Title
Markets
The Fulton Fish Market
Kathryn Graddy
This feature explores the operation of individual markets. Patterns of behavior
in markets for specific goods and services offer lessons about the determinants and
effects of supply and demand, market structure, strategic behavior, and govern-
ment regulation. Suggestions for future columns and comments on past ones
should be sent to James R. Hines Jr., c/o Journal of Economic Perspectives, Department
of Economics, University of Michigan, 611 Tappan Street, Ann Arbor, MI 48109
1220.
Introduction
The Fulton Fish Market was a colorful part of the New York City landscape that
operated on Fulton Street in Manhattan for over 150 years. In November 2005 the
market moved from the South Street Seaport in lower Manhattan to a large facility
built for the market at Hunts Point in the South Bronx. The Fulton Fish Market
now called The New Fulton Fish Marketis one of the worlds largest fish markets,
second in size only to Tsukiji, the famous fish market in Tokyo.
To economists, it may seem that a large centralized market with well-informed
buyers and sellers should also be a very competitive market. But fish is a highly
differentiated product. From 100 to 300 varieties of fish are sold at the Fulton
marketwith 15 to 20 different varieties of shrimp alone. Within each variety, there
are large fish, small fish, fish that were just caught, and fish that have been around
awhile (sometimes longer than they should be). Restaurateurs, retailers and even
consumers often wish to examine fish themselves or to have their agents do so. The
y Kathryn Graddy is University Lecturer and Fellow of Exeter College, University of Oxford,
Oxford, United Kingdom. Her e-mail address is [emailprotected].
Journal of Economic PerspectivesVolume 20, Number 2Spring 2006 Pages 207220
centralized market performs a matching functionmatching fish to buyers of
fishand large centralized markets make this matching easier and more efficient
for everyone. Indeed, the heterogeneity of fish explains why centralized fish mar-
kets have tended to survive, whereas methods of supplying other goods have often
become decentralized with time. Unlike, say, many agricultural and industrial
products, fresh fish is not easily treated as a well-defined commodity. Fish is more
perishable, supply is more unpredictable, and individual fish are more heteroge-
neous than most agricultural products.
The high level of product differentiation and the institutional structure in the
Fulton fish market can lead to patterns of behavior that suggest imperfect compe-
tition and a segmented market. Buyers at the fish market act as agents for customers
with very different elasticities of demand, from restaurants selling fishballs in
Chinatown, to expensive East Side restaurants, to fry shops serving fish sandwiches
in Harlem, to suburban fish retail outlets. At times in the past, the repeated nature
of price setting and extensive knowledge of the sellers may have created the basis
for tacit collusion and allowed the dealers to gather economic rents by exploiting
the different elasticities and buying patterns. Additional economic rents at the
market were created by subsidized rents, and lax regulation created fertile ground
for organized crime to operate.
History
The Fulton market was historically located in lower Manhattan, near the
Brooklyn Bridge, just a few blocks from Wall Street. The market first opened on
that site in 1807 on land donated to New York City, and at first was a general market
for both fish and goods other than fish. In 1822 the fish merchants occupied a new
Fulton market building, located on South Street between Fulton and Beekman
Streets. Prior to 1850, housekeepers from Brooklyn and nearby areas would pur-
chase fish directly from the market. However, since that time, wholesale customers
have been the primary buyers (Slanetz, 1986). The market gradually gained in
importance, and in 1924 the market sold 384 million pounds of fish, 25 percent of
all seafood sold in the United States (Miller, 1991).
However, the relative importance of the Fulton market declined in recent
years because of competing geographic markets such as Philadelphia and the New
Jersey docks, and also because of the increasing use of national or regional
suppliers. Philadelphias market, although much smaller in size, was more conve-
nient for merchants located near the area. The New Jersey docks allowed buyers to
purchase some types of fish directly from the fishermen. Based on estimates from
the New York Times of Fulton fish market volume (Bowen, 2005) and estimates of
U.S. fish consumption per capita from the National Oceanographic and Atmo-
spheric Administration (NOAA, 2005), about 5 percent of U.S. seafood sales flows
through the Fulton market. Restaurants and merchants outside coastal areas usu-
ally use national seafood brokers to supply their fresh fish, bypassing a centralized
208 Journal of Economic Perspectives
market completely. Without doubt, overnight shippers have increased the market
for seafood brokers. This option is available to merchants in the New York area, but
many merchants prefer to choose their fresh fish themselves, given the quality and
variety that is available at the Fulton market.
For comparison, Table 1 shows volumes at various well-known fish markets.
Tsukiji is by far the largest, with a volume of 615,409 tons per year, the Fulton fish
market is second, at around 115,000 tons per year, and Billingsgate in London is a
distant third at around 25,000 tons per year. All of these markets are located in
coastal cities. With modern transportation, having a fish market in a coastal city is
not strictly necessary. However, some coastal cities have built up a group of buyers
with the taste and appreciation for fish that allows them to sustain large fish
markets.
The Fulton market was primarily located in two open air structures, the Tin
Building and the New Building, in which various dealers rented stalls from the
Port Authority of New York with closed offices at the back of the stalls. The New
Building was opened in 1939 by Mayor La Guardia, after pilings of the old market
building gave way in 1936 and the entire building slid into the river.
Not only was the marketplace old and established, but many of the wholesalers
at the Fulton market were well-established firms. For example, the grandfather of
Robert Samuels, an owner of Blue Ribbon Seafood which supplies New York
restaurants such as La Bernardin and Bouley, started his business in the very early
part of the twentieth century. The traders also had established histories and family
ties to the business. For example, Martin (Buddy) Brady, a halibut man, comes
from a distinguished history of workers. Ive been here 38 years, he said, since I
was 17. My father did it before me, and my granddad before him. Mr. Brady
worked as a loader for 15 years before graduating to selling (Blau, 1985). Most of
Table 1
Yearly Volumes at Selected Fish Markets
Market Yearly volume in 2004 (tons)
Tsukiji (Tokyo, Japan) 614,409
Fulton Street (New York, NY) 115,000
Billingsgate (London, UK) 25,000
Sydney (Sydney, Australia) 15,000
Portland Fish Exchange (Portland, Maine) 11,000
Boston (Boston, MA) 10,000
Source: Data on all markets except for Fulton was taken from official websites for the
markets. Data on Fulton is an approximation taken from the New York Times (Bowen,
2005).
Notes: Some major regional markets for which volumes were not readily available
include Pikes Peak in Seattle, the Marseille fish market, the Newquay fish auction in
Cornwall in the United Kingdom and numerous markets throughout India and
Southeast Asia.
Kathryn Graddy 209
the wholesalers and traders have moved to the New Fulton Fish Market in the
Bronx.
How the Market Worked at Fulton Street
Fish began arriving at the market around midnight. Historically, fish was
received at the port of New York City by boat, but subsequently, all fish was brought
in by truck or air from other areas. The market was open from three to nine in the
morning on Monday and Thursday and from four to nine on Tuesday, Wednesday,
and Friday. Teams of loaders transported the fish from the trucks to the stalls by
hand trucks and small motorized pallet trucks. Once the buyers chose their fish, the
loaders reloaded the fish onto the customers trucks. Many of the market details are
similar at the New Fulton Fish Market in the Bronx. However, some details have
changed. For example, the loading and unloading of fish is done very differently,
primarily because of better access to the dealers stalls.
The market had three main sets of players: the buyers who purchased on
behalf of retail fish stores and markets, the wholesalers or dealers at the market,
and the suppliers who caught the fish in the first place. A common sight at the
market was buyers of tuna for sushi going around and sampling different tuna. The
buyers would take a sample, look at the color, rub the sample between their fingers
to determine the oiliness and taste it. Buyers who were interested in different
varieties of high-quality fish for their restaurant or retail shop were not inclined to
leave the purchasing decision to an agent. Conversely, some buyers were happy with
fish of a lower quality (primarily fish that were not quite so fresh), if the quality was
at least acceptable and they could get the boxes cheaply. To determine acceptable
quality and negotiate a good price, these owners or their agents would inspect the
fish and negotiate the price themselves.
Before the move to the Bronx, there were 60 registered dealers in the market.
The number seemed to be rising over time, since when I spent time at the market
in the early 1990s, I counted a total of 35 wholesalers actively operating in the
market. However, not all dealers carried all types of fish. In the early 1990s, there
were only six major dealers in whiting.
The supply arrangements for the different types of fish differed. I am most
familiar with the supply arrangements for whiting, since I observed the situation of
a particular whiting dealer.1 The whiting was supplied by fishermens cooperatives,
packing houses, and smaller fishing boats in New Jersey, Long Island, and Con-
necticut each day before the market opened. However, the price that a dealer
would pay a supplier for a particular days supply of whiting was determined at the
end of the day by the prices the dealer had received for the whiting on that
1 I chose to study whiting for three main reasons. First, at the time that I was at the market, more
transactions took place in whiting than in any other fish. Second, whiting do not vary as much in size and
quality as other fish. Finally, the dealer from whom I collected the data suggested that the whiting
salesman would be the only salesman amenable to having an observer.
210 Journal of Economic Perspectives
particular day. Each supplier serviced multiple dealers. A supplier would refuse to
deliver fish to a dealer from whom he continually received a poor price relative to
other dealers. In this way, according to my dealer, dealers competed with each
other for fish supplied in the future by the amount they paid the suppliers for fish
on a particular day. The dealer stated that he kept 515 cents on each pound of
whiting sold (although this margin could not be independently verified).
Due to custom and the practicalities of loading and unloading fish, the quantity of
whiting received for a particular day was received in its entirety before the market
opened. The quantity that each of the dealers received was readily observable by all
dealers. If one dealer did not receive what he perceived to be enough whiting, another
dealer would sell him whiting before the market opened. The price was set not at the
time of sale, but was determined toward the end of the trading day.
A dealer often had a good idea before the markets close of the quantity that
would be available the next day. Quantity supplied was primarily determined by
weather conditions. Wind and waves are the greatest determinant of the quantity of fish
likely to be caught. Other determinants of supply in the Fulton market were prices in
other places such as Philadelphia and the New Jersey docks, and the price received for
whiting relative to the prices received for other fish on days in the recent past. If the
price of whiting was relatively low, fishermen would seek other types of fish.
Anyone can purchase fish at the Fulton fish market, but small quantities were
not sold. For whiting, the minimum quantity that was sold was one box, approxi-
mately 60 pounds (except as favors to regular customers). Most of the customers
were repeat buyers, but buyerseller relationships varied. Some buyers regularly
purchased from one seller, and other buyers purchased from various sellers.
Most buyers followed distinct purchasing patterns. Monday, Thursday and
Friday were big days, but Tuesday and Wednesday were relatively quiet. Whiting
stays sufficiently fresh to sell for at most four days after it is received. However, it was
usually sold from the market either the day it was received or the day after. About
half of the time, total sales were less than total quantity received. As would be
expected, there was a small amount of oversupply as some inventory loss is in-
evitable in the selling process, especially given the perishable nature of fish. Over
111 days in late 1991 and early 1992, the total quantity received exceeded the total
quantity sold by 11,237 pounds, which amounts to about 1.6 percent of total sales.
The bulk of this difference consisted of fish not sufficiently fresh to sell; the surplus
was literally thrown out.
Most buyers purchased their fish on credit, with charges payable the following
week. It does not appear that there was a premium for credit purchases. Interest
rate effects were minimal and defaults were generally rare, as buyers were generally
repeat customers.
Pricing
The Fulton fish market had no posted prices and each dealer was free to
charge a different price to each customer. If a customer wished to buy a particular
Markets: The Fulton Fish Market 211
quantity of fish, he would ask a seller for the price. The seller would quote a price
and the customer would usually either buy the fish or walk away. Prices were
generally quoted in five-cent increments, which can seem quite large, considering
the average price was about 85 cents per pound. While sometimes the customer
would respond with a price that was five cents lower and back and forth bargaining
would take place (.65, .70, .65, .70 and so on), overt bargaining generally only
occurred with a very few large customers. Sellers were discreet when naming a
price. A particular price was for a particular customer. When I was observing,
sometimes a salesman and a regular purchaser would walk away, arm in arm,
discussing price very quietly.
My price data on whiting come from two sources. First, from April 13 to May
8, 1992, I spent everyday at the market collecting data on whiting prices from one
dealer. If I did not hear the final price, the salesman would report it back to me.
The dealer also gave me his inventory sheets for the period December 2, 1991,
through May 8, 1992. Each transaction is recorded on the inventory sheet, along
with a customer number for repeat customers. The inventory sheets matched up
well with the hand-collected data (Graddy, 1995).
Prices varied tremendously. For example, the average price in a whiting
transaction on Friday, May 1, 1992, was $.33/lb., and the average price on Friday,
May 8, was $1.75/lb., an increase of almost six-fold. Figure 1 is a high-low price
chart, where the bottom of the vertical line shows the low price for that day and the
top of the vertical line shows the high price for that day. The high intraday volatility
was primarily due to changes in price throughout the day, but some of the volatility
can be attributed to differences in qualities of fish. There was often a large decline
in price after 7:00 a.m., but the pattern before 7:00 a.m. was not easily predictable.
There was a higher intraday volatility on Friday than on other days.
Figure 2 indicates total daily sales, which varied considerably. Daily supply
shocks, primarily caused by weather, were largely responsible for the high volatility
in day-to-day prices. The simple correlation between total daily sales and average
daily price is .32.
Fish of the same variety at the same point in time can also differ tremendously
in price because of differences in quality. While I was collecting the data about
whiting, I placed the available whiting into one of five categories ranging from very
poor quality to very good quality, based on sight, feel and smell. I did not find that
the quality of whiting sold declined throughout the morningalthough this may
have been partly because during my study, the temperature outside was near
freezing, and the fish were well preserved by ice. Average price certainly differed by
quality; for the month that I spent at the market, fish of medium or better quality
averaged 83 cents per pound whereas fish of poor or very poor quality averaged 46
cents per pound. This price differential helps to explain why most fish merchants
wish to choose the fish themselves, matching the quality of whiting to the eventual
use.
About 60 percent of the purchasers at the market were Asian. I noticed that
sellers were quoting lower prices to Asian customers for the same box of fish. The
212 Journal of Economic Perspectives
price difference would usually be about 5 cents, though it could go to as large as
10 cents. Regressions that carefully control for other variables, including time of
sale and quality, indicated that whites pay on average 6.3 cents per pound more
on each transaction for the same type and quality of fish than do Asian buyers
(Graddy, 1995). While quality controls could naturally be imperfect, these
Figure 1
Daily Prices for Whiting
0
1
2
3
1/15 2/27
Date (December 2, 1991 May 8, 1992)
P
ri
ce
(
$/
p
o
u
n
d
)
4/9 5/812/2
Figure 2
Daily Volumes of Whiting
0
10000
20000
12/2 1/15 2/27 4/9 5/8
Date (December 2, 1991 May 8, 1992)
Q
u
an
ti
ty
(
p
o
u
n
d
s)
Kathryn Graddy 213
regressions were consistent with my observations. Why and how can such a price
difference occur?
On the demand side, there was strong anecdotal evidence that Asian buyers
were more elastic than white buyers. Asians buyers would resell their whiting as
whole fish in retail shops, fish sandwiches in fry shops, and would make them into
fish balls. Most of these establishments were located in very poor neighborhoods.
Resellers had little scope to raise the price of whiting for their ultimate customers.
Consequently, they would bargain very hard at the market. A New York Times article
described how one dealer named Donald Julich Sr. spent a half hour haggling over
a box of lobster tails with John Kim, the owner of a fish store in Queens. Kim left,
but came back an hour later, and paid the $17.50 (Ackman, 2000). Furthermore,
some of the Asian buyers purchased their fish at the Fulton market, and then resold
it in Chinatown. Chinatown has a reputation as the cheapest place to buy seafood
in New York City. Store owners claimed they must keep their prices low due to
fierce competition and the fact that most of their customers blue-collar
workerssimply cannot pay more (Lii, 1995).
Asian buyers appeared to be more organized than white buyers. The Korean
buyers had an active retailers organization. Shortly before my time at the market,
the association organized a boycott of one of the dealers because he was allegedly
shortchanging the pounds of fish included in a box. The dealers at Fulton Street
may have recognized that this organized group had more options, like purchasing
from fishing boats or intermediate wholesalers that supply the dealers at Fulton
Street.
In contrast, many white buyers (though not all) had more scope to pass on
prices to customers. For example, for the local fish dealer in Princeton, New Jersey,
whiting was a very small part of his purchase; he would spend no time shopping
around for price on whiting. If prices were especially high at the market, he could
explain to his customers that fish were expensive at Fulton Street and his customers
would often be willing to absorb costs by paying higher prices.
Why didnt the Asian buyers arbitrage the market, buying low and then
reselling to other customers? In all likelihood, it is very unlikely that either white
buyers or Asian buyers actually knew this was happening. Prices were quoted
discretely for particular customers. White buyers rarely socialized with Asian buyers
and hence communication between the two groups was limited.
These conditions establish why price discrimination might have been profit-
able on the demand side. However, why werent any extraordinary profits com-
peted away? The structure of the Fulton market for sales of whiting, at least in the
early 1990s, appeared conducive to tacit collusion, rather than perfect competition.
Remember that only a small number of dealers at the market (six at the time)
carried whiting. Many of the dealers at the market had been operating for many
years. Dealers would trade with each other before the market opened. The dealers
would price fish on a daily and sometimes hourly basis and would receive feedback
from the buyers on other dealers prices. Dealers in such a situation could easily
become very good at tacit communication. Moreover, entry into this market may
214 Journal of Economic Perspectives
well have been limited because suppliers provided the whiting on credit to estab-
lished dealersand it is not clear they would be willing to provide fish on credit to
a new entrant. The presence of organized crime in the market may also have
discouraged entry.
Wind and Waves
The quantity of fish brought to market is determined by many factors, but the
weather is an important determinant of supply because strong winds and high
waves make it difficult to catch fish. Quantities fall and prices rise when there are
storms, and quantities rise and prices fall in good weather. Without centralization,
it is unclear whether the price response could have been as effective in allowing
demand to match supply as buyers might have suspected gouging on the part of the
sellers if prices were not easily comparable.
As a demonstration of the price and quantity response to weather, in my price
and quantity data on the whiting market in 19911992, I classified the weather as
stormy when wave heights are greater than 4.5 feet and wind speed is greater than
18 knots. Wind speed and wave height are moving averages of the last three days
wind speed and wave height before the trading day, as measured off the coast of
Long Island and reported in the New York Times boating forecast. Holding day of
the week constant, the average quantity on a clear day was 2,371 pounds more than
on a stormy day. Conversely, the average price was 32 cents less per pound than on
a clear day. Average price on all days was 88 cents per pound and average quantity
was 6,335 pounds.
Variations in weather offer a way to use price and quantity data to estimate a
demand curve. The problem in estimating demand curves, of course, is that
different observations of price and quantity may be affected by shifts in either
supply and demand. However, both theory and evidence suggest that weather only
affects the supply of fish caught.2 For example, the consistent and significant price
reactions to stormy weather also suggest that the availability of close substitutes, or
at least those that are unaffected by the same storms, is limited. Fish restaurants and
retailers feel a need to offer fresh fish, even if the weather has not been ideal. The
data exhibit no price rigidities, and no excess supply and demand. Because of the
strong effect that weather has on supply and the price response allowed by a
centralized market, the price and quantity data from the Fulton fish market allow
a researcher to estimate a market demand curve and the elasticity of demand
(Angrist, Graddy and Imbens, 2000; Lee, 2004).
Table 2 shows estimates of elasticitiesthat is, the percentage change in
2 As noted in Angrist and Krueger (2001) and Stock and Trebbi (2003), P.G. Wright (1928) suggested
using yield per acre as an instrumental variable for estimating the demand elasticities of butter and
flaxseed, because yield per acre is primarily determined by the weather. This is one of the first known
solutions to the identification problem.
Markets: The Fulton Fish Market 215
quantity demanded in response to a percentage change in pricefirst using
ordinary least squares and then using instrumental variables with stormy weather as
an instrument. In the regressions, fish has been treated as an approximately
homogeneous product. The first column is an ordinary least squares regression
with log quantity as the dependent variable and log price as the independent
variable. The quantity is the total amount sold on a day and the price is the average
price for that day.3 A higher price has a negative effect on quantity. The second
column shows that this estimate is unchanged by including dummy variables for the
day of the week (Friday is the omitted day), and for measures of the weather on
shore.
The third column then uses an instrumental variables approach. That is, first
a regression is run with log price as the dependent variable and the storminess of
the weather as the explanatory variable. This regression seeks to measure the
variation in price that is attributable to stormy weather. The coefficients from this
regression are then used to predict log price on each day, and these predicted
values for price are inserted back into the regression. The third column shows that
the impact of these predicted values of price on quantity are double the ordinary
3 There does not appear to be any correlation between stormy weather and the quality of whiting sold.
Table 2
Ordinary Least Squares and Instrumental Variable Estimates of Demand
Functions with Stormy Weather as an Instrument
Variable
Ordinary least squares
(dependent variable:
log quantity)
Instrumental
variable
(1) (2) (3) (4)
Log price 0.54 0.54 1.08 1.22
(0.18) (0.18) (0.48) (0.55)
Monday 0.03 0.03
(0.21) (0.17)
Tuesday 0.49 0.53
(0.20) (0.18)
Wednesday 0.54 0.58
(0.21) (0.20)
Thursday 0.09 0.12
(0.20) (0.18)
Weather on shore 0.06 0.07
(0.13) (0.16)
Rain on shore 0.07 0.07
(0.18) (0.16)
R 2 0.08 0.23
No. of Obs. 111 111 111 111
Source: The data used in these regressions are available by contacting the author.
Note: Standard errors are reported in parentheses.
216 Journal of Economic Perspectives
least squares estimates. Adjusting for the day of the week, as is done in the fourth
column, increases the coefficient on the price still further. As would be expected,
the ordinary least squares estimates of the impact of price on quantity are much
smaller than the instrumental variables estimates.
In addition, it is interesting that the point estimates of the average elasticities
in the third and fourth columns are both about 1. This finding has several
interesting implications. First, it is consistent with pricing power on the part of the
fish dealers. A price-setting firm will raise price to the point where the percentage
change in the quantity demanded is at least as large as the percentage change in
price; otherwise, it would make sense to raise the price even more. Second, when
demand has a unitary elasticity, it means that the percentage change in quantity
would always equal the percentage change in price, and the weather would there-
fore not have much effect on a sellers revenue, keeping fishermens incomes
relatively constant. If demand were not approximately unit elastic, then sellers
would be subject to widely varying revenues with price swings. In such a setting,
market organization might well be different, with the sellers less willing to have
prices change on a regular basis. Third, unit elasticities could also result from
budget constraints on the part of some buyers. Undoubtedly, some buyers go to the
fish market expecting to spend a certain amount. If the weather has been terrible,
resulting in lower quantities and higher prices, they will be forced to buy less fish.
Prices may rise and quantities may fall to the point where buyers are spending
approximately the same amount at the market, regardless of the weather.
Crime at the Fish Market
Crime and the mafia have had a long history at the Fulton fish market. Law
enforcement experts point out that the mafia may find it easier to gain a foothold
when a relatively small number of people and businesses operate in a single
location. At one point in the early 1900s, oystermen and fishermen refused to enter
the harbor without police guard (Slanetz, 1986). Since then, there were continued
reports of extortion, tapping(that is, stealing fish), income tax evasion and
obtaining false loans. All transporting of fish at the market was done by the loading
teams, a system which evolved over 75 years ago (Raab, 1996), and these loading
companies formed a monopoly over the loading and unloading of fish. Loading
companies also controlled parking in city streets adjacent to the market, partially
through extortionate measures (Raab, 1995b). Effectively, one could not park on
city parking spaces near the market without paying.4
The involvement of the mafia was openly admitted in the early 1990s. At about
that time, the city gave the dealers at the market a vastly subsidized rental deal. A
4 In 1992, I paid $5.00 to park my car in a city space that should have been free parking. I would arrive
at the market at about 3:00 a.m., and I was happy to pay the money as I felt relatively safe walking from
my car to the market, which was not true of most parts of New York City at that time of the morning.
Kathryn Graddy 217
federal study in 1992 estimated the subsidy to be about $2.7 million per year
(Editorial Desk, 1995). From 1982 to 1995, rents were virtually unchanged for
dealers at the market. City officials claimed that part of the problem was that
previous city administrations had not wanted to sign long-term leases with dealers
who might later be found to have mafia ties (Raab, 1994). Of course, this argument
could be completely circular. The low rents in effect subsidized the presence of the
mafia as the lower cost base allowed excess profits. In 1995, with Rudolph Guiliani
as mayor, the administration began negotiations to raise rents for 40 dealers who
currently paid about $270,000 a year. They were able to force an agreement with
the wholesalers that would bring the citys intake on rents up to about $2 million
by 1998 (Raab, 1995a).
It is likely that organized crime chose to operate at the Fulton market not only
because of economic rents that were present, but also because it was relatively
unregulated. This situation changed in the mid-1990s. In the early 1990s, about ten
loading teams shared the market. However, as part of Rudolph Guilianis move
against organized crime, in 1995 the Guiliani administration held a sealed-bid
auction to award unloading licenses to companies. The six incumbent unloading
companies bid the exact same amount$1.35 per 100-pound crate. One cleaning
company, the Laro Maintenance Corporation, that had no ties to the market or
experience in unloading fish, bid $1.10 per 100-pound crate and Laro won the
entire loading contract (Van Natta, 1995). Since this crackdown on the market in
1995 and the increase in rents, there were few reports of continued mafia involve-
ment and by 1998, the Guiliani administration claimed to have cleared the mafia
out of the market.
Conclusion
Fish markets and the demand for fish have long been a favorite example and
subject of academic economists. Alfred Marshall used fresh fish for a variety of
examples in his Principles of Economics (1890 [1930]). More recently, Bell (1968)
examined the effect on the price of fish of the Papal-Bishops decree in 1966 that
Catholics no longer had to abstain from meat on Fridays, except during Lent.
Kirman and Vignes (1991) and Hardle and Kirman (1995) investigated the price
dynamics of the Marseille fish market and find significant instances of price
dispersion.
This article has suggested that the Fulton fish market is a centralized market
that exists because of the heterogeneity of fish and the differing tastes and needs
of customers. This heterogeneity has led to market im
