argued that monopsonistic competition in the labor market is
the relevant case (Bhashkar and To (1999), Dickens et al. (1999)). That is, workers are
not indifferent between employers, even if all employers pay the same wage. One plausible
explanation is geography. Employers are located in different places and transportation costs
are large relative to earnings of minimum wage workers. Thus a worker is willing to take
a low paying job in order to be relatively near home. Alternatively, teenagers may want to
26
work at the same restaurant as their friends. More generally, certain aspects of one employer
may be disagreeable to some workers but not others.
In order to simplify the problem, we make six assumptions beyond the usual axioms of
firm behavior:
Assumption 1 There is a fixed number N of firms.
Assumption 2 All firms have an identical production function, Qn = Q(Kn , Ln ) where
Qn , Kn , Ln are output, an aggregator of non-labor input (that includes capital), and labor at
the nth firm.
Assumption 3 The production function is increasing in all inputs, concave, continuous and
twice differentiable.
Assumption 4 K and L are complementary inputs in the production function (Q12 > 0).
Assumption 5 The utility function of the representative consumer is U = (1 − α)Q01−η +
1
1
P
1−η
1−ηZ 1−ηZ
N
1−η
,
Q
, where Q0 is the numeraire good, α is close to zero, Q̃ ≡
αQ̃
n
n=1
and Qn denotes output at the nth restaurant. Concavity implies η > 0 and ηZ ∈ [0, 1).
Assumption 6 The firm is a price taker in the capital market and purchases Kn at price
r. However, the firm is potentially a monopsonist in the labor market. The quantity of labor
supplied to the firm is LSn = L(wn , w−n ), where w−n is the average wage paid by all other
firms,
dL(wn ,w−n )
dwn
> 0, and
w−n =wmin
dL(wn ,w−n )
dwn
> 0.
w−n =wn
Under these assumptions, firms sell their products at a price p(Q) and choose inputs to
maximize profits π :
πn (Kn , Ln ) = p(Qn )Qn − wn Ln − rKn .
(3)
These assumptions are standard, although a few require some elaboration. Assumption
5 gives rise to monopolistic competition in the product market. Markets are perfectly competitive if ηZ = 0 and firms operate as monopolists if ηZ = η. Assumption 6 states that the
quantity of labor supplied to the firm need not be perfectly elastic and, therefore, firms face a
27
monopsonistically competitive labor market. Consequently, firms face a weakly upward sloping inverse labor supply curve, w(Ln ), where
dw(Ln ,w−n )
dLn
≥ 0. However, because the minimum
wage potentially binds, the offered wage is wn = max{w(Ln , w−n ), wmin }.38
Firms may be price takers in the labor market because the labor supply curve that the
firm faces is perfectly elastic or the minimum wage is sufficiently high that it destroys the
firm’s monopsony power. Either way, if firms are price takers in the labor market, Theorem
1 holds.
Theorem 1 Given the assumptions above, and if firms are price takers in the labor market,
the industry level demand curve for labor slopes down.
Proof: see Appendix A. Theorem 1 is more general than the discussion in Section 4.1.
There, it is presumed that firms make employment decisions given a fixed M RP (L) curve,
an assumption that is appropriate for monopolists. But, under monopolistic competition,
minimum wage changes potentially shift the M RP (L) curve by altering the decision of other
firms, and thus influencing aggregate prices. Theorem 1 also differs from the Weak Axiom
of Profit Maximization, which assumes perfect competition in both the product and factor
markets.39
Section 4.2 discusses why employment can rise under monopsony. Given assumption 6,
this is true under monopolistic competition as well. Together with Theorem 1, this shows that
minimum wage hikes cause employment to fall under competition and rise under monopsony.
The next theorem shows that we can use prices to infer the importance of monopsony
power in the labor market.
Theorem 2 Given an increase in a binding minimum wage, prices rise under perfect competition and, so long as wmin < w∗∗ , prices fall under monopsony.
38
The assumption of capital and labor being complementary inputs (i.e. Q12 > 0) rules out situations where
the profit maximizing choice would be to switch from a capital intensive, high output technology to a labor
intensive, low output technology. An example of this is a firm that is capital efficient only up to a certain size.
After this size, capital cannot be efficiently used. For example, suppose p = 1, r = 1, Q = K .5 L.5 if L < 10 and
Q = L.5 if L ≥ 10. Increasing L from 9 to 10 would reduce output but depending on the cost of labor, could
increase profits. However, this rather extreme counter-example appears to go against the empirical evidence.
For example, it is difficult to reject the hypothesis that production functions are constant returns to scale,
and constant returns production functions implicitly assume Q12 > 0.
39
See Varian (1984) and Kennan (1998) for proofs of Theorem 1 under competition and monopoly, respectively. We have also proved the theorems in this section for the case where firms are Cournot competitors in
the output market. Proofs are available from the authors.
28
Proof: see Appendix A. The intuition for Theorem 2 was discussed in Section 4.2.
Finally, there is the quantitative importance of price pass through when there is monopolistic competition in the product market. Theorem 3 shows that if the production function
is constant elasticity of substitution, then firms still push 100% of the higher labor costs onto
consumers in the form of higher prices.
Theorem 3 If Q(., .) is a constant elasticity of substitution aggregator, and if firms are price
takers in the labor market, then
d ln p
d ln wmin
= minimum wage labor’s share.
Proof: see Aaronson and French (2007). The intuition for this result is straightforward.
Given the assumptions above, firms have a constant marginal cost and thus have a horizontal
supply curve. Thus, in a perfectly competitive market, all higher labor costs will be pushed
onto consumers in the form of higher prices. In the case of monopolistic competition in the
product market, there is a constant mark-up over marginal cost. Thus the supply curve is
still horizontal and all labor costs are pushed onto consumers in the form of higher prices.
Furthermore, Aaronson and French give predicted price and employment responses under
monopsonistic competition. They show that if the employment response is large and positive,
then the price response will be large and negative. For example, if the employment elasticity
is +0.2, which is possible under monopsony, then the price response will be -0.05. These price
responses vary notably from what is reported in Table 2.
The only assumption that we view as not innocuous is the first. Although the size of a
business is allowed to change in response to a higher minimum wage, firm exit or entry is
precluded. We think this is a reasonable assumption given the existing, albeit rather meager,
empirical evidence.40 Moreover, in this paper, we are interested in a short-term response that
likely severely limits entry and exit decisions.
The main reason for assuming no exit is that under monopsony, minimum wage hikes increase employment per restaurant, but likely reduce the total number of restaurants. There40
Card and Krueger (1995) and Machin and Wilson (2004) find no effect in the U.S. and U.K., respectively.
We have done some analysis of restaurant entry and exit using the Census’ Longitudinal Business Database.
Consistent with the literature, our preliminary findings suggest negligible entry and exit effects in the year
following a minimum wage change. These results stand in contrast to those of Campbell and Lapham (2004),
who find a significant amount of retail net entry along the U.S.-Canada border within a year of exchange rate
movements. We suspect these different results reflect the importance of exchange rates relative to minimum
wage levels in terms of firm costs.
29
fore, the industry level employment response is ambiguous. In this sense, we view the assumption of no exit as supporting the monopsony argument.41
4.4
Efficiency Wage Models
Efficiency wage models (where increased wages increase effort or reduce turnover costs),
often give monopsony like predictions. Manning (1995), Rebitzer and Taylor (1995), and
Deltas (1999) all present models where increases in the minimum wage can increase employment. None of these models allows for capital-labor substitutability, and only Deltas (1999)
allows for endogenous prices. Below we present an efficiency wage model with endogenous
prices and capital-labor substitutability.
We follow Solow (1979), who argues that the wage affects morale and effort amongst
other things, and let the wage enter the production function directly. If an increase in the
wage causes increased effort, more meals can be produced with the same amount of labor and
capital. Therefore, it is not necessarily more costly to produce meals when the minimum wage
increases. This can attenuate the disemployment effects of the minimum wage. However, if
employment does not fall (and other factors do not fall either) and productivity rises, total
output will rise and product prices will fall.
Let the production function be:
Q = Q(K, L, w) = (1 − α)K ρ + α(Lwθ )ρ
where σ ≡
1
1−ρ
1
ρ
,
(4)
is the partial elasticity of substitution between K and Lwθ in the production
of Q, and Lwθ is “effective labor”. The parameter 0 ≤ θ < 1 may be greater than 0
because increases in the wage increase effort, which could happen for a variety of reasons.
Furthermore, assume that Assumptions 1 to 5 in Section 4.3 hold. Then the price and
41
Nevertheless, we also understand that, given our assumed market structure, entry and exit can change
the market price for a given industry output. This is potentially important because Bashkar and To (1999)
argue that an increase in the minimum wage could reduce the number of firms in the market but increase
employment per restaurant, causing total employment to potentially increase. Because the number of firms
decrease, market power of survivors increase. Consequently, both prices and output may increase. However,
given the small observed exit rates in response to minimum wage hikes, we doubt that these effects would be
large enough to overturn the basic presumption that market output and price move in opposite directions.
30
employment responses are:
d ln p
= s(1 − θ)
d ln w
d ln L
= −(1 − θ) (1 − s)σ − sη − θ
d ln w
(5)
(6)
where s is the share of total costs going to labor. When θ = 0, equations (5) and (6) give
the textbook response to the minimum wage. However, when θ > 0 (and all else is equal),
a one percent increase in the wage increases effective labor θ percent, causing the marginal
cost of effective labor to only rise 1 − θ percent. This has implications for both prices and
employment responses. The price response is attenuated, rising by s(1−θ) percent, a fraction
(1 − θ) less than without endogenous work effort. The employment response is also muted
by (1 − θ) due to an increase in the marginal cost of labor. However, the same amount of
effective labor and output can be produced with fewer bodies, lessening the need for labor
(the −θ term at the end of equation (6)).
Regardless, while a smaller employment response relative to the case without endogenous
work effort is possible, the model clearly predicts a smaller price response as well. However,
our estimates indicate large price responses to the minimum wage. Thus, our price results
provide evidence against the hypothesis that endogenous work effort is responsible for the
small observed employment responses to minimum wage hikes.
4.5
Other Models of the Employment Effects of the Minimum Wage
We have argued that the price responses to minimum wage hikes are useful for distinguishing between competition and monopsony in the labor market. Likewise, our estimated
price responses help shed light on other explanations of the small employment response found
in the minimum wage literature.
Some researchers (e.g. Kennan 1995, MaCurdy and OBrien-Strain 2000) suggest higher
income resulting from a minimum wage increase causes low wage workers to buy more minimum wage products, attenuating the disemployment effect of the minimum wage. For the
restaurant industry, Kennan refers to this possibility as the ”hungry teenager hypothesis.”
The price response reported in this paper is a key parameter for such a calculation. Aaronson
31
and French (2006) write down a model that allows for such demand-induced feedbacks and
show that increases in the minimum wage reduce real income for non-minimum wage workers
(because prices rise) but increase real income of minimum wage workers (because their wage
rises). They find that if minimum wage workers spend a large fraction of their income on fast
food, then the rise in incomes for fast food workers can at least partly offset the disemployment effect of the minimum wage. Using data from the Consumer Expenditure Survey and
US Department of Agriculture’s Continuing Survey of Food Intake by Individuals, they find
that minimum wage workers spend between 20 and 100% more of their income on fast food
as those who are not minimum wage workers.42 Given these estimates (and other calibrated
parameters) the increased income going to minimum wage workers can offset between 25 and
40% of the output loss, and 10 to 30% of the employment loss of a model that does not
account for the increased income and increased expenditures of minimum wage workers.
Our price responses are less consistent with the idea that that minimum wage laws merely
cause firms to reshuffle compensation packages from non-wage benefits to wages. For example, fast food restaurants could stop giving workers free meals after minimum wage hikes.
Hashimoto (1982) and Neumark and Wascher (2001) argue that firms may reduce training
after minimum wage hikes. As a result, there is no increase in the cost of labor faced by firms.
However, if the minimum wage does not increase the cost of labor, it is unclear why there
are price increases after minimum wage changes. Although shifting compensation packages
from non-wage to wage benefits may occur, our results indicate that firms still bear a sizeable
fraction of the cost of minimum wage hikes. In this sense, our findings are consistent with
Card and Krueger (1995), who also find very little substitution between wage and non-wage
benefits after minimum wage hikes and also find no evidence of minimum wage hikes on
training.
Finally, our analysis has not considered search models (e.g. Burdett and Mortensen
1998), which also give monopsony implications for employment. A full analysis of the variety
of search models is well beyond the scope of this paper, especially since most of them would
need to be modified to account for endogenous product prices and capital labor substituability. However, it seems likely that employment and prices would move in opposite directions
in most standard applications of search. For example, given that the Burdett and Mortensen
42
MaCurdy and OBrien-Strain (2000) also find that low income individuals spend a greater fraction of their
budget on products produced by minimum wage workers.
32
production technology is linear in labor (i.e. no substitution among inputs), increases in
employment increase output and should presumably reduce the product price were it endogenized. It is also worth noting that most estimated search models, including Van den Berg
and Ridder (1998), Flinn (2006), Ahn and Arcidiacono (2003), find some disemployment in
response to a minimum wage increase.
Arguably, our empirical results themselves - that marginal cost shocks are passed onto
consumers through higher output prices - may be consistent with small disemployment effects.
If restaurants face factors that limit their ability to raise prices, say because it is costly to
switch prices, or because the price elasticity of demand for food away from home is infinitely
elastic, the predicted disemployment effects of a minimum wage increase would be larger
than if these factors did not hinder price behavior. If firms cannot pass cost increases onto
consumers, then profits will be squeezed and firms may sharply cut their workforce. Given
that we find rather large price increases in response to minimum wage hikes, firms seem to
be able to push costs onto consumers, and are not having their profits greatly reduced.
Instead, we interpret our results to be consistent with the moderate disemployment effects
reported in Aaronson and French (2007). They calibrate a structural model of labor demand
that incorporates the price responses found here to show that a 10 percent increase in the
minimum wage reduces restaurant employment by 2 to 3 percent, a short run response that
is within the range of estimates found in the literature. Moreover, the total (low plus high
skill) restaurant employment response may be as small as 1 percent.
5
Conclusion
Much work has looked at the employment implications of raising the minimum wage,
with a range of estimates reported in the literature. We offer new empirical evidence using
output prices both at the store-level and aggregated to the city-level. In both cases, prices
unambiguously increase in response to a minimum wage change. Furthermore, the results are
similar across three sources of variation in the data: cross-state differences in the size of the
minimum wage change, cross-restaurant type differences in the tendency to pay at or near
the minimum wage, and cross-metro differences in the fraction of workers paid at or near
the minimum wage. There is no evidence that prices fall in response to a minimum wage
increase.
33
We interpret these findings within a simple yet quite general model of employment determination that shows that monopsony and perfect competition have opposite implications for
not only employment but output prices as well, so long as the minimum wage is not set too
high. In particular, under monopsony, an increase in a binding minimum wage causes employment to rise and output prices to fall. Under competition, employment falls and output
prices rise. Therefore, our price results appear to provide evidence against the hypothesis that
monopsony power is important for understanding the small observed employment response
to minimum wage changes. Indeed, our estimated price responses provide evidence against
other explanations of the small employment response, including the potential substitution of
nonwage for wage compensation and the importance of endogenous work effort. Rather, we
interpret our price results to be fairly consistent with the textbook model of labor demand.
34
References
[1] Aaronson, Daniel “Price Pass-Through and the Minimum Wage,” Review of Economics and Statistics, February 2001, 83(1), 158-169.
[2] Aaronson, Daniel and Eric French “Product Market Evidence on the Employment
Effects of the Minimum Wage,” Journal of Labor Economics, January 2007, 25(1), 167200.
[3] Aaronson, Daniel and Eric French “Can Income Effects Explain the Small Observed
Disemployment Reponse to Minimum Wage Changes ,” working paper, Federal Reserve
Bank of Chicago, 2006.
[4] Abowd, John, Francis Kramarz, David Margolis, and Thomas Philippon,
“The Tale of Two Countries: Minimum Wages and Employment in France and the
United States, working paper, INSEE-CREST, 2000.
[5] Ahn, Tom and Peter Arcidiacono “Minimum Wages and Positive Employment
Effects in General Equilibrium,” working paper, Duke University, 2003.
[6] Ashenfelter, Orley, and Stepan Jurajda, “Cross-country Comparisons of Wage
Rates: The Big Mac Index,” working paper, Princeton University, 2001.
[7] Bhaskar, V. and Ted To, “Minimum Wages for Ronald McDonald Monopsonies: A
Theory of Monopsonistic Competition,” Economic Journal, April 1999, 109(455), 190203.
[8] Bils, Mark and Peter Klenow, “Some Evidence on the Importance of Sticky Prices,”
Journal of Political Economy, October 2004, 112(5), 947-985.
[9] Boal, William and Michael Ransom, “Monopsony in the Labor Market,” Journal
of Economic Literature, March 1997, 35(1), 86-112.
[10] Brown, Charles, “Minimum Wages, Employment, and the Distribution of Income,”
in Ashenfelter and Card, eds. Handbook of Labor Economics, Vol 3B, 1999, Amsterdam:
North-Holland, 2101-2163.
35
[11] Brown, Charles, Curtis Gilroy, and Andrew Kohen, “The Effect of the Minimum
Wage on Employment and Unemployment,” Journal of Economic Literature, June 1982,
20(2), 487-528.
[12] Burdett, Kenneth, and Dale Mortensen, “Wage Differentials, Employer Size, and
Unemployment,” International Economic Review, May 1998, 39(2), 257-273.
[13] Burkhauser, Richard, Kenneth Couch, and David Wittenburg, “A Reassessment of the New Economics of the Minimum Wage Literature with Monthly Data from
the Current Population Survey,” Journal of Labor Economics, October 2000, 18(4), 653680.
[14] Campbell, Jeffrey and Beverly Lapham “Real Exchange Rate Fluctuations and the
Dynamics of Retail Trade Industries on the U.S.-Canada Border,” American Economic
Review , September 2004, 94(4), 1194-1206.
[15] Card, David and Alan Krueger, Myth and Measurement: The New Economics of
the Minimum Wage, Princeton, NJ: Princeton University Press, 1995.
[16] Card, David and Alan Krueger, “Minimum Wages and Employment: A Case Study
of the Fast-Food Industry in New Jersey and Pennsylvania: Reply,” American Economic
Review, December 2000, 90(5), 1397-1420.
[17] Chevalier, Judith, Anil Kashyap, and Peter Rossi, “Why Don’t Prices Rise During Periods of Peak Demand? Evidence from Scanner Data,” American Economic Review, March 2003, 93(1), 15-37.
[18] Converse, Muriel, Richard Coe, Mary Corcoran, Maureen Kallick, and James
Morgan, “The Minimum Wage: An Employer Survey,” in Report of the Minimum Wage
Commission, 1981, Washington DC: US Government Printing Office, 241-341.
[19] Deere, Donald, Kevin M. Murphy, and Finis Welch, “Employment and the 19901991 Minimum-Wage Hike,” American Economic Review, May 1995, 85(2), 232-237.
[20] Deloitte and Touch, 2003 Restaurant Industry Operations Report, Washington, DC:
National Restaurant Association, 2003.
36
[21] Dickens, Richard, Stephen Machin, and Alan Manning, “The Effects of Minimum Wages on Employment: Theory and Evidence from Britain,” Journal of Labor
Economics, January 1999, 17(1), 1-22.
[22] Dixit, A. and J. Stiglitz, “Monopolistic Competition and Optimum Product Diversity,” American Economic Review, June 1977, 67(3), 297-308.
[23] Flinn, Christopher, “Minimum Wage Effects on Labor Market Outcomes under
Search, Bargaining, and Endogenous Contact Rates,” Econometrica, forthcoming, 2006.
[24] Fuchs, Victor, Alan Krueger, and James Poterba, “Economists’ Views about
Parameters, Values, and Policies: Survey Results in Labor and Public Economics,”
Journal of Economic Literature, September 1998, 36(3), 1387-1425.
[25] Hashimoto, Masanori, “Minimum Wage Effects on Training on the Job,” American
Economic Review , December 1982, 72(5), pp. 1070-1087.
[26] Kashyap, Anil, “Sticky Prices: New Evidence from Retail Catalogs,” Quarterly Journal of Economics, February 1995, 110(1), pp. 245-274.
[27] Katz, Lawrence and Alan Krueger, “The Effect of the Minimum Wage on the Fast
Food Industry,” Industrial and Labor Relations Review, October 1992, 46(1), 6-21.
[28] Kennan, John, “The Elusive Effects of Minimum Wages,” Journal of Economic Literature, December 1995, 33(4), 1950-1965.
[29] Kennan, John, “Minimum Wage Regulation,” in Newman, ed. The New Palgrave
Dictionary of Economics and the Law, 1998, New York: Norton.
[30] Kim, Taeil and Lowell Taylor, “The Employment Effect in Retail Trade of California’s 1988 Minimum Wage Increase,” Journal of Business and Economic Statistics,
April 1995, 13(2), 175-182.
[31] Lee, David, “Wage Inequality in the United States During the 1980s: Rising Dispersion
or Falling Minimum Wage?,” Quarterly Journal of Economics, August 1999, 114(3), 9771023.
[32] Machin, Stephen and Joan Wilson, “Minimum Wages in a Low-Wage Labour Market: Care Homes in the UK,” The Economic Journal, March 2004, 114(494), C102-C109.
37
[33] MacDonald, James and Daniel Aaronson, “How Firms Construct Price Changes:
Evidence from Restaurant Responses to Increased Minimum Wages,” American Journal
of Agricultural Economics, May 2006, 88(2), 292-307.
[34] MaCurdy, Thomas and Margaret OBrien-Strain, “Increasing the Minimum Wage:
California’s Winners and Losers,” working paper, Public Policy Institute of California,
2000.
[35] Manning, Alan, “How Do We Know That Real Wages Are Too High?,” Quarterly
Journal of Economics, November 1995, 110(4), 1111-1125.
[36] Manning, Alan,
Monopsony in Motion: Imperfect Competition in Labor Markets,
Princeton, NJ: Princeton University Press, 2003.
[37] Neumark, David and William Wascher, “Reconciling the Evidence on Employment
Effects of Minimum Wages: A Review of Our Research Findings,” in Kosters, ed. The
Effects of Minimum Wages on Employment, 1996, Washington, DC: American Enterprise
Institute, 55-86.
[38] Neumark, David and William Wascher, “Minimum Wages and Employment: A
Case Study of the Fast-Food Industry in New Jersey and Pennsylvania: Comment,”
American Economic Review, December 2000, 90(5), 1362-1396.
[39] Neumark, David and William Wascher, “ Minimum Wages and Training Revisited,” Journal of Labor Economics, July 2001, 19(3), 563-595.
[40] Rebitzer, James and Lowell Taylor, “The Consequences of Minimum Wage Laws:
Some New Theoretical Ideas,” Journal of Public Economics, February 1995, 56(2), 245255.
[41] Rubin, Alissa, “Popular Minimum Wage Hike Gets Solid Senate Approval,” Congressional Quarterly, July 13, 1996, 1964-1965.
[42] Solow, Robert, “Another Possible Source of Wage Stickiness,” Journal of Macroeconomics, 1979, 1(1), 79-82.
[43] Stigler, George, “The Economics of Minimum Wage Legislation,” American Economic
Review, June 1946, 36(3), 358-365.
38
[44] van den Berg, Gerard and Geert Ridder, “An Empirical Equilibrium Search Model
of the Labor Market,” Econometrica, September 1998, 66(5), 1183-1223.
[45] Varian, Hal, “The Nonparametric Approach to Production Analysis,” Econometrica,
May 1984, 52(3), 579-597.
[46] Weisman, Jonathan, “Republican Defectors Help Propel Minimum Wage Bill to Passage,” Congressional Quarterly, May 25, 1996, 1461-1464.
39
Appendix A: Proofs
Proof of Theorm 1
Dixit and Stiglitz (1977) point out that two-stage budgeting techniques can be used
to analyze the consumer demand problem described in the text. In the second stage the
consumer solves:
max
{Qn }N
n=1
N
X
Z
Q1−η
n
n=1
1
1−ηZ
(7)
subject to
N
X
pn Qn = X,
(8)
n=1
where X is total expenditure on Q̃. The consumer’s first order condition for utility maximization yields
pn = λ−1
Qn
Q̃
!−ηZ
(9)
where λ is the Lagrange multiplier on the budget constraint for expenditure on Q̃, i.e., it is
the multiplier on the first stage constraint. If N is sufficiently large, then Qn is small relative
to Q̃ and thus the firm does not take into account the effect of Qn on Q̃ when assessing the
effect of Qn on pn .
Inserting equation (9) into equation (3) yields
Z
π(Kn , Ln ) = ΩQ1−η
− wn Ln − rKn
n
(10)
where Ω = λ−1 Q̃ηZ . The first order condition for maximization of the profit function (10)
with respect to Kn yields
Z
Ω(1 − ηZ )Q−η
Qn,1 = r
n
where
∂Qn
∂Kn
(11)
≡ Qn,1 . If the firm is a price taker in the labor market, then employment is
40
determined by maximization of equation (10) with respect to Ln , which yields
Z
Ω(1 − ηZ )Q−η
Qn,2 = w.
n
where
∂Qn
∂Ln
(12)
≡ Qn,2 . First, we discuss the case where the firm is a price taker in the labor
1
market. Given the assumptions above, all firms are of equal size, so Q̃ = N 1−ηZ Qn . Because
of this, equation (9) shows that
d ln λ
d ln p̃
=−
,
d ln w
d ln w
where the price index associated with Q̃ is p̃ ≡
and Assumption 5, one can show that
d ln Q̃
d ln p̃
(13)
ηZ −1
P
ηZ
N
n=1 pn
= −η.
ηZ
ηZ −1
. Using these definitions
Because price changes are responses to supply shocks, price changes identify the demand
curve for the aggregate product, so
d ln Q̃
d ln p̃
= −η ∗ for some η ∗ ∈ (0, η], where η ∗ = η if the
production technology is constant returns. Therefore, we may write
d ln p̃
=
d ln w
where we are assuming that
d ln N
d ln w
d ln Q̃
d ln w
d ln Q̃
d ln p̃
1
=− ∗
η
d ln Qn
d ln w
!
(14)
= 0. Taking logs and differentiating equations (11) and
(12) with respect to ln w and inserting equations (13) and (14) yields
An
dKn
dLn
+ Cn
=0
dw
dw
(15)
Bn
dKn
dLn
+ Dn
=1
dw
dw
(16)
and differentiating equation (11) and (12) with respect to ln r and inserting equations analogous to (13) and (14) yields
An
dLn
dKn
+ Cn
=1
dr
dr
(17)
Bn
dLn
dKn
+ Dn
=0
dr
dr
(18)
41
where
An ≡ (−
wQn,1 wQn,11
+
)
η ∗ Qn
Qn,1
(19)
Bn ≡ (−
wQn,1 wQn,12
+
)
η ∗ Qn
Qn,2
(20)
Cn ≡ (−
wQn,2 wQn,12
+
)
η ∗ Qn
Qn,1
(21)
Dn ≡ (−
wQn,1 wQn,22
+
).
η ∗ Qn
Qn,2
(22)
In matrix form this can be rewritten as
An
Cn
Bn Dn
dKn
dr
dKn
dw
dLn
dr
dLn
dw
=
1 0
0 1
which can be rewritten as
dKn
dr
dKn
dw
dLn
dr
dLn
dw
Dn −Cn
1
=
An Dn − Bn Cn
−Bn An
where solving for An Dn − Bn Cn shows that the term is positive if the production function is
concave.
Solving for
dKn
dw
and
dLn
dw
yields
An
dLn
=
dw
An Dn − Bn Cn
(23)
−Cn
dKn
=
.
dw
An Dn − Bn Cn
(24)
The term An is negative. Therefore, the quantity of labor demanded falls when the price
rises. The market employment change is N times the firm level employment change.
42
We consider the monopsony case next. Because both the production and utility functions
are smooth, small changes in the minimum wage result in small changes in the M RP (Ln ),
denoted in equation (12). Therefore, employment is determined by labor supply, not labor
demand. Figure 3 clarifies this point. Given the assumed upward sloping labor supply curve,
employment rises in response to an increase in the minimum wage.
QED.
Proof of Theorem 2
Consider the case of perfectly competitive labor markets first. Define p, Q and L as the
market price, output and employment. Because all firms produce the same amount, p = pn ,
Q = N Qn and L = N Ln for all n. The price response to the wage change is
dp dQ
dp
=
dw
dQ dw
(25)
where a Taylor’s series expansion shows
!
dQ
dKn
dLn
.
= N Qn,1
+ Qn,2
dw
dw
dw
(26)
Inserting equations (23) and (24) into equation (26) yields
dQ
=N
dw
!
−Qn,1 Cn + Qn,2 An
.
An Dn − Bn Cn
(27)
The numerator of this expression is negative and the denominator is positive if Qn,12 > 0.
Therefore,
dQ
dw
< 0. Since
dp
dQ
< 0 by assumption,
dp
dw
> 0 under perfectly competitive labor
markets.
Now consider the monopsony case. Recall that under the monopsony case, labor is determined by labor supply, not labor demand. Therefore, we can take labor supply as exogenous.
Therefore, we may rewrite equation (25) as:
dp dQ
dp
dp
=
=
dw
dQ dw
dQ
Since by assumption,
dp
dQ
< 0 and
dL
dw
!
dQ dL
.
dL dw
> 0, all that remains is to prove that
is taken as exogenous in order to prove that
dp
dw
43
< 0 under monopsony.
(28)
dQ
dL
> 0 when dL
Taking the log of equation (11) and differentiating with respect to ln L yields
d ln Ω d ln Qn,1
+
=0
d ln L
d ln L
Because
d ln λ
d ln L
(29)
ln p̃
= − dd ln
L and for reasons discussed in the proof of Theorem 1, equation (29)
may be rewritten as
d ln Qn,1
d ln p̃
+
=0
d ln L
d ln L
Using
d ln p̃
d ln L
=
d ln Q̃
d ln L
d ln Q̃
d ln p̃
(30)
we may rewrite equation (30) as
−1
η∗
for some η ∗ ∈ (0, η], where
d ln N
d ln L
d ln Qn d ln N
+
d ln L
d ln L
!
+
d ln Qn,1
=0
d ln L
(31)
= 0. Using the chain rule,
dKn
dQn
= Qn,1
+ Qn,2 ,
dL
dL
(32)
dQn,1
dKn
= Qn,11
+ Qn,12 .
dL
dL
(33)
Inserting equations (32) and (33) into (31) yields
L
− ∗
η Qn
Solving for
dKn
dL
!
dKn
Qn,1
+ Qn,2
dL
!
+
L
Qn,1
!
dKn
Qn,11
+ Qn,12
dL
!
=0
(34)
and inserting this into equation (32) yields
1
1
dQn
η ∗ Qn Qn,2 − Qn,1 Qn,12
= Qn,1
1
dL
− η∗1Qn Qn,1 + Qn,1
Qn,11
where the term in parentheses is
dKn
dL .
!
+ Qn,2
(35)
The denominator of this term is negative. The first
term in the numerator is positive but the second term is non-positive, making the sign of the
numerator term in parentheses ambiguous. However,
44
dQn
dL
is positive. To see this, note that
| −
Qn,1
η ∗ Qn |
<| −
dQn
> Qn,1
dL
Qn,1
η ∗ Qn
−
+
Qn,2
η ∗ Qn
Qn,1
η ∗ Qn
Qn,11
Qn,1 |.
!
Therefore, |
Qn,2
η ∗ Qn
Q
− η∗n,1
Qn
Q
+ Qn,1
− Qn,12
n,1
Q
− η∗n,1
Qn +
Qn,11
Qn,1
!
!
|>|
!
| and thus
Q
+ Qn,2 = Qn,1
which is greater than 0 if Qn,12 is greater than 0. Therefore,
dp
dw
Qn,2
η ∗ Qn
Q
Q
+ Qn,11
− η∗n,1
Qn
n,1
dQn
dL
− Qn,12
n,1
Q
− η∗n,1
Qn +
Qn,11
Qn,1
> 0 and thus
dQ
dL
!
(36)
> 0 and
< 0 under monopsony. QED.
Appendix B: Price and Employment Responses in Models with Endogenous
Work Effort
In this appendix we assume that firms are price takers in all markets (so ηZ = 0), although
using the results in Aaronson and French (2004) it is straightforward to account for the case
where ηZ > 0. That paper provides a similar and more detailed derivation than the one
below, although they assume θ = 0.
Maximization of equation (3) with respect to K and L, and taking logs, yields:
σ(ln r − ln p) = σ ln(1 − α) + ln Q − ln K ,
σ(ln w − ln p) = σ ln α + ln Q − ln L + σρθ ln w.
(37)
(38)
Assume that product prices potentially change in response to a change in the minimum wage,
but the price of capital does not. Differentiating equations (37) and (38) with respect to ln w,
and using results in Aaronson and French (2007), yields:
d ln Q d ln K
1 d ln Q
=−
+
,
η d ln w
d ln w
d ln w
σ
d ln L
d ln Q
1−
+ (1 − σ)θ + σ,
−
=−
d ln w
d ln w
η
−σ
(39)
(40)
where η is the elasticity of demand for the product (in absolute value), as in the text. Furthermore, differentiating the production function with respect to the wage yields:
dQ
∂Q dK
∂Q dL ∂Q
=
+
+
.
dw
∂K dw
∂L dw ∂w
Solving for
∂Q
∂w
as a function of
∂Q
∂L ,
noting that
45
∂Q
∂L
=
w ∂Q
p , ∂K
(41)
= pr , and denoting s =
wL
pQ
as
labor’s share, some algebra shows that equation (41) can be rewritten as:
d ln K
d ln L
d ln Q
= (1 − s)
+s
+ sθ
d ln w
d ln w
d ln w
(42)
Using equations (39), (40), and (42), we can solve for the unknowns:
d ln Q
= −ηs(1 − θ),
d ln w
(43)
d ln K
= s(1 − θ)(σ − η),
d ln w
(44)
d ln L
= −(1 − θ) (1 − s)σ − sη − θ.
d ln w
(45)
The price response will be
d ln p
= s(1 − θ).
d ln w
46
(46)
Working Paper Series
A series of research studies on regional economic issues relating to the Seventh Federal
Reserve District, and on financial and economic topics.
Does Bank Concentration Lead to Concentration in Industrial Sectors?
Nicola Cetorelli
WP-01-01
On the Fiscal Implications of Twin Crises
Craig Burnside, Martin Eichenbaum and Sergio Rebelo
WP-01-02
Sub-Debt Yield Spreads as Bank Risk Measures
Douglas D. Evanoff and Larry D. Wall
WP-01-03
Productivity Growth in the 1990s: Technology, Utilization, or Adjustment?
Susanto Basu, John G. Fernald and Matthew D. Shapiro
WP-01-04
Do Regulators Search for the Quiet Life? The Relationship Between Regulators and
The Regulated in Banking
Richard J. Rosen
Learning-by-Doing, Scale Efficiencies, and Financial Performance at Internet-Only Banks
Robert DeYoung
The Role of Real Wages, Productivity, and Fiscal Policy in Germany’s
Great Depression 1928-37
Jonas D. M. Fisher and Andreas Hornstein
WP-01-05
WP-01-06
WP-01-07
Nominal Rigidities and the Dynamic Effects of a Shock to Monetary Policy
Lawrence J. Christiano, Martin Eichenbaum and Charles L. Evans
WP-01-08
Outsourcing Business Service and the Scope of Local Markets
Yukako Ono
WP-01-09
The Effect of Market Size Structure on Competition: The Case of Small Business Lending
Allen N. Berger, Richard J. Rosen and Gregory F. Udell
WP-01-10
Deregulation, the Internet, and the Competitive Viability of Large Banks
and Community Banks
Robert DeYoung and William C. Hunter
WP-01-11
Price Ceilings as Focal Points for Tacit Collusion: Evidence from Credit Cards
Christopher R. Knittel and Victor Stango
WP-01-12
Gaps and Triangles
Bernardino Adão, Isabel Correia and Pedro Teles
WP-01-13
A Real Explanation for Heterogeneous Investment Dynamics
Jonas D.M. Fisher
WP-01-14
Recovering Risk Aversion from Options
Robert R. Bliss and Nikolaos Panigirtzoglou
WP-01-15
Economic Determinants of the Nominal Treasury Yield Curve
Charles L. Evans and David Marshall
WP-01-16
1
Working Paper Series (continued)
Price Level Uniformity in a Random Matching Model with Perfectly Patient Traders
Edward J. Green and Ruilin Zhou
WP-01-17
Earnings Mobility in the US: A New Look at Intergenerational Inequality
Bhashkar Mazumder
The Effects of Health Insurance and Self-Insurance on Retirement Behavior
Eric French and John Bailey Jones
WP-01-18
The Effect of Part-Time Work on Wages: Evidence from the Social Security Rules
Daniel Aaronson and Eric French
WP-01-20
Antidumping Policy Under Imperfect Competition
Meredith A. Crowley
WP-01-21
WP-01-19
Is the United States an Optimum Currency Area?
An Empirical Analysis of Regional Business Cycles
Michael A. Kouparitsas
WP-01-22
A Note on the Estimation of Linear Regression Models with Heteroskedastic
Measurement Errors
Daniel G. Sullivan
WP-01-23
The Mis-Measurement of Permanent Earnings: New Evidence from Social
Security Earnings Data
Bhashkar Mazumder
WP-01-24
Pricing IPOs of Mutual Thrift Conversions: The Joint Effect of Regulation
and Market Discipline
Elijah Brewer III, Douglas D. Evanoff and Jacky So
WP-01-25
Opportunity Cost and Prudentiality: An Analysis of Collateral Decisions in
Bilateral and Multilateral Settings
Herbert L. Baer, Virginia G. France and James T. Moser
WP-01-26
Outsourcing Business Services and the Role of Central Administrative Offices
Yukako Ono
WP-02-01
Strategic Responses to Regulatory Threat in the Credit Card Market*
Victor Stango
WP-02-02
The Optimal Mix of Taxes on Money, Consumption and Income
Fiorella De Fiore and Pedro Teles
WP-02-03
Expectation Traps and Monetary Policy
Stefania Albanesi, V. V. Chari and Lawrence J. Christiano
WP-02-04
Monetary Policy in a Financial Crisis
Lawrence J. Christiano, Christopher Gust and Jorge Roldos
WP-02-05
Regulatory Incentives and Consolidation: The Case of Commercial Bank Mergers
and the Community Reinvestment Act
Raphael Bostic, Hamid Mehran, Anna Paulson and Marc Saidenberg
WP-02-06
2
Working Paper Series (continued)
Technological Progress and the Geographic Expansion of the Banking Industry
Allen N. Berger and Robert DeYoung
WP-02-07
Choosing the Right Parents: Changes in the Intergenerational Transmission
of Inequality Between 1980 and the Early 1990s
David I. Levine and Bhashkar Mazumder
WP-02-08
The Immediacy Implications of Exchange Organization
James T. Moser
WP-02-09
Maternal Employment and Overweight Children
Patricia M. Anderson, Kristin F. Butcher and Phillip B. Levine
WP-02-10
The Costs and Benefits of Moral Suasion: Evidence from the Rescue of
Long-Term Capital Management
Craig Furfine
WP-02-11
On the Cyclical Behavior of Employment, Unemployment and Labor Force Participation
Marcelo Veracierto
WP-02-12
Do Safeguard Tariffs and Antidumping Duties Open or Close Technology Gaps?
Meredith A. Crowley
WP-02-13
Technology Shocks Matter
Jonas D. M. Fisher
WP-02-14
Money as a Mechanism in a Bewley Economy
Edward J. Green and Ruilin Zhou
WP-02-15
Optimal Fiscal and Monetary Policy: Equivalence Results
Isabel Correia, Juan Pablo Nicolini and Pedro Teles
WP-02-16
Real Exchange Rate Fluctuations and the Dynamics of Retail Trade Industries
on the U.S.-Canada Border
Jeffrey R. Campbell and Beverly Lapham
WP-02-17
Bank Procyclicality, Credit Crunches, and Asymmetric Monetary Policy Effects:
A Unifying Model
Robert R. Bliss and George G. Kaufman
WP-02-18
Location of Headquarter Growth During the 90s
Thomas H. Klier
WP-02-19
The Value of Banking Relationships During a Financial Crisis:
Evidence from Failures of Japanese Banks
Elijah Brewer III, Hesna Genay, William Curt Hunter and George G. Kaufman
WP-02-20
On the Distribution and Dynamics of Health Costs
Eric French and John Bailey Jones
WP-02-21
The Effects of Progressive Taxation on Labor Supply when Hours and Wages are
Jointly Determined
Daniel Aaronson and Eric French
WP-02-22
3
Working Paper Series (continued)
Inter-industry Contagion and the Competitive Effects of Financial Distress Announcements:
Evidence from Commercial Banks and Life Insurance Companies
Elijah Brewer III and William E. Jackson III
WP-02-23
State-Contingent Bank Regulation With Unobserved Action and
Unobserved Characteristics
David A. Marshall and Edward Simpson Prescott
WP-02-24
Local Market Consolidation and Bank Productive Efficiency
Douglas D. Evanoff and Evren Örs
WP-02-25
Life-Cycle Dynamics in Industrial Sectors. The Role of Banking Market Structure
Nicola Cetorelli
WP-02-26
Private School Location and Neighborhood Characteristics
Lisa Barrow
WP-02-27
Teachers and Student Achievement in the Chicago Public High Schools
Daniel Aaronson, Lisa Barrow and William Sander
WP-02-28
The Crime of 1873: Back to the Scene
François R. Velde
WP-02-29
Trade Structure, Industrial Structure, and International Business Cycles
Marianne Baxter and Michael A. Kouparitsas
WP-02-30
Estimating the Returns to Community College Schooling for Displaced Workers
Louis Jacobson, Robert LaLonde and Daniel G. Sullivan
WP-02-31
A Proposal for Efficiently Resolving Out-of-the-Money Swap Positions
at Large Insolvent Banks
George G. Kaufman
WP-03-01
Depositor Liquidity and Loss-Sharing in Bank Failure Resolutions
George G. Kaufman
WP-03-02
Subordinated Debt and Prompt Corrective Regulatory Action
Douglas D. Evanoff and Larry D. Wall
WP-03-03
When is Inter-Transaction Time Informative?
Craig Furfine
WP-03-04
Tenure Choice with Location Selection: The Case of Hispanic Neighborhoods
in Chicago
Maude Toussaint-Comeau and Sherrie L.W. Rhine
WP-03-05
Distinguishing Limited Commitment from Moral Hazard in Models of
Growth with Inequality*
Anna L. Paulson and Robert Townsend
WP-03-06
Resolving Large Complex Financial Organizations
Robert R. Bliss
WP-03-07
4
Working Paper Series (continued)
The Case of the Missing Productivity Growth:
Or, Does information technology explain why productivity accelerated in the United States
but not the United Kingdom?
Susanto Basu, John G. Fernald, Nicholas Oulton and Sylaja Srinivasan
WP-03-08
Inside-Outside Money Competition
Ramon Marimon, Juan Pablo Nicolini and Pedro Teles
WP-03-09
The Importance of Check-Cashing Businesses to the Unbanked: Racial/Ethnic Differences
William H. Greene, Sherrie L.W. Rhine and Maude Toussaint-Comeau
WP-03-10
A Structural Empirical Model of Firm Growth, Learning, and Survival
Jaap H. Abbring and Jeffrey R. Campbell
WP-03-11
Market Size Matters
Jeffrey R. Campbell and Hugo A. Hopenhayn
WP-03-12
The Cost of Business Cycles under Endogenous Growth
Gadi Barlevy
WP-03-13
The Past, Present, and Probable Future for Community Banks
Robert DeYoung, William C. Hunter and Gregory F. Udell
WP-03-14
Measuring Productivity Growth in Asia: Do Market Imperfections Matter?
John Fernald and Brent Neiman
WP-03-15
Revised Estimates of Intergenerational Income Mobility in the United States
Bhashkar Mazumder
WP-03-16
Product Market Evidence on the Employment Effects of the Minimum Wage
Daniel Aaronson and Eric French
WP-03-17
Estimating Models of On-the-Job Search using Record Statistics
Gadi Barlevy
WP-03-18
Banking Market Conditions and Deposit Interest Rates
Richard J. Rosen
WP-03-19
Creating a National State Rainy Day Fund: A Modest Proposal to Improve Future
State Fiscal Performance
Richard Mattoon
WP-03-20
Managerial Incentive and Financial Contagion
Sujit Chakravorti, Anna Llyina and Subir Lall
WP-03-21
Women and the Phillips Curve: Do Women’s and Men’s Labor Market Outcomes
Differentially Affect Real Wage Growth and Inflation?
Katharine Anderson, Lisa Barrow and Kristin F. Butcher
WP-03-22
Evaluating the Calvo Model of Sticky Prices
Martin Eichenbaum and Jonas D.M. Fisher
WP-03-23
5
Working Paper Series (continued)
The Growing Importance of Family and Community: An Analysis of Changes in the
Sibling Correlation in Earnings
Bhashkar Mazumder and David I. Levine
WP-03-24
Should We Teach Old Dogs New Tricks? The Impact of Community College Retraining
on Older Displaced Workers
Louis Jacobson, Robert J. LaLonde and Daniel Sullivan
WP-03-25
Trade Deflection and Trade Depression
Chad P. Brown and Meredith A. Crowley
WP-03-26
China and Emerging Asia: Comrades or Competitors?
Alan G. Ahearne, John G. Fernald, Prakash Loungani and John W. Schindler
WP-03-27
International Business Cycles Under Fixed and Flexible Exchange Rate Regimes
Michael A. Kouparitsas
WP-03-28
Firing Costs and Business Cycle Fluctuations
Marcelo Veracierto
WP-03-29
Spatial Organization of Firms
Yukako Ono
WP-03-30
Government Equity and Money: John Law’s System in 1720 France
François R. Velde
WP-03-31
Deregulation and the Relationship Between Bank CEO
Compensation and Risk-Taking
Elijah Brewer III, William Curt Hunter and William E. Jackson III
WP-03-32
Compatibility and Pricing with Indirect Network Effects: Evidence from ATMs
Christopher R. Knittel and Victor Stango
WP-03-33
Self-Employment as an Alternative to Unemployment
Ellen R. Rissman
WP-03-34
Where the Headquarters are – Evidence from Large Public Companies 1990-2000
Tyler Diacon and Thomas H. Klier
WP-03-35
Standing Facilities and Interbank Borrowing: Evidence from the Federal Reserve’s
New Discount Window
Craig Furfine
WP-04-01
Netting, Financial Contracts, and Banks: The Economic Implications
William J. Bergman, Robert R. Bliss, Christian A. Johnson and George G. Kaufman
WP-04-02
Real Effects of Bank Competition
Nicola Cetorelli
WP-04-03
Finance as a Barrier To Entry: Bank Competition and Industry Structure in
Local U.S. Markets?
Nicola Cetorelli and Philip E. Strahan
WP-04-04
6
Working Paper Series (continued)
The Dynamics of Work and Debt
Jeffrey R. Campbell and Zvi Hercowitz
WP-04-05
Fiscal Policy in the Aftermath of 9/11
Jonas Fisher and Martin Eichenbaum
WP-04-06
Merger Momentum and Investor Sentiment: The Stock Market Reaction
To Merger Announcements
Richard J. Rosen
WP-04-07
Earnings Inequality and the Business Cycle
Gadi Barlevy and Daniel Tsiddon
WP-04-08
Platform Competition in Two-Sided Markets: The Case of Payment Networks
Sujit Chakravorti and Roberto Roson
WP-04-09
Nominal Debt as a Burden on Monetary Policy
Javier Díaz-Giménez, Giorgia Giovannetti, Ramon Marimon, and Pedro Teles
WP-04-10
On the Timing of Innovation in Stochastic Schumpeterian Growth Models
Gadi Barlevy
WP-04-11
Policy Externalities: How US Antidumping Affects Japanese Exports to the EU
Chad P. Bown and Meredith A. Crowley
WP-04-12
Sibling Similarities, Differences and Economic Inequality
Bhashkar Mazumder
WP-04-13
Determinants of Business Cycle Comovement: A Robust Analysis
Marianne Baxter and Michael A. Kouparitsas
WP-04-14
The Occupational Assimilation of Hispanics in the U.S.: Evidence from Panel Data
Maude Toussaint-Comeau
WP-04-15
Reading, Writing, and Raisinets1: Are School Finances Contributing to Children’s Obesity?
Patricia M. Anderson and Kristin F. Butcher
WP-04-16
Learning by Observing: Information Spillovers in the Execution and Valuation
of Commercial Bank M&As
Gayle DeLong and Robert DeYoung
WP-04-17
Prospects for Immigrant-Native Wealth Assimilation:
Evidence from Financial Market Participation
Una Okonkwo Osili and Anna Paulson
WP-04-18
Institutional Quality and Financial Market Development:
Evidence from International Migrants in the U.S.
Una Okonkwo Osili and Anna Paulson
WP-04-19
Are Technology Improvements Contractionary?
Susanto Basu, John Fernald and Miles Kimball
WP-04-20
7
Working Paper Series (continued)
The Minimum Wage Restaurant PricesDQG/DERU0DUNHW6WUXFWXUH
Daniel Aaronson, Eric French and James MacDonald
WP-04-21
8
@FedFRASER