Contract Options For Buyers and Sellers of Talent in Professional Sports 1St Ed Edition Duane W Rockerbie Full Chapter

Contract Options for Buyers and Sellers
of Talent in Professional Sports 1st ed.

Edition Duane W Rockerbie
Visit to download the full and correct content document:
https://ebookmass.com/product/contract-options-for-buyers-and-sellers-of-talent-in-pr
ofessional-sports-1st-ed-edition-duane-w-rockerbie/
PALGRAVE PIVOTS IN SPORTS ECONOMICS
Contract Options
for Buyers
y and
Sellers of Talent in
Professional Sports
Duane W Rockerbie
Stephen T. Easton
Palgrave Pivots in Sports Economics
Series Editors
Wladimir Andreff
Emeritus Professor
University Paris 1 Panthéon-Sorbonne
Paris, France
Andrew Zimbalist
Department of Economics
Smith College
Northampton, MA, USA
This mid-length monograph series invites contributions between 25,000–
50,000 words in length, and considers the economic analysis of
sports from all aspects, including but not limited to: the demand for
sports, broadcasting and media, sport and health, mega-events, sports
accounting, finance, betting and gambling, sponsorship, regional devel-
opment, governance, competitive balance, revenue sharing, player unions,
pricing and ticketing, regulation and anti-trust, and, globalization. Sports
Economics is a rapidly growing field and this series provides an exciting
new publication outlet enabling authors to generate reach and impact.
More information about this series at

http://www.palgrave.com/gp/series/15189
Duane W Rockerbie · Stephen T. Easton
Contract Options
for Buyers and Sellers
of Talent
in Professional Sports
Duane W Rockerbie Stephen T. Easton
Department of Economics Department of Economics
University of Lethbridge Simon Fraser University
Lethbridge, AB, Canada Burnaby, BC, Canada
ISSN 2662-6438 ISSN 2662-6446 (electronic)

Palgrave Pivots in Sports Economics
ISBN 978-3-030-49512-1 ISBN 978-3-030-49513-8 (eBook)
https://doi.org/10.1007/978-3-030-49513-8
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer
Nature Switzerland AG 2020
This work is subject to copyright. All rights are solely and exclusively licensed by the
Publisher, whether the whole or part of the material is concerned, specifically the rights
of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on
microfilms or in any other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc.
in this publication does not imply, even in the absence of a specific statement, that such
names are exempt from the relevant protective laws and regulations and therefore free for
general use.
The publisher, the authors and the editors are safe to assume that the advice and informa-
tion in this book are believed to be true and accurate at the date of publication. Neither
the publisher nor the authors or the editors give a warranty, expressed or implied, with
respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
This Palgrave Macmillan imprint is published by the registered company Springer Nature
Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Sports economics is a relatively new field of research that is gaining popu-

larity and momentum. The authors of this book have produced research
and publications in the field for the last two decades, focusing mainly
on sports labor market issues and issues relating to sports league poli-
cies, such as revenue sharing, parity, and changes in rules of the game.
It has always been our purpose (and pleasure) to introduce models and
methods borrowed from other economic research fields, such as auction
models, models of multiple equilibria, consideration of talent conjectures,
exchange rate risk exposure, time series methods, and other innovations.
We do this to expand the sports economics literature and bring it to a
level of analysis consistent with other fields. This is an ambitious, and
sometimes challenging, endeavor but we strive to produce new research
that breaks new ground in sports economics that does not merely add to
the list of previously published results that are well-known and accepted.
Our purpose in this short book is to import another established
method from a field of economics outside of sports economics to attempt
to answer a problem. Professional athletes are paid large sums relative
to other professions, to play a game accessible to most children at a
much higher level that fans will pay to watch. Many publications exist
in sports economics that address the question of what financial compen-
sation a professional athlete should receive using profit-maximizing princi-
ples. However very few consider the issue of how players and team owners
actually bargain for salaries using the incentives and leverages available
v
vi PREFACE
to them. Instead, sports economists vaguely refer to “bargaining power”

when a player or a team owner gets the upper hand over salary nego-
tiations. Our motivation is provided by a relatively standard estimation
of marginal revenue product for baseball players that is well-known in
the sports economics literature. We find that certain groups of players
exhibit a consistent ability to exploit team owners, while other groups
have no ability to do so. Our “import” in this book is options theory as a
logical explanation for why some players are able to extract vast amounts
of money from owners, while others appear to be exploited. When one
thinks of the argument we make, it is clearly logical, however formal-
izing the argument into an economic model of a profit-maximizing team
owner is not so easy. The payoff is that we are able to provide insights
into the salary negotiation process that would not otherwise be possible
by developing a formal model. It is our hope that our contribution to this
literature is well-received and offers up an incentive for future research in
the field, perhaps in other professional sports than baseball.
We would like to thank Anthony Krautmann for specific comments
and encouragement, as well as the attendees of our NAASE session at
the 94th annual conference of the Western Economics Association in San
Francisco, June 2019.
Lethbridge, Canada Duane W Rockerbie

Burnaby, Canada Stephen T. Easton
May 2020
Contents
1 On the Rise: Player Compensation and Multi-year

Contracts 1
1.1 Introduction 1
1.2 A Brief History of Multi-year Contracts: Early Years
(1876–1975) 5
1.3 Early Years of Free Agency 11
1.4 Free Agency Since 2000 12
1.5 Multi-year Contracts in Basketball, Football,
and Hockey 13
1.6 Player Salaries and Marginal Revenue Products:
the Scully Method 16
1.7 Real Options 19
References 20
2 The Puzzle of Overpaid and Underpaid Players 23

2.1 Introduction 23
2.2 A Revenue Model 27
2.3 Estimates of the Revenue and Winning Percentage
Functions 32
2.4 Calculating a Player’s Net MRP 36
2.5 A Linear Probability Model 47
2.6 Summary 49
References 51
vii
viii CONTENTS
3 Contract Options for Buyers and Sellers of Talent 53

3.1 Introduction 53
3.2 Employment Contracts that Include Real Options 54
3.3 Setup of the Real Options Model 57
3.4 The Player’s European Put Option 60
3.5 The Owner’s European Put Option 64
References 66
4 Extensions to the Put Option Model 69

4.1 The Player’s Option Value in a Three-Year Contract 70
4.2 The Owner’s Option Value in a Three-Year Contract 74
4.3 Option Values with Different Career Profiles 75
4.4 Contract Values with Player and Owner Put Options 78
4.5 Multi-year Contracts with Option Years 81
4.6 No-Trade Clauses in Multi-year Contracts 82
References 83
5 Concluding Remarks 85
5.1 An Empirical Test 86
5.2 Extensions to the Real Options Model 89
References 91
Index 93
List of Figures
Fig. 1.1 MLB average team payroll, average team revenue and
payroll share of revenue, 1980–2018 (Source https://sites.
google.com/site/rodswebpages/codes taken on December
20, 2019. These figures are taken from Financial World
and Forbes magazine) 2
Fig. 1.2 Monopsony model of talent acquisition (Source Author’s
creation) 17
Fig. 3.1 Possible states during the two-year player contract (Source
Author’s creation) 58
Fig. 3.2 Options available to the player after one season (Source
Fig. 4.1 Possible states during the three-year player contract (Source
Fig. 4.2 Possible paths for the player after two seasons (Source
Fig. 4.3 Player paths that give positive put option value for the
owner (Source Author’s creation) 75
Fig. 5.1 Scatter plot of surplus to team owner versus number of
years of player contract (Source Author’s creation) 87
ix
List of Tables
Table 2.1 Free-agent fielding players by playing position,

2000–2012 26
Table 2.2 Estimates of the (natural logarithm of the) revenue
function 29
Table 2.3 Estimates of the (natural logarithm of the) revenue
function (2.1) 34
Table 2.4 Estimates of the logistic winning percentage function in
(2.2) 36
Table 2.5 Estimated surpluses for MLB free agents, 2006–2012,
(η S = 2) 41
Table 2.6 Estimated surpluses for MLB free agents, 2000–2006,
(η S = 2) 42
Table 2.7 Exploitation rates for MLB free agents versus salary,
2000–2012, (η S = 2) 43
Table 2.8 Exploitation rates for MLB free agents versus salary,
2000–2012, (η S = ∞) 46
Table 2.9 Exploitation rates of free-agent players by contract
length, 2000–2012, (η S = 2) 47
Table 2.10 Exploitation rates for MLB free agents versus salary
using linear probability model, 2000–2012, (η S = 2) 49
Table 2.11 Exploitation rates of free-agent players by contract
length, 2000–2012, (η S = 2) 50
Table 5.1 Weighted least squares estimate of surplus model 88
xi
CHAPTER 1
On the Rise: Player Compensation

and Multi-year Contracts
Abstract A salary anomaly has been identified in the sports economics

literature suggesting that some athletes are overpaid relative to their
marginal revenue product, while others are underpaid. The use of real
options theory answers this anomaly by tying the under or overpayment
of salaries to contract lengths. We begin this chapter with a brief history
of multi-year contracts in Major League Baseball and find that multi-year
contracts are much more prevalent after the repeal of the reserve clause
in 1976. The so-called Scully method to estimate a player’s marginal
revenue product is standard in the sports economics literature and we
review it here. We close the chapter by intuitively explaining how real
options theory explains the salary anomaly to motivate the rest of the
book.
Keywords Real options · Marginal product · Contract lengths · Baseball
1.1 Introduction
Since the beginning of the 1990 season, the date at which systematic
financial data for Major League Baseball teams became available, it has
been a boom time for the professional sport. Attendance at games has
increased from an average of 26,000 per game to over 28,000 and overall
© The Author(s) 2020 1

D. W. Rockerbie and S. T. Easton, Contract Options for Buyers
and Sellers of Talent in Professional Sports, Palgrave Pivots in Sports
Economics, https://doi.org/10.1007/978-3-030-49513-8_1
2 D. W. ROCKERBIE AND S. T. EASTON
350 0.8
300 0.7
0.6
250
0.5
$ millions
200
0.4
150
0.3
100
0.2
50 0.1
0 0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Revenue Payroll Payroll/Revenue
Fig. 1.1 MLB average team payroll, average team revenue and payroll share
of revenue, 1980–2018 (Source https://sites.google.com/site/rodswebpages/
codes taken on December 20, 2019. These figures are taken from Financial
World and Forbes magazine)
league attendance has grown from 54.8 million to 68.5 million in 2019.1
Between 1990 and 2016 average revenue for clubs increased more than
sixfold, and even adjusting for inflation rose 550%. At the same time
payroll for the average major league club expanded even faster. From
1990 to 2016 the average payroll paid to players’ salaries rose by 800% in
real (inflation adjusted) terms.
Major League Baseball (MLB) finance has undergone considerable
change since the 1980s, most notably the rapid increase in salaries has
coincided with the even greater increase in revenues. Figure 1.1 plots
the average team payroll from 1980 to 2018, the average team revenue
from 1990 to 2018, and the payroll share of team revenue.2 The average
annual rate of increase of average team payroll is 28.4%. The average team
1 https://www.baseball-reference.com/leagues/MLB/misc.shtml taken March 27,

2020.
2 Source: https://sites.google.com/site/rodswebpages/codes taken on December 20,
2019. These figures are taken from Financial World and Forbes magazine.
1 ON THE RISE: PLAYER COMPENSATION AND MULTI-YEAR CONTRACTS 3
payroll increased from $17.3 million in 1990 (39% of total expenses) to

$154.7 million in 2019 (53.3% of total expenses).3 Average team revenue
increased from $51.5 million in 1990 to $329.8 million in 2018. The
payroll share of team revenue increased from 27.6% in 1990 to 32.6% in
2018, however this is still well below the 50% payroll share of revenue
in the salary cap systems used in the National Football League (NFL),
National Basketball Association (NBA), and National Hockey League
(NHL).4 Gate revenue’s share of total revenue has decreased from an
average of 33.3% in 1990 to 28.7% in 2019. The salary shares presented
in Fig. 1.1 are underestimates of the true figures as a number of excluded
payroll categories are difficult to measure (Zimbalist 2011). Some of these
include deferred salary payments, signing bonuses, non-roster players,
player benefits, minor league payrolls, and whether the payroll is measured
at the start of the season, mid-season, or end of the season. It is also
not clear what revenues should be included in reported revenues and to
what extent these are related to baseball operations. For instance, lands
adjacent to ballparks that are owned by the team owner that generate
annual revenue, and would not otherwise generate revenue if not for
the operation of the team in the adjoining stadium, are not included in
baseball-related income (BRI). We rely on Forbes estimates throughout
this book, but we acknowledge the possible errors in their calculations.
Team values have increased from a median value of $102.5 million in
1990 to $1.58 billion in 2019, yielding an average annual growth rate
of 48.0% (without compounding). New expansion clubs have appeared
in Miami, Tampa Bay, Denver, Phoenix, and one club relocated from
Montreal to Washington, DC. A new, more extensive revenue sharing
system was adopted in 2002 that is even more extensive today, and in the
late 1990s a competitive balance tax on payrolls above a given threshold
level was phased in. Increased attendance resulted in the construction
of 21 new ballparks at a total estimated cost to clubs and taxpayers of
$7.8 billion.5 Even after general inflation (the all items Consumer Price
3 Source: https://sites.google.com/site/rodswebpages/codestaken on December 20,

2019. These figures are taken from Financial World and Forbes magazine.
4 1994 is an exception as the baseball lockout shortened season inflated the share going
to players.
5 Source: http://www.ballparks.com/ taken on February 8, 2017.
Index increased by 85.6% between 1990 and 2016), MLB operates on a

much larger dollar scale that it did in the 1980s.6
The impressive growth in player salaries in MLB begs the question of
whether players are being paid what they are worth. Economics suggests
that player salaries are proportional to the revenue a player generates for
the team owner, more specifically, the player’s marginal revenue product
(MRP). Players do not generate revenue directly for their owners, rather
their talent contributes to a team production function that generates
winning results. These wins then translate into a greater demand for
the team’s product (tickets, television and media subscriptions, etc.) that
result in greater revenue for the owner. The owner pays the player for
his skills on the field and keeps a share of the revenue generated by these
skills: what we term the surplus from talent. Owners and players deter-
mine the relative shares when negotiating salary contracts in MLB. Player
salaries in the National Football League (NFL), National Basketball Asso-
ciation (NBA), and National Hockey League (NHL) have definite ceilings
due to salary caps agreed to in their collective bargaining agreements
(CBAs), however MLB has no salary caps, so players are better able to
capture the full value of their MRPs.
Our task in this monograph is to suggest an answer to the apparent
anomaly in MLB that many players are paid well above the MRPs. This
evidence is supported by a number of studies that we discuss later. An
early study by Scully (1974) found convincing evidence that players were
paid well below their MRPs in the reserve clause period (defined in the
next section). This is understandable since the reserve clause prevented
the movement of players to new teams by limiting free agency. The stan-
dard monopsony model of industry structure predicts that the wage rate
paid to players will be well below their MRPs when there is only a single
buyer of talent. We do not refute Scully’s results. However, subsequent
analyses using more recent salary and performance data after the aboli-
tion of the reserve clause suggest that the situation has become reversed
for many players, but in particular, for highly talented players. One can
justify this anomaly by pointing to the greatly increased bargaining power
of these players after the fall of the reserve clause, however it is not clear
what this bargaining power actually is.
6 Source: http://www.inflationdata.com/inflation/Consumer_Price_Index/ taken on

February 8, 2017.
We attempt to clarify the process of bargaining by introducing the

concept of real options. A player and an owner who wish to agree on
a multi-year contract each give up valuable options in comparison to a
series of one-year contracts. Players give up the option to move to another
team if more attractive ones become available. Owners give up the option
to leave the player un-signed in future years if the player’s performance
diminishes and his MRP falls. Salary negotiations then involve each party
placing values on these options and agreeing on their net value. If the net
value favors the player, such as a younger player moving into the prime
of his career, the player will demand and receive a salary greater than
his expected MRP. If the net value favors the owner, such as an older
player whose skills and MRP could diminish, the owner will demand and
pay a salary less than the player’s expected MRP. Hence our explanation
for the apparent salary–MRP anomalies in multi-year contracts. In single-
year contracts, the player and owner do not have options of any value, so
players should be paid a salary equal to their expected MRP.
The remainder of this chapter provides brief histories of salaries and
multi-year contracts of more notable players, and a brief review of the
Scully (1974) approach to estimating MRPs that is essential to finding
evidence for the real options approach. We develop a technique to esti-
mate free-agent player MRPs and estimate them in Chapter 2. Our
findings confirm the salary–MRP anomaly for a recent sample period and
motivate the real options model that follows in Chapter 3. Chapter 4
provides extensions to the real options model and Chapter 5 provides
concluding remarks.
1.2 A Brief History of Multi-year

Contracts: Early Years (1876–1975)
The history of the labor market in MLB can be divided into three distinct
periods that progressed from a system of indentured servitude for players
to the lucrative market that exists today in which players have tremendous
bargaining power. These divisions mark a century and a half of contract
history: the reserve clause period from 1876 to 1975, the early years of
free agency from 1976 to 2000, and the years since 2001 when television
and media rights fees exploded.
The first period of contract history spanned almost a century. The
inception of the National League (NL) of Baseball in 1876 created a
closed league with only eight teams operating in some of the largest cities
in the eastern United States. Amateur leagues quickly dissolved as a conse-

quence of the higher quality of baseball played in the NL. Players in the
NL were full-time employees during the playing season, meaning they did
not need to hold other jobs to earn a living, and they could devote their
time to honing their baseball skills. The NL also saw bargaining power
for salaries transferred from the players to the owners with the inclusion
of a reserve clause in the standard playing contract (fully implemented
in 1879) and a salary cap of $2000 when annual earnings of non-farm
employees was $403.7 . The reserve clause generally stated that if a player
and an owner could not agree on a salary for the coming MLB season,
the owner had the right to “reserve” the player at the salary paid in the
season just ended. The owner could then invoke the reserve clause at the
end of each season that the owner wished to retain the player, effectively
tying the player to one team for the length of his playing career. If the
owner chose not to retain the player’s services, the player was a free agent
able to sign a contract with any team that would have him.
Contracts were for only a single season in the early years of base-
ball—multi-year contracts were unheard of—however, the players did not
complain as most were happy to earn a good living playing baseball. A
player would receive his contract in the mail at the end of the playing
season, sign it, and mail it back to the team owner without question.
Players were not represented by agents, instead contract negotiations were
conducted between the player and the team owner, as owners refused
to negotiate with player agents. Team owners were allowed to release a
player from his contract at any time during the playing season in exchange
for a small compensation payment. Thus players had very little job secu-
rity. Cash sales of players were allowed as well with the player receiving
none of the proceeds. In this way, team owners kept the surplus value of
the player’s services among themselves. Connie Mack used this strategy
twice during his tenure with the Philadelphia Athletics (now operating
in Oakland) during years in which his team struggled financially. Mack
sold many players during the 1914–1916 seasons, the most notable of
which were Eddie Collins, sold for $50,000 to the Chicago White Sox
7 Table Ba4280-4282 in Historical statistics of the United States, earliest times to the
present: Millennial edition, edited by Susan B. Carter, Scott Sigmund Gartner, Michael
R. Haines, Alan L. Olmstead, Richard Sutch, and Gavin Wright. New York: Cambridge
University Press, 2006. http://dx.doi.org/10.1017/ISBN-9780511132971.Ba4214-454
410.1017/ISBN-9780511132971.Ba4214-4544.
in 1914, and Home Run Baker, sold to the New York Yankees in 1916
for $37,500. Even a World Series appearance in 1931 could not keep the
team afloat, subsequently, Mack sold Mickey Cochrane for $100,000 to
the Detroit Tigers, Lefty Grove for $125,000 to the Boston Red Sox,
and Al Simmons, Jimmy Dykes and Mule Haas to the Chicago White
Sox for $100,000. These cash sales effectively moved the Athletics out of
contention by the 1934 season.
The reserve clause effectively eliminated the need for multi-year
contracts, even for the best players. However, there were a few excep-
tions. Data on player contracts is sparse for the early years of baseball all
the way up to the 1960s, and generally, the data that are available are
for the star players. Ty Cobb negotiated a three-year contract with the
Detroit Tigers in 1910 for an annual salary of $9000, significantly more
than the average player salary of $3000. Cobb did not report to spring
training for the 1913 season and missed a number of games while holding
out but received a one-year contract for what is thought to have been the
first five-figure salary in MLB history ($12,000). He negotiated a two-
year contract with Detroit in 1914 for annual salaries of $15,000 and
$20,000, and inked multi-year contracts thereafter until becoming the
player-manager of the Tigers in 1921. Cobb was arguably the top player in
MLB over the 1910–1920 seasons, achieving a weight-average8 batting,
on-base percentage and slugging percentage (weighted by at bats) of
0.382, 0.453, and 0.462, respectively, and was the American League (AL)
MVP in 1911.9 Cobb led the AL in batting average in nine of those
seasons and led in total hits in five. After being released by the Tigers
in 1927, Cobb signed a one-year contract for an MLB record salary of
$85,000 with the Philadelphia Athletics.10 His investments in real estate
and the stock market made Cobb a very wealthy man, far beyond his
wealth from baseball.
Yankee Stadium is still known as the “house that Ruth built,” despite
not being the original stadium in which Ruth played. That “house” was in
the Bronx, built in 1923, renovated in 1976, and closed in 2008. Ruth’s
legacy lives on through his amazing playing statistics and his financial
8 Each season’s batting average weighted by the number of at bats to form an overall
batting average for the period.
9 All playing statistics hereafter taken from http://www.baseball-reference.com.
10 To put this salary in perspective, the Dodger’s great Sandy Koufax earned the same
salary in his 1965 season, some 38 years after Cobb’s.
rewards. Ruth signed a string of one-year contracts with the Boston Red
Sox and New York Yankees up to the 1922 season, at which point he
negotiated a three-year contract at $52,000 per year (40% of the Yankee’s
team payroll). A two-year contract followed in the 1925 season for the
same salary, followed by a three-year contract in the 1927 season for
$70,000 per season. Ruth played another four seasons for the Yankees,
each a one-year contract, until he retired. Ruth’s playing statistics over
the 1922–1930 seasons are phenomenal: a weight-average batting, on-
base percentage, and slugging percentage (weighted by at bats) of 0.351,
0.479, and 0.714, respectively, His season average of 46 home runs would
be impressive today, however Ruth played when the playing rules dictated
that hits that cleared the fences but curve outside the foul pole did
not count as home runs. He earned AL MVP honors in 1923 and the
Yankees won three World Series championships (appearing in six) over
1922–1930.
At the time of Babe Ruth, the Yankees “murderer’s row” lineup also
included Lou Gehrig. Over the 1926–1937 seasons, Gehrig averaged 37
home runs per season (weighted by each season’s at bats) and featured
a weight-average batting, on-base percentage, and slugging percentage
of 0.348, 0.457, and 0.651, respectively. He was AL MVP in 1927 and
1936 and won World Series titles in four of these seasons. Gehrig was
truly the ironman of MLB in his playing days, playing in almost every
regular season game. Despite these statistical accolades that are, in many
ways, comparable to Ruth, Gehrig was paid far less than Ruth and only
once negotiated a multi-year contract—a three-year deal in 1928 paying
him $25,000 per season. This was a significant increase in salary compared
to the $8000 he was paid in the 1927 season. Gehrig’s annual salary was
consistently in the $25,000–$30,000 range over much of his career with
the Yankees, suggesting he was severely underpaid.
Jimmie Foxx began his playing career with the Philadelphia Athletics
in 1925 and quickly established himself as one of the elite players in base-
ball. Foxx batted 0.354 in 1929, slugged 33 home runs and contributed
118 runs batted in (RBI). He was rewarded with a three-year contract
paying a total of $50,000 by Connie Mack, manager, treasurer, and part
owner of the Athletics. The team won World Series championships in
1929 and 1930 and Foxx won MVP titles in 1932, 1933, and 1938.
His career statistics compare very favorably to Cobb, Ruth, and Gehrig
in many respects, yet his annual salary averaged far less, likely due to the
financial difficulties constantly plaguing Mack and the Athletics.
Many other star players toiled in the major leagues until the 1960s,
perhaps most notably Joe DiMaggio, Ted Williams, and Willie Mays.
DiMaggio (the “Yankee Clipper”) played for the New York Yankees for
the 1936–1942 and 1946–1951 seasons, with three years of military
service during World War II. Although an excellent player throughout
his career, DiMaggio’s prewar statistics are impressive: a 0.339 batting
average (weighted by at bats), 0.403 on-base percentage, and 0.606 slug-
ging average, earning MVP titles in 1939 and 1941.11 In salary history,
DiMaggio is best known as the first six-figure player in MLB history,
earning $100,000 for the 1949 and 1950 seasons. He is also known
for holding out at the start of the 1939 season, demanding a one-year
contract at $40,000 when the Yankees offered $25,000. With nowhere
else to play, DiMaggio accepted the Yankees offer with Yankees owner,
Colonel Jacob Ruppert, quipping “I hope the young man has learned his
lesson.”12 Despite having an outstanding career and being a tough nego-
tiator, DiMaggio never negotiated a multi-year contract. Ted Williams is
generally thought to be the highest paid player in MLB during 1951 to
1960 seasons, consistently earning $100,000 or more per season. One of
the greatest hitters to have ever played, Williams played his entire career
with the Boston Red Sox and negotiated a number of short multi-year
contracts toward the end of his career. Beginning in 1951, Willie Mays
played his entire 22-year career with the New York/San Francisco Giants,
winning MVP honors in the 1954 and 1965 seasons. Mays 0.345 batting
average and 41 home runs helped power the Giants to a World Series title
in 1954, but the team achieved few other successes during his lengthy
playing career. In his best season (1954), Mays earned $12,500 and was
severely underpaid afterward with a series of one-year contracts. A three-
year contract was agreed in 1966 that paid Mays $105,000 per season.
This contract set the standard for the famous dual contract hold outs
of Sandy Koufax and Don Drysdale in 1966. Believing that they could
negotiate more effectively as a pair, Koufax and Drysdale demanded three-
year contracts totaling $500,000 each, but ultimately capitulated to the
11 In 1941 DiMaggio also hit in 56 straight games, a record that still stands.
12 D. Gaffney, “DiMaggio’s contract hold-out”. https://www.pbs.org/wgbh/americane
xperience/features/dimaggio-contract-hold-out/.
Dodgers and accepted one-year contracts of $125,000 and $110,000,

respectively.13
Data for average salaries by season are not available prior to 1967,
the year in which the MLB Player’s Association began publishing salary
figures. However, Haupert (2012) has used a number of sources to
compile the highest salary for each season. It is unreasonable to assume
that the average salary increased as quickly as the highest salary, never-
theless the highest salaries are instructive. Albert Spalding earned $4000
in 1876 playing for the Chicago White Stockings of the NL (later to
form the Spalding sporting goods company) when the average salary
was approximately $1700 (Kahn, 2000). By 1920, the aforementioned
Ty Cobb topped the salary list at $20,000, while the average salary was
approximately $5000. Jumping forward, Willie Mays’ $125,000 was the
highest salary in 1967, an increase of $105,000, or 525% since 1920.
The average salary in 1967 was only $19,000, an increase of $14,000, or
280%, however not really a king’s ransom. With prices increasing by 67%
over the same period,14 the net increase in the average salary was 213%,
roughly doubling in 57 years.
Multi-year contracts were rare up to 1976 and were only awarded to
the very best players. The presence of the reserve clause in the stan-
dard MLB contract eliminated the need for multi-year contracts since
players had no options to move to other teams except by the decision
of the team owner to either trade or release the player. Owners could
minimize their risk of injured players, or players with declining skills,
by signing players to a series of one-year contracts. The very best star
players could hold out by not reporting to spring training if they had
other lucrative opportunities outside of baseball. Ty Cobb, Babe Ruth,
and Ted Williams were well-known celebrities, often appearing on maga-
zine and newspaper covers, and sometimes in movies and radio. They
may have convinced team owners of available opportunities in corporate
boardrooms or politics. If this were a credible threat to leave the team,
it was in the best interest of the owner to retain the player’s services in
a multi-year contract. However, for all other players, multi-year contracts
were not even a remote consideration.
13 The Dodgers negotiated wisely with Koufax as he retired following the 1966 season
due to arm injuries.
14 https://www.usinflationcalculator.com/inflation/consumer-price-index-and-annual-
percent-changes-from-1913-to-2008/.
1.3 Early Years of Free Agency

The road to free agency in MLB began with the 1970 collective
bargaining agreement (CBA) that allowed for a binding arbitration
dispute settlement mechanism.15 Player’s Association (MLBPA) President
Marvin Miller convinced pitchers Andy Messersmith and Dave McNally
to play the 1975 season without signing their new one-year contracts.
Immediately following the end of the 1975 season, the MLBPA launched
a challenge, arguing that the reserve clause did not apply to Messersmith
and McNally since the players could not be reserved for the 1976 season
having not signed their contracts from the previous season. The arbi-
tration panel decided in favor of the players and deemed the two were
free agents. The owners failed to have the decision overturned in subse-
quent courts, hence the 1976 CBA included provisions for free agency
(minimum six years of MLB service). This case set the stage for a multi-
tude of lucrative multi-year contracts for many players, 32 before the
1977 season and 50 the year after. Four star players left the Oakland
Athletics after winning three World Series championships: Reggie Jackson
inked a four-year contract with the New York Yankees for $525,000 per
season (including bonuses and incentives) in 1977; Joe Rudi signed a five-
year contract with the California Angels for $400,000 per season; Rollie
Fingers signed a six-year contract with the San Diego Padres for $266,667
per season (including bonuses and incentives) in 1977; Bert Campaneris
signed a five-year contract with the Texas Rangers for $190,000 per
season. Free agency allowed for a quick reshuffling of star players to the
teams that valued them the most.
The average salary in MLB increased from $44,676 in 1975 to $1895,
630 by 2000, or 4143% in just 25 seasons.16 However moving players
to new teams via cash sales was effectively eliminated, although not
formally banned. Following the 1975 season, Oakland owner Charlie
Finley reached an agreement with the Boston Red Sox to sell Rollie
Fingers and Joe Rudi for $1 million each. On the same day, Finley sold
Vida Blue to the New York Yankees for $1.5 million. Commissioner
15 The 1970 CBA also established the 10 and 5 rule. Players with at least ten years of
MLB service and the last five years with the same club could veto any trade. This rule
was put in place in response to the Curt Flood legal challenge. Although unsuccessful,
Flood’s challenge to the reserve clause brought attention to the contract strictures.
16 https://www.latimes.com/sports/mlb/la-sp-mlb-salaries-chart-20160329-story.html.
Bowie Kuhn vetoed the deals arguing that the cash sales violated Article
1, Section 4 of the 1921 Major League Agreement that granted the
Commissioner powers to protect the interests and morale of the game. In
response to Kuhn’s decision, Finley refused to play his three star players
but eventually yielded to a threat by the team not to play any games
until the three players were reintroduced to the lineup. Significant cash
sales have been discouraged since the Kuhn decision and none have taken
place.
1.4 Free Agency Since 2000

The 2001 MLB season was a landmark year for media rights. MLB agreed
to grant the FOX network exclusive rights to broadcast MLB games
nationally for six seasons in exchange for $2.5 billion, the largest national
television contract in MLB history. The previous rights deals signed in
1996 paid MLB a total just under $1 billion. In 2006, FOX and TBS
paid MLB $3 billion for television rights to 2012, after which, FOX and
TBS paid MLB just over $6.8 billion for the national television rights
for through the 2021 season. The monies each team received from these
rights deals fueled an accelerated period of free-agent salaries and contract
lengths. The average salary in MLB increased from $1.99 million in 2000
to $4.1 million in 2018, or 106% in 18 years. However, the salaries for
the star players increased much more dramatically. Kevin Brown’s six-year
contract with the Los Angeles Dodgers at a $15.7 million annual salary
topped the list in 2000. Most pundits at the time were shocked at the total
value of the contract, however they did not foresee the even more lucra-
tive contracts to follow. Alex Rodriguez set the standard in 2001 when
he signed a ten-year $252 million contract with the Texas Rangers—the
first ten-year contract in MLB history. After being traded to the New
York Yankees, Rodriguez opted out of his contract in 2007 and re-signed
with the Yankees for another ten years at $275 million. Nolan Arenado’s
eight-year contract signed in 2019 with the Colorado Rockies will pay
him $260 million. Mike Trout set the record for any player contract in
North America, negotiating a 12-year $426 million contract with the Los
Angeles Angels of Anaheim at the age of 27.
1.5 Multi-year Contracts

in Basketball, Football, and Hockey
Prior to the formal striking down of the reserve clause in baseball in
1976, the National Football League (NFL), National Basketball Associa-
tion (NBA), and the National Hockey League (NHL) each had a variation
of the reserve clause in its standard player contract. Like MLB, one-year
contracts were the norm and multi-year contracts a rarity. However, there
were some exceptions for star players. Historical data for the details of
player contracts are scarce outside of baseball, so we provide only a few
examples.
Multi-year contracts were almost nonexistent in the NFL prior to the
1950s. From the league’s inception in 1922 to the 1960s, the NFL paled
in popularity compared to college football. Players needed to find second
jobs in the off-season due to the low salaries in the NFL. Star players in
college typically found it more lucrative to use their popularity to find
employment in business or government, hence the opportunity cost for
playing in the NFL was high. In 1936, only 24 of 81 players drafted
by NFL teams signed contracts. The rules of the game and the playing
schedules were not very formalized, and franchise turnover was frequent.
This resulted in low salaries and a predominance of short-term player
contracts, many players being paid by the game. Harold “Red” Grange, a
star college player at the University of Illinois, was paid a share of the gate
receipts by the Chicago Bears in the late 1920s, thus limiting his income
to the games he played in. Number one draft pick in 1936, Jay Berwanger
took a job as a foam rubber salesman instead of signing a contract to pay
him $150 per game. The merger with the All-American Football Confer-
ence (AAFC) in 1950 solidified the NFL into a set number of teams that
lasted until its merger with American Football League (AFL) in 1970.
Star players, such as Frank Gifford, Johnny Unitas, and others, typi-
cally earned in the $20,000 per year range, but only negotiated one-year
contracts.17 NFL teams avoided multi-year contracts due to the high
probability of injury that could limit a player’s playing career.18 The salary
landscape changed dramatically with the signing of Joe Namath, from the
University of Alabama, by the New York Jets (AFL). Namath’s three-year
17 Average annual income in the United States was about $4000 in 1960.
18 Currently the average playing life is 3.3 years in the NFL. https://careertrend.com/
how-long-is-the-average-career-of-an-nfl-player-3032896.html Accessed May 4, 2020.
deal totaled $427,000 (with a signing bonus, a new car, and other incen-
tives), a pro football record at the time. This contract set the stage for
other star players to sign lucrative multi-year contracts in the years leading
up to the NFL–AFL merger. Heisman Trophy winner O. J. Simpson
signed a five-year deal with the Buffalo Bills for a total of $650,000
in 1969. Competition for players between the NFL and AFL drove up
salaries, particularly for rookie players who also received large signing
bonuses. After the merger in 1970, salary growth and contract lengths
were reduced until the abolishment of the reserve clause in the late 1970s.
The longest contract in NFL history is Donovan McNabb’s 12-year deal
with the Philadelphia Eagles in 2002 for a total of $115 million, and 8- to
10-year deals are not uncommon today for the league’s best star players.
However, NFL contracts are not guaranteed unless a player negotiates
guarantees in the contract.
Prior to the start of the NBA, George Mikan signed a five-year contract
with the Chicago American Gears of the National Basketball League
(NBL) for an annual salary of $12,000, a record at that time for a
professional basketball player. Mikan’s 6-foot 10-inch height made him a
dominant player in the NBL. The NBL morphed into the Basketball Asso-
ciation of America (BAA) that morphed into the NBA. The NBA played
its first season in 1949, making it the youngest of the four major profes-
sional sports leagues in North America. The 1950s witnessed unstable
franchises and low attendances for most teams. Like the NFL, the NBA
competed with the college game and was a distant second in popularity.
Bob Cousy of the Boston Celtics was one of the best players in the young
league, but only negotiated a one-year deal for $26,000 in 1959. Bill
Russell and Wilt Chamberlain each earned $100,000 for the 1966 season,
but only on one-year contracts. It was not until competition with the
upstart American Basketball Association (ABA) in 1968 that contracts
increased in value and length, despite the reserve clause. The move to the
new league started with the case of Rick Barry who signed a three-year
contract with the ABA Oakland Oaks worth $500,000 plus a 15% owner-
ship share and 5% of gate revenue. Other new players followed to the new
league for lucrative multi-year contracts, including Julius Erving (4 years,
$500,000), Artis Gilmore (10 years, $2.5 million) and David Thompson
(3 years, $1.45 million). Salaries and contract lengths continued to climb
after the NBA–ABA merger in 1979.
Kobe Bryant, Shaquille O’Neal, Tim Duncan, and Chris Webber all
share the longest contracts in NBA history at seven years. Bryant’s deal
with the Los Angeles Lakers paid a total of $136.4 million commencing
in 2004, with O’Neal’s contract with the Lakers totaling $120 million in
1996. Tim Duncan signed a seven-year deal worth $122 million in 2003
with the San Antonio Spurs. The Sacramento Kings inked Chris Webber
to a deal worth $123 million in 2001. However, like the NFL, the NBA
is not friendly to longer-term contracts, perhaps due to the uncertainty
of injury and the small roster size for each team (17 players) relative to
the number of available players in college to be drafted. Many NBA star
players negotiate four- and five-year deals, but few longer than that.
The NHL (incepted in 1917) has a better-documented history than
the NFL or NBA. Again, one-year contracts were the norm in the league,
with a few exceptions for star players. Maurice Richard signed a two-
year contract with the Montreal Canadiens in 1943 worth $5000 per
season (with performance bonuses) and signed two-year deals until the
mid-1950s. The Montreal Canadiens also signed Jean Beliveau in 1953
to a three-year contract worth a total of $100,000, a large amount for its
times. Beliveau had a strong business relationship with Molson Breweries
in Montreal and could have left the Canadiens to assume an executive
position, hence his opportunity cost to play hockey was significant and
the Canadiens knew it. Bobby Hull negotiated a five-year contract with
the Chicago Black Hawks (NHL) in 1960 for $20,000 per season. The
Detroit Red Wings signed Doug Harvey to a three-year contract as a
player-coach for $27,000 per season. Bobby Orr, one of hockey’s greatest
players, signed a three-year with the Boston Bruins in 1968 for $50,000
per season. The establishment of the rival World Hockey Association
(WHA) in 1972 created a higher level of competition for star players
that increased salaries and contract lengths rapidly. The most notable of
these being the ten-year, $1 million contract signed by Guy Lafleur with
the Montreal Canadiens, the largest NHL contract up to that time, and
the ten-year $2.75 million contract signed by Bobby Hull with the WHA
Winnipeg Jets.
The longest multi-year contracts we could find in the four professional
sports leagues in North America were the 15-year NHL contracts signed
by Rick DiPietro of the New York Islanders in 2006 and Ilya Kovalchuk
with the New Jersey Devils in 2010. DiPietro was to receive $4.5 million
per season, however the Islanders bought out his contract in 2013 due
to his frequent injuries. Under the terms of his buyout, DiPietro will
receive $1.5 million per year until 2026. Kovalchuk turned down a 12-
year $101 million contract with the Atlanta Thrashers to sign with the
Devils for a total of $100 million. The Devils initially agreed to a 17-year
contract, but the terms of the deal violated the NHL collective bargaining
agreement and was disallowed. Just three years into the deal, Kovalchuk
retired from the NHL to play hockey in his native Russia.
1.6 Player Salaries and Marginal

Revenue Products: the Scully Method
The workhorse method to estimate MRPs of professional athletes paid
on a salary basis is Scully (1974), hereafter referred to as the “Scully
method.” The method requires estimation of a revenue function and a
team production function for the team the player is employed by. These
can be specified in general form as Ri = R(wi , Ai ) and wi = w(X i ), where
Ri is local revenue for team i, wi is the team winning percentage, Ai is
a vector of market-specific variables, and X i is a vector of team perfor-
mance measures that are proxies for the stock of team talent ti . These
two functions should have the characteristics ∂ Ri /wi > 0, ∂ 2 Ri /∂wi2 < 0,
∂wi /∂ X i > 0 and ∂ 2 wi /∂ X i2 < 0 for concavity. The reserve clause was
still in place at the time of Scully (1974) therefore a monopsony model
of talent acquisition was quite appropriate. The monopsonist is the only
buyer of talent and will acquire talent up to the point where the increase
in team payroll (the marginal resource cost [MRC]) is just equal to the
marginal revenue product. Note that the owner is buying units of talent
that are continuous, not discrete numbers of players since MLB teams
have fixed roster sizes of 25 players. An owner cannot buy additional
players but can acquire a fixed roster of players that possess a greater stock
of talent. The supply curve of talent is assumed to be upward sloping
due to talent scarcity, so the MRC is above the supply curve of talent in
Fig. 1.2. Technically, the profit maximization problem in general form is
M AX Ri = R(wi (ti ), Ai ) − Z i (ti )ti − FCi (1.1)
The wage rate per unit of talent is Z i and fixed costs are FCi . Maxi-
mizing profit with respect to stock of team talent gives the first-order
condition

∂ Ri ∂wi d Zi
− ti + Z i (ti ) = 0 (1.2)
∂wi ∂ti dti
Fig. 1.2 Monopsony model of talent acquisition (Source Author’s creation)
The first term in (1.2) is the MRP of talent and the bracketed term is
the MRC that includes the wage rate per unit of talent plus the increase
in the wage rate paid to all units of talent when one more unit of talent
is acquired (the total being the MRC). Figure 1.2 demonstrates that a
profit-maximizing owner will page a wage rate per unit of talent Z i when
acquiring a team stock of talent Ti , much less than the MRP, hence players
are “exploited.” It is important to note that it is a mistake to conclude
that each player will be paid his MRP in the absence of monopsony
power. Each player contributes a different MRP of each unit of talent,
but the wage rate per unit of talent is determined by the last unit of
talent acquired. The player’s salary is then the profit-maximizing value of
Z i multiplied by the amount of talent units ti j that player j contributes to
the team stock of talent. More talented players will earn higher salaries,
but in equilibrium, the MRP is the same for all units of talent and thus
players. If all players earned their MRPs, there would be no surplus left
for the owner, however in the case of the monopsonist, the owner earns
an additional area given by (M R Pi − Z i )Ti .
Assuming specific functional forms and estimating slope coefficients for
the independent variables allows for the computation of the MRP for each
player on the roster. The first step is to estimate the team revenue func-
tion. This is straightforward given the availability of data. Unfortunately,
comprehensive revenue data was difficult to obtain 50 years ago, and the
Forbes magazine estimates were not produced at that time. Scully (1974)
computed gate revenue by multiplying an average ticket price by atten-
dance for each team, then added the share of national broadcast revenue.
This introduces two problems. First, the average ticket price should be a
weighted average of ticket prices, the weights being the share of seats in
each seating section. It is not clear if the original Scully (1974) study used
weighted average ticket prices or just some sort of other average. Second,
Scully admits that his revenue data were not post-revenue sharing. Up to
the late 1990s, visiting clubs received a fixed share of gate revenue in the
American League and a fixed price per ticket sold in the National League.
Team revenues that are available to the owner to pay players will be over-
estimated for teams above the average team revenue and underestimated
for teams that fall below. This is true whether the revenue sharing formula
is the current pooled system, or the old gate-sharing system (Rockerbie
2009a), however in the latter system, the ranking of each team in the
league revenue hierarchy can change using team revenue after sharing.
Scully specified a revenue function and a winning percentage function
that are each linear in their components. This could be an issue if one is
concerned with concavity of the revenue function (diminishing marginal
revenue), however if the winning percentage function is nonlinear and
concave, a downward MRP schedule is still assured. The A-vector in the
Scully method is composed of the same independent variables that we
employ in Chapter 2 and summarized in an appendix.19 Scully found that
a one percentage point increase in the team winning percentage was esti-
mated to increase team revenue by $10,330 and the regression model fit
reasonably well with an adjusted R2 of 0.75.
The winning percentage function is also linear in its components with
the same independent variables that we employ in Chapter 2 and summa-
rized in an appendix. The fit was quite good with an adjusted R2 of 0.88
and slope coefficients for the team slugging average and team strikeout to
walk ratio of 0.92 and 0.90, respectively. The linearity of the revenue and
winning percentage functions result in a perfectly elastic MRP schedule, a
problem that has been corrected in later studies described in Chapter 2.20
19 In addition to the independent variables used in Scully (1974), we include dummy

variables for new television contracts.
20 We specify a logistic winning percentage function in Chapter 2.
Each player’s contribution to the team winning percentage is

computed in a simple way. For hitters, each player’s slugging percentage
(total bases/total at bats) is multiplied by 0.0833 since Scully assumed 12
positional players on a roster. In this way, each player contributes to the
team slugging average with an equal weight. In reality, the team slugging
average that is computed by MLB and reported in statistical services is
the total team bases/total team at bats, so the Scully method is not accu-
rate unless Scully computed his own team slugging average (he did not
discuss this point). The appropriate weighting is the share of team at bats
that we use in Chapter 2. The Scully method computes the contribution
of each starting pitcher by assuming eight starters on the team roster and
then multiplying each pitcher’s strikeout to walk ratio by 0.125, on the
basis that this is the contribution to the team average strikeout to walk
ratio. This method suffers from the same weighting issue as for hitters. A
better weight is the number of innings pitched or the number of batters
faced. When the marginal value of a win is $10,330, for a hitter with a
good 0.350 slugging average, the contribution to the team revenue (the
MRP) is 0.0833(350)(0.92)($10,330) = $277,077.45. A starting pitcher
with a good strikeout to walk ratio of 2 contributes an MRP equal to
0.125(2)(0.90)($10,330) = $232,425.
Scully (1974) estimated the MRP of every positional player and
starting pitcher in MLB for the 1968 and 1969 seasons. Unfortunately,
salary data for only 88 players in the sample were available so Scully
estimated a log salary regression model with past performance statistics
as independent variables. This allowed for the comparison of estimated
net MRPs with estimated salaries. The Scully method deducts estimates
of player development costs (10% of gross MRP) from the gross MRP
figure to arrive at a net MRP. Rather than present results for each player,
Scully grouped players into performance categories (mediocre, average,
and star) using arbitrary hitting and pitching criteria. As one would
expect, mediocre players were the most exploited, while the star players
were paid estimated salaries close to their MRPs.
1.7 Real Options

Scully (1974) found that all MLB players were paid a salary below
their estimated MRP during the reserve clause era. Free agency appears
to have changed the salary–MRP relationship so much so that recent
studies described in Chapter 2 suggest that many players are paid a salary
above their estimated MRP. This does not fit with the standard monop-
sony model. One explanation is that some star players bring more to a
team than just their on-field performance. Acquiring a superstar player
could result in higher local media revenue, apparel revenue, promotion
revenue, or even result in a team developing a world brand. In other
sports, there are players who bring a sort of cachet to an organiza-
tion to which fans respond. Christiano Ronaldo and Neymar’s move to
Juventus and Paris-St. Germain in European football came with record
transfer fees (112 million euros and 222 million euros) that give an indi-
cation of their worth to their clubs beyond their performance on the
field. David Beckham’s move from Real Madrid to LA Galaxy brought
tremendous exposure to a previously mediocre Major League Soccer in
North America. The top European clubs can earn large bonus revenues
from competing deep into the annual Champion’s League competition.
Shirt sales often garner enough revenue to recover a significant portion
of salaries for these superstars.
Today there are few examples in MLB of these players whose popu-
larity extends well beyond their performance on the field. Babe Ruth,
Joe DiMaggio, and Ted Williams are players from the distant past now.
The skew in the salary–MRP relationship is more systematic in MLB,
suggesting that there is an important feature of bargaining that did not
exist in the reserve clause period. One factor at the aggregate level is
that the talent market cannot be characterized as a monopsony for free-
agent players. Competitive bidding between clubs drives a player’s salary
much closer to their expected MRP and could drive it above that if
owners are willing to bid away some or all of their expected surplus
(Rockerbie 2009b). This sort of “winner’s curse” argument could hold
for the very best free agents, but not likely for the large number of free
agents observed to be “overpaid” in Chapter 2.
References
Haupert, M. (2012). MLB’s annual salary leaders since 1874. Society for Amer-
ican Baseball Research (SABR) Newsletter. https://sabr.org/research/mlbs-
annual-salary-leaders-1874-2012.
Kahn, L. (2000). The sports business as a labor market laboratory. Journal of
Economic Perspectives, 14(3), 75–94.
Rockerbie, D. (2009a). Free agent auctions and revenue sharing: A simple
exposition. Journal of Sport Management, 23(1), 87–98.
Rockerbie, D. (2009b). Strategic free agency in baseball. Journal of Sports

Economics, 10(4), 278–291.
Scully, G. (1974). Pay and performance in major league baseball. American
Economic Review, 64(5), 915–930.
Zimbalist, A. (2011). Circling the bases: Essays on the challenges and prospects of
the sports industry. Philadelphia: Temple University Press.
CHAPTER 2
The Puzzle of Overpaid and Underpaid Players
Abstract To use real options theory to explain the salary anomaly in

Major League Baseball, it is necessary to estimate a player’s marginal
revenue product (MRP). In this chapter, we specify a team revenue model
and winning percentage function to specify a player’s MRP in theory.
We then estimate an MRP for a large sample of free-agent players over
the 2000–2012 baseball seasons. Our method is novel in that it utilizes
features not found in previous published studies. The empirical estimates
of player surpluses confirm the salary anomaly result for comparisons
relative to the league average salary and by single- or multi-year contracts.
Keywords Marginal revenue product · Surplus · Logistic · Monopsony
2.1 Introduction
One of the important issues in the sports economics literature is whether
players are paid according to what economic theory predicts: profit-
maximizing firms pay a salary equal to a player’s marginal revenue product
(MRP). Scully (1974) tackled the problem decades ago during the last
remnants of the reserve clause era in MLB. His method and results are
reviewed in the last chapter, and his paper sparked a burgeoning literature
© The Author(s) 2020 23

D. W. Rockerbie and S. T. Easton, Contract Options for Buyers
and Sellers of Talent in Professional Sports, Palgrave Pivots in Sports
Economics, https://doi.org/10.1007/978-3-030-49513-8_2
on the subject. More recent contributions include Scully (1989), Brug-

gink and Rose (1990), Fort (1992), Zimbalist (1992), MacDonald and
Reynolds (1994), Leeds and Kowalewski (2001), Bradbury (2007), and
Berri et al. (2015). Each of these papers utilizes a two-equation regres-
sion model to estimate the appropriate salary given an individual player’s
production on the field by looking first at the relationship between team
performance and revenue, and then estimating the relationship between
player performance and team performance. The method places strains
on the data, although it is not computationally burdensome. Kraut-
mann (1999) developed a simpler, more direct approach that requires less
data and utilizes only a single regression equation. This has become the
method of choice over the last few years (Krautmann and Ciecka 2009;
Brown and Jepsen 2009, to name a few), however it does not provide
an estimate of a player’s MRP, rather it is a method to predict a play-
er’s salary assuming that the MRP rule holds (see Bradbury [2013] for a
critique of the method).
Previous studies have estimated MRP in an ex ante sense, that is, they
attempt to estimate how a club owner forms an expectation of a player’s
future MRP from past performance data. Utilizing the Scully method,
this MRP expectation for a particular season is then compared to the
actual salary a player earned to determine if this is consistent with the
expectation. The method does not compare the negotiated salary (which
assumedly is the expectation the owner has formed plus any monopsony
rent plus any additional surplus the player has been able to negotiate from
the owner) with the realization of the player’s future MRP. Abstracting
from player injuries, the ability to accurately forecast player MRPs is an ex
post test of the MRP rule. An innovation in our analysis is to compare the
negotiated salary for free agents with their subsequent discounted MRP
in the following seasons of the contract period. By doing so we discover
an anomaly. Future MRP as traditionally estimated is not well indicated by
future salary. In the subsequent chapter we explore why this may occur.
The purpose of this chapter is to revisit the question of whether base-
ball players are paid their MRP in an ex post sense, by improving on past
studies. Our method follows Scully (1974) in flavor but introduces several
improvements. Accurate and reliable financial data for MLB clubs is more
available in 2020 than it was in the 1980s, and a consistent time series
2 THE PUZZLE OF OVERPAID AND UNDERPAID PLAYERS 25
of estimates has been provided by Financial World and Forbes maga-

zine since 1990.1 Data limitations for other variables in our approach,
however, constrain us to begin our analysis in 2000. Nonetheless, we
use a longer sample period than has been used in previous studies and
utilize average ticket prices in our revenue function. We estimate our
winning percentage function using a logistic regression, rather than the
standard linear probability model used in the past. Finally, rather than
estimate MRPs for every player, we estimate MRPs only for players who
signed contracts with new clubs after free agency. In this way, we isolate
the effect that the expectation of future performance has on the negoti-
ated salary. We compare the newly negotiated (average) annual salary of
each free-agent position player (excluding pitchers) to an estimate of the
(average) present value of the marginal revenue products over the lifetime
of the new contract.
More broadly, we explore how workers are paid in long-term contracts
by focusing on an industry in which MRP’s can be estimated and
compared to actual salaries: professional baseball players. Although a
traditional labor market demand-supply approach (Rosen and Sanderson
2001) results in the MRP = salary profit-maximizing condition, the
approach does not account for observed systematic failures in the condi-
tion for specific players other than to resort to untested monopsony
models or vague discussions of the bargaining power of players. Our
sample includes 678 free agents (excluding pitchers) who were signed
to MLB contracts over the 2000–2012 MLB seasons. The free-agent
market was more active in some seasons than others. Summary statis-
tics of free-agent activity by fielding position appear in Table 2.1. When a
player played in multiple fielding positions during the season immediately
following their free-agent signing, the most frequently played position
was used in the table.2
Our results in this chapter suggest that significant changes have
occurred in the MRP = salary relationship since the 1980s. We find
1 These revenue data are not without their limitations. See Zimbalist (2010) for an
in-depth analysis of salary and revenue data. Forbes revenue estimates are net of revenue
sharing of local revenues and net of debt payments on stadiums paid by the teams. See
Ouzanian and Badenhausen (2020) for a brief description of the methodology of revenue
calculations and team valuations.
2 The Designated Hitter (DH) position for American League teams was not used since
most teams did not employ a full-time DH, rather they split time among several players
at different fielding positions with the DH position.
Table 2.1 Free-agent fielding players by playing position, 2000–2012
Season Catcher 1B 2B 3B SS OF Total
2000 8 3 6 3 5 9 34
2001 7 5 4 5 4 13 38
2002 6 5 5 4 5 21 46
2003 7 7 6 5 3 14 42
2004 12 9 9 8 7 32 77
2005 11 7 6 5 9 19 57
2006 8 9 7 5 6 16 51
2007 9 9 14 10 3 25 70
2008 7 3 4 5 3 14 36
2009 7 8 5 4 7 17 48
2010 12 11 8 7 7 19 64
2011 15 8 4 6 6 19 58
2012 9 9 9 3 7 20 57
Total 118 93 87 70 72 238 678
Source Author’s research
that higher-priced free agents, defined as those paid above the league
average salary for all players, tend to be overpaid (or their expected MRP
overestimated) while other free agents are underpaid or paid appropri-
ately. Berri et al. (2015) found that NBA players were overpaid over
the 2001–2011 sample period, resulting in negative surpluses for team
owners. They suggest that the MRP rule works to some extent and that
players earn part of the owner’s surplus that arises from fixed revenues
through a bargaining process. Krautmann and Solow (2018) found that
most MLB free agents were overpaid using a different methodology to
estimate player MRPs. These results contrast sharply with the belief that
the player’s market is a monopsony (Humphreys and Pyun 2017; Leeds
and Leeds 2017). Krautmann et al. (2009) found that players with low
bargaining power tend to be underpaid in the National Football League,
the National Basketball Association, and MLB. Our empirical results in
this paper substantiate this claim and provide a number of further insights,
the most important being that free agents signing multi-year contracts
are overpaid relative to those signing one-year contracts. In Chapter 4 we
provide a theoretical structure to help explain this apparent anomaly.
2.2 A Revenue Model

The essence of the Scully (1974) method is that an improvement in player
performance increases a club’s winning percentage at the expense of other
clubs, which in turn draws more fans to their games and increases the
club’s revenue. In our specification, the revenue function for a single club
is given by:
R = X α (W P)β (2.1)
where
R = attendance revenue + television revenue + all other revenues

WP = season winning percentage
X = a vector of independent variables that affect revenue
Total revenue includes revenue that is specific to each club (attendance,

local television, concessions, etc.) and revenue that is shared by all clubs
(national television, apparel, etc.). Although some of these revenues may
not be a function of local player talent directly (such as national televi-
sion revenue), increases in these revenue sources can increase the MRP
of talent by shifting the MRP schedule for each club. This is the break-
down of team revenues provided by Forbes Magazine. The only way to
obtain more detailed (and accurate) revenue figures is to take them from
the team accounting statements. However, these are generally unavailable
with a few exceptions and include only a few seasons.3
The vector X includes per capita income (total metropolitan
income/metropolitan population) Y , a state unemployment rate U , a
weighted average real ticket price P, a dummy variable, NEWSTA, taking
on the value one for a period of five seasons after a new stadium is
opened4 (Coffin 1996; Clapp and Hakes 2005; Coates and Humphreys
3 Some of these are the Blue Ribbon Report and follow-up report commissioned by
MLB in 1999 and 2000. Deadspin leaked financial statements for several MLB teams for
specific seasons but not for long enough periods of time to be used in our study. See
https://sites.google.com/site/rodswebpages/codes for all of the Deadspin releases.
4 New or extensively renovated stadiums since 2001–2012 included Cincinnati, Miami,
Milwaukee, New York (NL), New York (AL), Philadelphia, Pittsburgh, San Diego, St.
Louis, Kansas City, and Minnesota. New stadiums opened before the start of the 2000
season were not included, even though the 5-year period extended into the sample period.
2005), and three dummy variables (TV1, TV2, and TV3) that take on
the value one in each season during or after a new national television
contract (2001, 2007, and 2012, respectively). Finally, team revenue, per
capita income, and the average ticket price were deflated by an individual
state consumer price index (2000 = 100). A list of sources for all variables
is contained in Table 2.2.
The revenue function in (2.1) is a departure from typical linear revenue
functions utilized in the literature, however we see it as an improve-
ment. Revenue functions that are linear in the winning percentage assume
strong separability and a completely elastic marginal revenue schedule.
The more general revenue function in (2.1) allows for a downward
sloping marginal revenue schedule that shifts with changes in the inde-
pendent variables. Since market conditions can differ between cities and
can change over time, it is reasonable to assume that the MRP schedule
is not constant.
The function relating team winning percentage to player performance
is given by the logistical function:
ln[W P/(1 − W P)] =δ + θ O P S + μ(K /B B) + π E R A+

ρC O N T + ωOU T + ε (2.2)
where
OPS = team slugging percentage (total bases/total at bats) +

team on-base percentage
K/BB = team strikeout–walk ratio
ERA = earned run average (earned runs surrendered by a
pitcher/9 innings pitched)
CONT = a dummy variable equal to one if a team finishes within
five games of the wild-card winner
OUT = a dummy variable equal to one if a team finishes twenty or
more games behind the wild-card winner
ε = a normal error
These stadiums included Detroit, Houston, San Francisco, and Seattle. It was thought best
not to include these as the variable would not capture what their revenues were prior to
the new stadium opening.
Table 2.2 Estimates of the (natural logarithm of the) revenue function
R = MLB club revenue from all sources Taken from http://www.rodneyfort.com

which is data compiled from Forbes
magazine and Financial World magazine
P = Average ticket price Taken from http://www.rodneyfort.com
which is data compiled from Team
Marketing Report
WP = Club winning percentage Taken from Sports Reference, LLC
http://www.sports-reference.com
Y = Metropolitan area income per capita Local metropolitan area personal income
and population taken from Bureau of
Economic Analysis, Regional Accounts,
http://www.bea.gov/regional/reis/
U = State-wide unemployment rate Taken from Bureau of Labor Statistics,
http://www.bls.gov/schedule/archives/
metro_nr.htm
CPI = State-wide consumer price index
(2000 = 100)
NEWSTA = dummy variable = 1 for a new Determined from http://www.ballparks.
stadium com/
SA = Team slugging average (total bases/at Taken from Sports Reference, LLC
bats) http://www.sports-reference.com
K/BB = team strikeout to walk ratio Taken from Sports Reference, LLC
http://www.sports-reference.com
CONT = a dummy variable = 1 if a club Determined from Sports Reference,
finished the season within five games of the LLC http://www.sports-reference.com
wild-card winner in each league.
OUT = a dummy variable = 1 if a club Determined from Sports Reference,
finished the season within twenty or more LLC http://www.sports-reference.com
games behind the wild-card winner in each
league
TV1, TV2, TV3 = 1 for the seasons covered
by a new national television contract (2001,
2007, and 2012, respectively)
Source For data in regression models
The independent variables OPS, K/BB, and ERA are measures of hitting
and pitching quality that have been justified by Scully (1974) and others
elsewhere. The independent variables CONT and OUT were also used by
Scully (1974) and others as proxies for team morale or perhaps manage-
rial ability. They represent a shift in the team production function based
on factors other than player performance. We have retained their use to
allow for comparison of our results to previous work. An earlier version
of the model also included fielding percentage (outs per fielding attempt)
as an independent variable in (2.2). It was omitted here as it lacked statis-

tical significance in any of the formulations. The sabermetric literature
suggests many different measures of player performance, such as Wins
After Replacement (WAR) and Wins After Replacement Plus (WARP),
among others.5 One could argue that some of these should be included as
independent variables in (2.2). We think otherwise for two reasons. First,
many of these variables are measures of a player’s output from a team
production function, instead of being an input into a team production
function. Using an independent variable like WAR would imply regressing
a team’s winning percentage on a measure of the contribution to wins
from each player, suggesting that some sort of adding up property should
apply across the members of the team. We view WAR as an intermediate
output, rather than an input. Second, it is difficult to judge the marginal
effect on winning percentage from a change in a sabermetric composite
statistic. The marginal change in OPS is well understood and the deriva-
tive to calculate a player’s MRP is straightforward. Although imperfect,
the contribution to the team is also easily calculated. Many of the saber-
metric statistics are measured in units that are not easily translated to a
team contribution, making the calculation of MRP difficult.
The fact that winning percentage is bounded between zero and one
justifies the use of the logistic function. Previous studies have relied on a
linear probability model for winning percentage. The predictions for the
marginal effect on winning percentage require estimation of the param-
eters and the choice of a baseline winning percentage (∂ W P/∂ O P S =
θ W P(1 − W P)). Teams with a 0.500 winning percentage will show the
largest marginal effect, holding player performance and team morale
constant. We chose to use the fitted value for the current season’s winning
percentage for each team as the baseline winning percentage. This places
each team on a logistic function at different positions with different
marginal effects. This could not be done with a linear probability model
and we think it is an improvement.
Taking the natural logarithm of (2.1), the calculation for MRP is
straightforward once (2.1) and (2.2) have been estimated. For a specific
hitter denoted as player i playing on team j, MRP is given by (where AB
denotes at bats):
M R Pi = (∂ Ri /∂ W Pi )(∂ W Pi /∂ O P Si )d O P Si j (2.3)
5 A useful reference is Baumer and Zimbalist (2013).

Totally differentiating (2.1) gives
1 α β
dR = dX + dW P
R X WP
The vector X is not a function of WP by assumption so dX = 0. Multi-
plying by R and dividing by dOPS gives the first required derivative in
(2.3), d Od RP S = R WβP ddOWPPS . Totally differentiating (2.2) and focusing only
on the OPS term gives

1−WP 1
dW P = θ d O P S
WP (1 − W P)2

1
dW P = θ d O P S
W P(1 − W P)
dW P
= W P(1 − W P)θ
dOPS
Inserting the necessary derivatives into (2.3) gives

ABi
M R Pi j = β θ R j 1 − W P j O P Si j (2.4)
AB j
The contribution to the team OPS by a specific hitter is calculated

using the hitter’s total at bats as a share of total team at bats multi-
plied by the hitter’s OPS for a season.6 If a hitter has an OPS of 0.750
and accounts for 10% of the team’s total at bats, his contribution to the
team OPS is 0.075. This differs from the method used by Scully (1974)
and others where the player’s contribution to the team OPS is the team
OPS multiplied by the share of team at bats of the individual player.
That measure might not be reflective of the player’s performance so we
use the player’s OPS since it provides individual information of perfor-
mance.7 The method assumes that the team performance is simply the
6 OPS is on-base percentage plus slugging percentage. The former is based on total plate
appearances, while the latter is based on total at bats. Total plate appearances includes
walks, while total at bats does not. We chose to use total at bats for the player contribution
to the team OPS, although we admit the choice is arbitrary.

7 Since the team OPS is calculated by T eam O P S = N (AB /T eam AB) × O P S
i=1 i i
we use the right-hand side value for each player as the player’s contribution. The Scully
linear summation of the individual player performances and ignores other

aspects of team production such as complementarities.
We found that the estimates for gross MRP for pitchers were very
unreliable when compared to actual salaries, so we do not report them.
Pitchers can be divided into three categories: starters, middle relievers,
and relievers. Starting pitchers are paid far less per inning pitched than
relief pitchers and this makes the results very hard to interpret. If the
samples of free agents were evenly divided between starters and relievers,
the differences might cancel out in the overall average gross MRP,
however our sample years of free agents were typically heavily weighted
toward starting pitchers or relief pitchers. Zimbalist (1992) also excluded
gross MRP estimates for pitchers on similar grounds. We also discovered
that a large proportion of free-agent pitchers spent a considerable amount
of time on the 60-day disabled list in the first season of their new contract.
Consequently, their gross MRP was smaller than expected, but not due
to their own performance or the mistake of the owner.
2.3 Estimates of the Revenue

and Winning Percentage Functions
The revenue function in (2.1) was estimated in natural log form using
data collected from the 2000 through 2012 MLB seasons, providing a
total pooled sample size of 364 observations per variable. The Montreal
Expos and Washington Nationals were excluded from the sample. It was
thought that the franchise move from Montreal to Washington for the
2005 season could distort the regression results. The Toronto Blue Jays
were also excluded since there were substantial swings in the Canadian-
U.S. Dollar exchange rate that affected the club’s revenues measured in
U.S. Dollars.
Total revenue for each club was taken from Forbes magazine and
includes revenue from all sources including net revenue sharing from
national media rights and revenue sharing of local revenues. Although
it would not be difficult to reverse the effects of revenue sharing on
team revenues using the revenue sharing formula for each season, it was
method uses (ABi /T eam AB) × T eam O P S which does not utilize any information about
the individual player and is not reflective of how the team OPS is calculated. The original
Scully (1974) paper used team batting average, not team OPS, but the team calculation
is the same.
left in place on the reasoning that each owner can more or less predict
team revenues after revenue sharing fairly accurately and base their talent
acquisition and payroll decisions on available revenues after sharing. Team
fixed effects were included to account for variations in team revenue
not explained by the independent variables in (2.1). Team fixed effects
were included, but year fixed effects were not.8 A weighted least squares
method was used to account for heteroskedasticity across teams. Although
Scully (1974) included a dummy variable to distinguish between National
League and American League teams, we did not. The National and Amer-
ican League used the same revenue sharing formula over our sample, so
we see little reason to include the dummy variable in (2.1). After testing
for and rejecting nonlinearity, we report the linear regression results (t-
ratios appear in parentheses) in Table 2.3. To test for robustness of the
team revenue model, results are also reported for the 2000–2008 and
2004–2012 subsamples.
The team fixed effects coefficient estimates are omitted from Table 2.3
for the sake of brevity. The high significance of the F-statistic and the very
acceptable adjusted R2 suggests that the log-linear model is appropriate
for explaining variations in team revenue. All the regression coefficients
are statistically significant at 95% confidence with the exception of the
state unemployment rate. The unemployment rate is a proxy for economic
conditions and the results suggest that MLB team revenues are not sensi-
tive (a-cyclical) to the business cycle. The team winning percentage is
the key variable in the marginal revenue calculation and its elasticity esti-
mate of 0.523 suggests that a ten percent increase in winning percentage
(say from 0.5 to 0.55) raises team revenue by about 5.23% indicating
that revenue is quite inelastic with respect to winning percentage. Given
that approximately 90% of teams in any given season fell within winning
percentages of 0.4 and 0.6, large revenue swings due to team performance
are not a frequent occurrence in MLB.
A new stadium has no significant effect on team revenue, but has a
significantly negative effect on revenue in the 2004–2012 seasons. This
result is counter to the short-run effects on attendance found by Coffin
(1996), Clapp and Hakes (2005), and Coates and Humphreys (2005).
8 The inclusion of year fixed effects created vectors of zeroes and ones whose combi-
nation was perfectly correlated with the combination of vectors from the NEWSTA, TV1,
TV2, and TV3 independent variables. Rather than omit one of these variables, it was
thought best to omit the year fixed effects.
Table 2.3 Estimates of

2000–2012 2000–2008 2004–2012
the (natural logarithm of
the) revenue function Constant 13.755 13.136 15.727
(2.1) (16.51)∗ (12.53)∗ (30.01)∗
WP 0.523 0.565 0.256
(6.82)∗ (7.21)∗ (6.01)∗
Y 0.873 1.023 0.493
(3.79)∗ (3.46)∗ (3.73)∗
U 0.252 0.348 0.267
(0.61) (0.25) (1.05)
P 0.457 0.476 0.368
(14.75)∗ (16.59)∗ (10.47)∗
NEWSTAD −0.009 0.037 0.025
(0.65) (1.29) (2.24)*
TV1 0.147 0.137
(6.89)∗ (4.64)∗
TV2 0.292 0.277 0.101
(11.74)∗ (7.75)∗ (4.76)∗
TV3 0.364 0.175
(17.44)∗ (9.02)∗
Adjusted R2 0.923 0.922 0.946
F 125.88∗ 88.46∗ 130.29∗
N 364 252 252
Note t-statistics appear in parentheses

∗ denotes statistical significance at 95% confidence
Source Author’s creation
However, Rockerbie and Easton (2019) find the same result on team
performance using a much longer sample period and a model that incor-
porates the availability of talent. The lack of statistical significance is likely
due to the inclusion of the real average ticket price in the team revenue
function. The previously cited papers did not include the real average
ticket price. A regression of the real ticket price on the opening of a
new stadium suggests that the average real ticket price increased by an
average of 20.8% (p-value = 0.001) in the first season. If ticket demand
is inelastic, the revenue increases noted by the previously cited papers will
show up as an increase in the average ticket price. Removing the real
average ticket price from the regression of (2.1) reveals a coefficient for
NEWSTA of 0.088 with a t-statistic of 4.15, suggesting that real team
revenue increased by 9.2% following the construction of a new stadium.
We chose to include the real average ticket price since it provides a useful
elasticity estimate.
Real per capita income has the largest positive effect on revenue
with an estimated elasticity 0.873, confirming that baseball’s composite
commodity is a normal good. Ticket price has a positive effect on revenue
that is inelastic (0.457), confirming that marginal revenue is less than the
ticket price. Each new national television contract increased team revenue
independently of the other variables, with percentage increases of 0.158,
0.339, and 0.439, respectively.9
The coefficients for the 2004–2012 subsample are all smaller than
those for the 2000–2008 subsample, although the statistical significance
is virtually unchanged. This can be explained by the declining annual real
revenue growth that occurred over the 2000–2012 seasons. Revenues
grew by an annual average of 7.3% in 2000–2008 but fell to 5.6% in
2004–2012. With the weighted average ticket price (P ) held constant,
revenue growth could have slowed from a decrease in attendance growth
and/or decreases in the growth of other revenue sources (concessions and
other local revenues).
The winning percentage function (2.2) was also estimated using fixed
effects and weighted least squares to account for heteroskedasticity. The
model was estimated separately for the National and American Leagues
owing to the use of the designated hitter in the American League. Team
hitting statistics are somewhat higher and pitching statistics are somewhat
lower for the American League relative to the National League where
pitchers are required to hit. The results for each league are given in
Table 2.4. Again, the fixed effects estimates are omitted for the sake of
brevity.
The degree of fit was quite high for both leagues and the F-statistic
for each regression suggests that the model specification is appropriate.10
All of the coefficients carried the expected signs and all of the coefficients
were statistically significant with the exception of the strikeout–walk ratio
for the National League. The marginal effects can be found dividing each
coefficient by 4 since the linear marginal effect is given by βi W P(1 − W P)
and W P = 0.5 over the entire sample. Since we focus on the perfor-
mance of hitters, the OPS variable is of key interest. A 100 point increase
in team OPS is predicted to increase the team winning percentage by
9 The effect on revenue expressed as a percentage change when using a dummy variable
is eb − 1 where b is the slope coefficient.
10 This measures the degree of fit for the nonlinear dependent variable, but not for the
predicted winning percentages themselves.
Another random document with
no related content on Scribd:
hauskoja muistoja noilta vuosilta, jolloin opetin sinua, ja olet sanonut
minulle useammasti kuin kerran, että elämäsi iloisin päivä tulee
olemaan se, jolloin Heven maakunnan jokainen asukas osaa lukea
ja kirjoittaa».
»Nyt olen kumminkin toista mieltä», sanoi András hieman

töykeästi, »ja silloin kun olette ruvennut puhumaan minulle tuosta
asiasta, olen kieltäytynyt keskustelemasta siitä kanssanne. Niin, olen
muuttunut suuresti sitten noiden päivien, jolloin te Bideskuty’n
kauhean tulipalon jälkeen esititte suunnitelmianne ensimmäisen
kerran minulle ja kunnioititte minua äärettömästi pyytämällä minua
auttamaan teitä sen toteuttamisessa. Sen jälkeen, isä, olen niin
katkerasti ja kiihkeästi ikävöinyt raakaa tietämättömyyttä, joka ei
kaipaa muuta kuin ruoskaa ja jolla ei ole muita ihanteita kuin viini,
hyvä mustalaissoitto ja lihavat talonpoikaistytöt. Olen ikävöinyt
äärettömästi noiden kaipausten ja ihanteiden suomaa onnea, jota
voidaan verrata kedoilla laitumella käypien eläinten tyytyväisyyteen.
En halua enää ollenkaan riistää tuota onnea tasankojen kylissä
asuvilta tovereiltani ja vertaisiltani, sillä heidän saamansa korvaus
siitä olisi niin sanomattoman mitätön».
»András», sanoi pappi hyvin ystävällisesti, »olet kyllä kärsinyt

paljon ja kuten kaikki muutkin, joille on sattunut joku kauhea
onnettomuus, katsot sinäkin senvuoksi tulevaisuuteen niin synkästi,
että ammut aivan harhaan kuin sokea. Sivistys antoi sinulle suuria
pyrinnöitä ja niiden siivillä kohosit sinä ihanteitten kuningaskuntaan,
mutta nopeasti tapahtui jotakin juuri silloin kun ojensit kätesi
poimiaksesi nuo jo melkein saatavissa olevat hedelmät, ja sinä
putosit jälleen maahan haavoitettuna ja loukkautuneena. Moitit
pyrkimyksiä, jotka kantoivat sinua ylöspäin, ja soimaat ihanteita, joita
koetit saavuttaa, huomaamatta, että niitä ehkä painoivat inhimilliset
intohimot, jotka vetivät ne takaisin maahan».
»Ei, isä, en ole soimannut ihanteitani, jotka ehkä olivat

saavuttamattomissa, kuten sanoitte. Ollessani nuori poika halusin
nähdä isäni pelloilla kasvavan viljan antavan suuremman sadon kuin
kenenkään muun, mutta sitten myöhemmin haaveilin omasta
kodista, jossa äitini istuisi suuressa nojatuolissaan ja minä hankkisin
hänelle kaikki mukavuudet, joita hän vain kaipaisi. Ehkä uneksin
myöskin kauniista vaimosta, jonka ottaisin sitten hänen kuoltuaan, ja
jonka ruoanlaittotaitoa ja kutomia kankaita kehuttaisiin kaikkialla
maakunnassa. Tällaiset ovat, isä, useimpien kylän poikien ihanteet,
jotka hyvin sopivatkin orjien jälkeläisille. Tuollaiset ihanteet voidaan
helposti saavuttaa, ja vaikka talonpojan mieli eksyisikin liian kauaksi
noihin seutuihin tullakseen siellä nöyryytetyksi, ei se kumminkaan
lennä liian korkealle. Hän putoo matalalta ja putoamisen aiheuttamat
haavat paranevat helposti. Tyytyväisyys ja jonkinlainen tyyni
mietiskely muuttavat vanhuuden hauskaksi. Mutta minä rikkauden
aiheuttamine ylpeyksineni rupesin haaveilemaan muista asioista.
Ojensin karkeat ruskeat käteni tunnustelevasti muita esineitä kuin
lapiota ja viikatetta kohti. Te ohjasitte horjuvat askeleeni oppimisen ja
sivistyksen uusiin seutuihin. Hyvinvoinnin suoma onnellinen tunne
täytti mieleni. Aloin ajatella, että tuo lumoava valtakunta olikin
todellinen kotini, ja otaksuin, että olin vapautunut ikuisiksi ajoiksi
talonpojan suuresta tietämättömyydestä, hänen rivoista
huvituksistaan ja hänen raa'asta tavallisesta luonteestaan. Luulin
voivani sen jälkeen kuljeskella mielin määrin ja häiritsemättömästi
yhä korkeammalle monien kauniiden seutujen kautta lumoavien
ihanteitteni maahan, jonka ihastunut katseeni oli nopeasti
näkevinään sen pilvimaan takana, jossa parhaillaan oleskelin. Sitten
rupesin unelmieni kehoituksesta hakemaan tuota ihannetta tähtien
joukosta, ja minä saavutinkin sen. Pilvet hajautuivat ja näin sellaisen
paratiisimaisen näyn, ettei kukaan kuoleva ihminen ole saanut
sellaista ennen katsella. Mutta ojentaessani käteni tarttuakseni
siihen, ilmestyikin eteeni nopeasti tuo julma ja heltymätön raivotar
'Ennakkoluulo' ja sulki minulta tien paratiisiin. Ivallisin sormin viittasi
tuo hirviö karkeihin käsiini, raskaaseen käyntiini,
lammasnahkaviittaani ja pellavapaitaani, alkaen sitten äänekkäästi ja
ivallisesti nauraen katkoa sydänjänteitäni ja kiskoa niitä rinnastani.
Lopulta heitti hän minut noista pyörryttävistä korkeuksista takasin
maahan ja helvettiin. Ah, isä, tuo putoaminen tuntui hirveältä, sillä
olinhan uskaltanut katsoa tähtiin. En kumminkaan moiti niitä. Ne
eivät voi mitään omalle saavuttamattomalle suloisuudelleen, ja
raivotar, joka suojelee heitä, on sitonut lujan siteen heidän silmilleen.
Soimaan ainoastaan omaa tyhmyyttäni, ylpeyttäni ja intohimojanikin,
jos niin haluatte, mutta en voi olla moittimatta tuota näkymätöntä
kättäkään, joka alussa tempaisi silmieni edestä tietämättömyyden
onnellisuuden verhon ja näytti minulle välähdyksiä tuosta paratiisista,
jota eivät meikäläiset voi milloinkaan saavuttaa. Katsokaa, isä,
putosin niin äärettömän korkealta. Olen kuolettavasti loukkautunut ja
haavoittunut, mutta huolimatta heikkoudestani on minulla kumminkin
voimia kuiskata: 'Älkää ponnistelko, houkat! Tietämätön tyytyväisyys
on vain onnellisuutta!'»
Hänen äänensä murtui nyyhkytyksiksi. Pappi ei vastannut. Hänen

kokemuksensa ihmisluonteesta, sellaisena kuin se tavallisesti
esiintyi hänelle tässä vaatimattomassa kansassa, joka tuli kertomaan
hänelle surujaan ja huoliaan, ei riittänyt ilmaisemaan, miten hänen oli
meneteltävä tämän omituisen ja intohimoisen miehen voimakkaan
surun lieventämiseksi, miehen, jonka luonteeseen sisältyivät
sivistyneen ihmisen hienotunteisuus ja raa'an unkarilaisen
talonpojan ajattelematon itsepäisyys. Hänen ystävällinen luonteensa
suhtautui osaaottavasti tuohon suruun, jonka laadun hän
epämääräisesti arvasi, mutta jota hän ei voinut ymmärtää. Kuin
lohduttaakseen Andrásta nosti hän kumminkin kätensä pelokkaasti
tämän olalle.
He olivat saapuneet teiden risteykseen, ja nuori talonpoika oli

pysähtynyt sanoakseen jäähyväiset. Hän oli puhunut hyvin
kiihkeästi, ja hänen kasvonsa näyttivät kalpeammilta ja
surullisimmilta kuin äsken. Tuntiessaan papin ystävällisen käden
olallaan säpsähti hän kovasti. Ystävällisesti otti hän sen omiinsa ja
puristi sitä lämpimästi.
Isä Ambrosius katsoi kauan ja myötämielisesti noihin tummiin

silmiin, joissa oli niin surullinen ja toivoton ilme.
»Eikö sinustakin tunnu, András», sanoi hän hyvin ystävällisesti,

»että mielesi huojentuisi suuresti, jos kertoisit minulle huolesi»?
Andráksen ote heltisi äkkiä, hän laski vanhan ystävänsä käden irti
ja hänen katseensa synkistyi huomattavasti.
»Ei minulla ole mitään kertomista», sanoi hän kartellen.
Isä Ambrosius huokaisi. Hän näytti pettyneeltä ja

loukkaantuneelta, ja hääräili hevosensa luona aikoen kiivetä sen
selkään.
»Ettekö halua sanoa minulle jäähyväisiäkään, isä»?
Pappi tarttui nuoren talonpojan hänelle ojentamaan käteen

katsoen häneen moittivasti.
»András, et näytä enää välittävän vanhasta ystävästäsi».

»Olette aivan väärässä, isä», sanoi András vakavasti. »Teidän on
pyydettävä piispalta erityinen synninpäästö, sillä niin luonnoton
valhe. Kas niin, nyt taasen — teidän on suotava minulle anteeksi.
Olen sellainen kiittämätön raukka. Isä, saatte minulta tuota
kouluanne varten niin paljon rahaa kuin haluatte. Aloittakaa
rakentaminen heti roudan haihduttua maasta. Teidän on
huolehdittava siitä kaikesta, sillä teidänhän se on suunnitelmannekin.
Toimikaa, kuten parhaaksi näette. Hyvästi nyt, ja sanokaa äidilleni,
että tulen kotiin huomenna».
»Jumala siunatkoon sinua, András! Minä…»
»Hsh, luultavasti täyttää hän pyyntönne myöhemmin. Nykyään ei

hän muista minua. Hyvästi»!
Vanha pappi oli noussut hevosensa selkään, mutta nähtävästi

tuntui lähtö hänestä hyvin vaikealta. Pari kertaa katsahti hän
taakseen Dandárin juostessa majesteettista ravia. Pitkä talonpoika
seisoi kauan teiden risteyksessä katsellen hänen jälkeensä. Isä
Ambrosius voi nähdä hänet hyvin kirkasta taivasta vasten. Hän
kaivoi taskustaan suuren nenäliinansa ja niisti nenäänsä
voimakkaasti, sillä ahdistava pala oli kohonnut hänen kurkkuunsa.
XXIX
»KUNNIOITA ISÄÄSI JA ÄITIÄSI»!
Bideskuty’n seinät kaikuivat jälleen ilosta ja naurusta. Suuret

keittiöt olivat jälleen täynnä ahkeria tyttöjä ja keittiöpoikia, kokonaiset
härät paistuivat suurissa vartaissa, lampaita ja karitsoita oli
teurastettu, ja suurilla portailla ja kivitetyillä käytävillä riennettiin
edestakaisin. Huomenna oli kreivittären syntymäpäivä, ja vihdoinkin
parin vuoden kuluttua iloittaisiin Bideskuty’ssa jälleen ja
kuunneltaisiin mustalaisten soittoa.
Kreivitär Irma oli kiihtyneempi ja hermostuneempi kuin hän

tavallisissa oloissa olisi pitänyt sopivanakaan. Paljon kauheita
tapahtumia oli sattunut sen jälkeen kuin vierasjoukko oli täyttänyt
Bideskuty’n suojat viimeksi ylimyksellisellä iloisuudellaan, josta
kreivitär niin paljon piti. Hän tiesi tuon ajan vihdoinkin koittaneen,
jolloin hänen oli kuunneltava vieraiden kaksimielisiä puheita ja
ivallisen myötätunnon osoituksia, ja vastattava ystäviensä
hämmästyneihin kysymyksiin maakunnan suloisimman perijättären
mahdottomasta avioliitosta. Kun hän määräili keittiössä ja navetassa,
ja valvoi ruokapöydän koristamista, viinin laskemista tynnyreihin ja
vierashuoneiden tuulettamista, voi hän tuskin todeta, että hänen
elämässään oli tapahtunut suuri muutos sen jälkeen kuin pari vuotta
sitten tuon peloittavan ja salaperäisen tulipalon jälkeen kaikki vieraat
olivat poistuneet nopeasti tästä silloin niin surullisesta ja autiosta
paikasta. Ainoastaan, kun hän kohtasi Ilonkan portailla ja kun
aterioiden aikana tuon pari vuotta sitten niin iloisen tytön omituinen
hiljaisuus kiusasi häntä tuoden hänen mieleensä sellaisen
kummallisen vaikutelman, ettei hän enää voinut hallita omaa
tytärtään eikä keksiä hänen häpeämättömän käyttäytymisensä syitä,
silloin vain muisti hän vastenmielisesti nuo monet tapahtumat, jotka
olivat seuranneet nopeasti toisiaan viimeisten vuosien kuluessa.
Osana hänen suunnitelmastaan oli aina ollut, että Ilonka poistuisi

niin pian kuin suinkin sivistymättömän miehensä kodista, eikä hän
tuotta ajatuksetta luultavasti olisi milloinkaan suostunutkaan tuohon
kauheaan avioliittoon. Hän oli kumminkin samalla aina luullut, että
hänen lapsensa palaisi hänen luokseen melkein samanlaisena kuin
hän oli ollut, ennenkuin tuo hirveä talonpoikainen kosija oli
varjostanut hänen elämänsä. Hän oli otaksunut, että Ilonka
suottaleskenä olisi vielä viehättävämpi miesten mielestä, ja
vapaampi ja hilpeämpi käytökseltään, mutta kumminkin melkein
sama kuin ennenkin ja hyvin iloinen saadessaan yhtyä äitinsä
kanssa yhteiseen ponnistukseen — kartoittaa kaikki tuon
talonpoikaisen aviomiehen muistotkin Bideskuty’n rajojen
ulkopuolelle.
Kun Ilonka hääpäivän jälkeisenä päivänä palasi entiseen kotiinsa

aivan entisen minänsä varjona, tunsi äidin sydän ensimmäisen
kerran surua ja katumusta. Täynnä todellista myötätuntoa koetti hän
puristaa järkytetyn lapsensa hellään äidilliseen syliinsä, ja
valmistautui kuuntelemaan kauhistuneena ja itkien syvästi
loukkautuneen ylimyksellisen tyttärensä kertomusta raa'an
talonpoikaisen miehensä kohtelusta.
Mutta sen sijaan olikin Ilonka, joka nähtävästi oli vielä heikko ja
sairas kaikesta kokemastaan, kieltäytynytkin kertomasta sanaakaan
lyhytaikaisesta olostaan miehensä kodissa. Hän oli ottanut vaitiollen
vastaan vanhempiensa suutelot, vaitiollen oli hän istuutunut entiselle
paikalleen pöytään ja tyynesti ruvennut jälleen asumaan entisessä
huoneessaan yön poissaolon jälkeen ja jatkamaan tuskin
keskeytyneitä hommiaan.
Kreivittären mielestä oli Ilonka omituisesti muuttunut. Hän näytti

kokonaan unhottaneen hymyilemisen, hän puhui hyvin vähän ja hän
oli omistanut muutamia hyvin omituisia mielipiteitä lyhyen
oleskelunsa ajalta talonpojan kodissa. Kreivitär oli menettänyt
vaikutusvaltansa häneen, ja kreivitär Irma tunsi itsekin, kuinka
naurettavaa olisi ollut koettaakaan ohjata tytärtä, joka kuitenkin oli
naimisissa oleva nainen ja niin ollen oikeutettu menettelemään,
miten hän vain halusi. Häntä harmitti hirveästi llonkan itsepäisyys
käydä vierailemassa Kisfalussa tuon vanhan talonpoikaisnaisen
luona, jota hän itsepäisesti nimitti »äidiksi». Nuoren suottalesken ja
hänen anoppinsa välille näytti syntyneen omituinen ystävyys.
Kreivitär Irma ei voinut sitä ymmärtää, mutta hän oli kuullut aivan
varmalta taholta, että Keményn András oli asunut Zárdassa
hääpäivästä alkaen. Hän oli senvuoksi johonkin määrin tyydytetty,
ettei hänen tyttärensä kumminkaan alentunut niin paljon, että hän
vieläkin seurusteli tuon hävittömän talonpojan kanssa, joka oli
uskaltanut tuppautua hänen miehekseen.
Mitä tuohon inhoittavaan Keményn Andrákseen tuli, ei kreivitär

vaivannut juuri aivojaan ajattelemalla häntä. Hääpäivän jälkeisenä
päivänä oli hän lähettänyt Bideskuty’yn tärkeän paperin, jossa hän
kirjallisesti lahjoitti Ilonkalle koko Bideskuty’n tilan. Tuohan oli
niinkuin sen olla pitikin, tietysti, sillä ei suinkaan hän voinutkaan
vaatia, että hänen ylimyksellinen vaimonsa olisi ollut riippuvainen
hänestä tahi tytön omista vanhemmista. Kreivitär Irmasta oli se
kaikkein suotuisin ratkaisu, eikä Ilonka sitäpaitsi ollut milloinkaan
tiennytkään, ettei maatila enää kuulunutkaan hänen vanhemmilleen
eikä hänen tarvinnut saada sitä nyt ikinä tietääkään. Ilonka ei
milloinkaan kysellyt mitään eikä näyttänyt huolehtivan ollenkaan,
mistä hän saa ylläpitonsa ja kuka hänet elättää ja vaatettaa. Tyttö
luuli varmaankin, että isä oli vielä tuo sama rikas aatelismies,
jollaiseksi hän oli kuullut isäänsä aina sanottavan, ja otaksui
luultavasti, että hän mainiosti voi elättää ainoan tyttärensä, ellei tämä
halunnut asua miehensä luona. Vanhemmat kohtelivatkin häntä
hyvin ystävällisesti. Hän sai kaikkea, mitä hän vain halusi, ja tuona
kauheana koleeran raivoamisaikanakin antoi Bideskuty hänelle niin
paljon rahaa, vehnää ja viiniä kuin hän vain halusi jakaa kylän
köyhille. Kreivitär Irma kohotti kauhistuneena kätensä, kun Ilonka
ilmaisi tarkoituksensa olevan jäädä Kisfaluun niin kauaksi, kunnes
kulkutauti lakkaa raivoamasta. Ilonka oli sanonut tuon »Menen
sinne»! sellaisin äänensävyin, ettei sitä voitu vastustaa. Ja mitenpä
häntä olisi voitukaan estää, sillä olihan hän naimisissa. Hän ei ollut
enää äitinsä holhouksessa.
Ilonka viipyi poissa neljä kuukautta eikä kreivitär Irma voinut

käsittää, mitä lapsi tuon ajan kuluessa teki. Keményn András oli
Zárdassa, jossa koleera raivosi pahemmin, ja Bideskuty’n jalo
kreivitär toivoi, että tauti täyttäisi velvollisuutensa tuohon vihattavaan
mieheenkin nähden.
Kylässä raivoava koleera ja alituinen tartunnan pelko riittivät isä
Ambrosiuksenkin käyntien lakkaamisen syiksi. Pappi olikin
omaksunut hyvin vastenmielisen tavan puhua aina Andráksesta ja
kaikista tämän kuluttamista rahamääristä kulkutaudin tuottamien
häviöiden korvaamiseksi. Ilonka oli poissa, joten hän ei ollut
kuulemassa tätä alituista ylistystä, mutta hän näytti kumminkin
omaksuneen jotakin sivistymättömän miehensä hellyydestä noita
Arokszállaksen likaisissa taloissa asuvia talonpoika-raukkoja
kohtaan.
Vihdoinkin talvella palasi Ilonka jälleen kotiinsa. Hän näytti

onnellisemmalta ja iloisemmalta kuin lähtiessään, mutta hän ei
milloinkaan halunnut kertoa äidilleen, miten hän oli viettänyt viimeiset
kuukaudet. Kreivitär Irma alkoi kumminkin vähitellen odottaa kevättä
ja kesää. Viime vuonna ei hänen syntymäpäiväänsä oltu ollenkaan
vietetty, sillä silloin raivosi juuri koleera, mutta tänä vuonna tiesi hän
monien tulevan, ja vanhat melkein unhottuneet perinnäistavat
uudistettaisiin. Uteliaisuudestakin tulisi Bideskuty’yn paljon vieraita.
Kreivitär Kantássy olisi kai vielä väsyttävämpi kuin ennen. Mariska
oli juuri mennyt naimisiin Bartoczin Zsigan kanssa, jolla oli hyvä virka
Lontoossa olevassa lähetystössä, Bideskuty’n Ilonkan, tuon
maakunnan kuuluisimman kaunottaren, saadessa tyytyä erääseen
isänsä maalla asuvaan talonpoikaan. Tuo vahinko korvautuisi
kumminkin vähitellen. Bideskuty’n vieraanvaraisuus tänä vuonna,
viinit, paistit ja hedelmät, voittaa varmaankin kaikkien naapurien
suurimmatkin vaatimukset.
Kreivillä oli nykyään aina runsaasti rahaa, ja tämän vuoden sato oli
ollut niin runsas, että se oli täydellisesti korvannut viimevuotisen
tulvan aiheuttamat vahingot.
Keményn András näytti todellakin ymmärtävän, miten maatilaa oli
hoidettava. Ja sehän olikin luonnollista, sillä kaikki talonpojat
ymmärsivät aina, miten vehnää ja turnipseja oli kasvatettava. András
hoiti kaiken eikä kreivin tarvinnut muuta kuin iloita saavutuksista ja
vuosituloistaan, joita hän ei enää jaksanut kokonaan tuhlatakaan.
»Tuntuu aivan siltä, kuin entiset ajat olisivat palanneet jälleen»,

sanoi kreivitär Irma miehelleen, joka istui tupakoimassa
työhuoneessaan. »Muistatko, Gyuri, että me pari vuotta sitten
istuimme juuri näin keskustelemassa syntymäpäiväni vietosta? Kuka
olisi silloin voinut aavistaakaan kaikkia sen jälkeen osallemme
sattuneita onnettomuuksia»?
Bideskuty’n Gyuria vaivasi leini niin, ettei hän voinut laskea

jalkaansa maahan, ja hän oli senvuoksi äreällä ja kärtyisellä tuulella.
Tupakoidessaan murahti hän jotakin vihaisesti vastaukseksi.
»Sinun on nyt myönnettävä, uskallan sen sanoa», lisäsi kreivitär,

»minun olleen aivan oikeassa varoittaessani sinua noista pirullisista
laitoksista. Ja muutkin varoittivat sinua, Gyuri. Huomaat kai nyt,
miten väärässä olit».
Bideskuty ei vastannut mitään, sillä hänen kärsivällisyytensä oli

loppumaton. Sitäpaitsi oli häntä moitittu ja syytetty niin usein, ettei
hän välittänyt sellaisista puheista enää ollenkaan. Hän tupruutteli
piipustaan suuria savupilviä, murahtaen silloin tällöin tuskasta ja
kiroillen aina tämän tästä partaansa. Sitten kuin hänen vaimonsa
lopetti, sanoi hän tyynesti:
»Olen varma, etteivät kaikki ruokalistasi ole vielä valmiit. Ja puoli

tuntia sitten sanoit meneväsi leikkaamaan kukkia pöytäkoristeiksi».
»Haluat siis karkoittaa minut pois. Odotatko ketään»?
»Odotan».
»Ketä sitten»?
»Erästä liiketuttavaani».
»Liiketuttavaako, Gyuri?» kysyi kreivitär epäilevästi.
»Et suinkaan ketään juutalaista koronkiskuria»?
»En, en! Mitä taivaan nimessä se sinua liikuttaa, ketä minä

odotan»?
»En ymmärrä tuota sanaa 'liikeasiat'. Vävysihän hoitaa nuo asiat

tavallisesti. Sanoit, ettet odota ketään juutalaista. Kunhan vain
et…?»
»Odotan juuri häntä! Minulla on luullakseni oikeus käskeä talooni

kenen ikinä vain haluan!»
»Aiotko todellakin sanoa, että olet kutsunut tuon miehen tänne»?
»Ja miksi en olisi»?
»Mutta Ilonkahan on täällä»!
»Ei suinkaan hän syö tyttöä».
»Gyuri, sinun on ajateltava hieman tyttäresi tunteita. Ilonka ei saa

nähdä tuota miestä täällä».
»Joutavia! András on hänen miehensä, eikö olekin? Kuvitteletko
heidän aina elävän erillään, kuten tähän asti? Ja jos sellainen
osoittautuu mahdolliseksi, voin vain sanoa, että Ilonkan on täytynyt
kiusata häntä äärettömästi, kuten sinäkin minua. Koska hänellä ei
ollut leiniä, pääsi hän pakoon».
»Gyuri, en voi uskoa, että tuo sivistymätön talonpoika on lumonnut

sinutkin, kuten hän on lumonnut tuon tyhmän vanhan papin. Puhut
kiusaamisesta. Elämäni on yhtä vaikeaa kuin marttyyrien kuullessani
kaikkien ylistävän häntä. Sanon häntä kauheaksi raakalaiseksi
hänen käytöksensä vuoksi tytärtäsi kohtaan. Ilonka on liian hellä ja
itseensä sulkeutunut kertoakseen sinulle kaikkia, mitä hän miehensä
puolelta on saanut kestää, mutta varmasti ei hän olisi poistunut
miehensä luota niin pian, ellei mies olisi käyttäytynyt vielä
sivistymättömämmin kuin olin hänestä luullutkaan».
»Teillä naisilla ei ole ollenkaan kunniantuntoa», jyrisi Bideskuty.

»Puhut tuosta miehestä kuin jostakin raakalaisesta ja pedosta, ja
kumminkin olet valmis ottamaan häneltä vastaan kaikki hänen
jalomieliset lahjansa. Hän pelasti koko tämän talon, jossa elämme,
tuon verenimijän, Rosensteinin, kynsistä, ja lahjoitti Ilonkalle koko
maatilan, joka kokonaisuudessaan on kai maksanut hänelle
satojatuhansia. Tyynesti toimittaa hän kaikki puolestamme, että
Ilonka voisi elää niin ylellisesti, ettei hänen eikä meidän ylpeytemme
joutuisi kärsimään. Ja sinä puhut hänen häpeällisestä
käyttäytymisestään tytärtämme kohtaan. Mitä sinä sitten sanot
Ilonkan käytöksestä»?
»Gyuri, sinun on muistettava, ettei lapsi raukka ollenkaan

tiedäkään kotimme luovuttamisesta hänelle. Hän ei tiedä mitään
sinun ja tuon miehen välisistä raha-asioista».
»Minun syytäni ei ole ollenkaan noiden inhoittavien salaisuuksien
keksiminen. Ilonka oli kylliksi vanha saadakseen tietää kaiken.
Pelimme miestä kohtaan ei ollut rehellinen. Koetit parhaasi mukaan
häväistä häntä lapsesi kuullen luovuttaen Ilonkan kumminkin hänelle
ja sallien hänen yksinään taistella taistelunsa Ilonkan voittamiseksi.
Sellainen ei ollut rehellistä peliä».
»Gyuri, et tiedä enää, mitä puhut. Jos olisin kertonut Ilonkalle

sinun ja Keményn välisistä asioista, olisit joutunut hyvin omituiseen
valoon lapsesi silmissä. Olisitko voinut sitten vaatia Ilonkaa
kunnioittamaan isäänsä ja äitiään? Silloin olisi hän menettänyt
kaiken myötämielisyytensä meitä kohtaan».
»Mielestäni olisi se paljon suotavampaa, että hän kunnioittaisi

meitä hieman vähemmän ja miestään enemmän. Murrun melkein
tuon kiitollisuuden taakan alle, jonka Ilonka ja kaikki me muutkin
olemme velkaa tuolle miehelle».
»Gyuri, puheestasi päättäen olet omaksunut lukiessasi noita

ulkomaalaisia kirjoja yhä enemmän noita uusia aatteita, jotka jo
kerran veivät sinut vararikkoon».
»Älä koske niihin kieliin, Irma, tahi…»
»Hsh, malttisi menettäminen ei hyödytä mitään, Gyuri, sillä tehtyä

ei saada enää tekemättömäksi. Meidän on koetettava tehdä Ilonka
niin onnelliseksi kuin suinkin, että hän voi unhottaa menneisyyden.
Hän on vielä niin nuori ja András on luullakseni jo miehuutensa
parhaassa iässä. On hyvin mahdollista, että Ilonkasta tulee leski,
ennenkuin hän ennättää täyttää kolmeakymmentäkään. Silloin tulee
hänestä niin rikas, että hän voi mennä naimisiin kenen kanssa hän
vain haluaa, ja varmasti on hän ensimmäinen kiittämään meitä
tavasta, jolla olemme suunnitelleet hänen elämänsä. Sillä aikaa on
minun koetettava pitää häntä poissa talon tästä osasta. Tänään…»
Ovi aukeni hiljaa ja Ilonka tuli huoneeseen hymyillen niin iloisesti,

että hän oli melkein entisen näköinen. Kreivitär katsahti häneen
epäluuloisesti, mutta tyttö näytti aivan tietämättömältä ja
iloisemmalta kuin moniin kuukausiin.
»No niin, minun on mentävä katsomaan kukkakoristuksiani», sanoi

kreivitär välinpitämättömästi. »Mielestäni, voisit sinä, Ilonka, tulla
minua auttamaan. Saat koota minulle suuren korillisen noita kauniita
kukkia lavojen takaa. Palvelijattaret eivät osaa niitä katkoa, vaan
repivät kukat juurineen maasta. Tarvitsen niitä melko paljon. Saat
Pannalta korin, ja kun olet täyttänyt sen, tule luokseni
leipomatupaan, jossa sitten järjestän ne».
»Tulen aivan heti, äiti. Mutta», lisäsi hän viehättävästi hymyillen,

»enkö saa puhella hieman isäni kanssa ensin»?
»Ainoastaan muutamia silmänräpäyksiä vain. Tarvitsen noita

kukkia, ja isäsi odottaa erästä liiketuttavaansa luokseen».
»Tulen luoksesi, äiti, viiden minuutin kuluttua».
Poistuminen tuntui kreivitär Irmasta hyvin vastenmieliseltä. Hän ei

halunnut milloinkaan jättää tytärtään isän huostaan. Katsahdettuaan
varoittavasti mieheensä poistui hän kumminkin.
Ilonka odotti, kunnes äidin askelten ääni lakkasi kuulumasta

käytävästä. Sitten hän kääntyi isäänsä päin ja sanoi tyynesti:
»Isä, etkö halua kertoa minulle, mikä tuo 'kiitollisuuden taakka' on,
jonka minä ja me muut kaikki olemme velkaa miehelleni»?
»Ilonka, olet kuunnellut»!
»Tahtomattani, vannon sen. Olin juuri tulossa huoneeseen, kun

tuo lause kantautui korviini aikoessani avata oven. Tunnustan, että
koetin kuulla lisääkin, mutta äidin puheesta en saanut mitään selvää.
Kai kerrot minulle, etkö kerrokin»?
»Puhuin vain ylimalkaisesti», sanoi Bideskuty hermostuneesti. »Et

varmaankaan kuullut oikein»?
»Nyt isä», sanoi Ilonka taivuttavasti, »toivon sinun koettavan

muistaa, etten ole enää aivan sellainen lapsi kuin olen ollut. Kaksi
vuotta on jo pitkä aika», lisäsi hän miettivästi, »ja niiden kuluessa
olen kokenut kaikenlaista. Olen nyt naimisissa ja olen paljon
vanhempikin. Luullakseni on minulla oikeus tietää, miksi olemme
kiitollisuuden velassa tuolle miehelle, jonka nimi minulla nyt on».
»Sinun on taivutettava äitisi kertomaan sinulle kaikki nuo toivomasi

asiat», sanoi Bideskuty.
»Tiedät aivan hyvin, isä, ettei hän kerro minulle mitään. On aivan
hyödytöntä taistella sitä vastaan, rakkaani, sillä en aio poistua tästä
huoneesta, ennenkuin saan tietää, mitä tahdon».
»Siinä ei ole mitään kerrottavaa».
»Millaisessa kiitollisuudenvelassa olen minä miehelleni»?
»Ymmärsit sanani aivan väärin», väitti Bideskuty itsepäisesti.
»Isä, olen kysynyt sinulta niin kunnioittavasti kuin lapsenasi vain

voin. Älä pakota minua vaatimaan, mitä minulla on oikeus saada
tietää».
»Ilonka, olet luonnoton. Mitä sinua hyödyttää kuulla asioita, jotka
koskevat vain minua ja äitiäsi?»
»Kuinka paljon rahaa antoi Keményn András sinulle saadakseen

mennä naimisiin kanssani»?
»Ilonka, olet menettänyt järkesi», sanoi Bideskuty vihaisesti.
»Enkä ole! Et halua kertoa minulle totuutta ja senvuoksi on minun

pakko tehdä johtopäätöksiä. Jos kieltäydyt kertomasta minulle
kaikkea, on minun mahdotonta oleskella kattosi alla enää tuntiakaan,
ja…» lisäsi hän änkyttäen, »koska ei tietysti mieheni huoli minua
luokseen, on minun mentävä johonkin muualle».
»Kuuntele nyt, Ilonka. Teissä naisissa ei ole hituistakaan

johdonmukaisuutta. Et ole mielestäsi enää mikään lapsi, ja niin ollen
kai ymmärrät, että tulipalo toisena ja tulva toisena vuonna voivat
köyhdyttää rikkaankin maanviljelijän. Sen lisäksi uhkasi ja petti
minua muudan verenimijä, koronkiskuri, kunnes kaikki maani
joutuivat vieraisiin käsiin. Miehesi on lainannut minulle paljon rahaa
maitani vastaan ottamalla niistä vain kohtuullisen koron. Tuo
Rosenstein roisto, jonka kuulemma piru vihdoinkin on korjannut,
pakotti minut allekirjoittamaan muutamia papereita, joiden
perusteella hän sitten pakotti minut maksamaan suunnattomia
korkoja. Maksoin niitä vuosikausia tietämättä ollenkaan, että ne
menivät juutalaisen taskuihin Keményn saamatta niistä penniäkään.
Tuli ja vesi täydensivät tuon koronkiskurin työn. Minusta tuli köyhä
mies. Silloin ilmestyivät jälleen nuo allekirjoittamani paperit näkyviin,
joita en ollut lukenut läpikään, ja sain tietää niiden olevan antamiani
velkakirjoja rahoista, joita en milloinkaan ollut saanutkaan.
Rosenstein uhkasi minua kaikin mahdollisin keinoin, joita en nyt
muistakaan. Hänellä tuntui olevan oikeus puolellaan, koska olin
allekirjoittanut nuo paperit. Keményn András tuli silloin luokseni. Hän
tukki juutalaisen suun kullalla, lunasti nuo paperit minulle takaisin ja
maksoi tämän rakennuksen kiinnityksen, josta Rosenstein uhkasi
karkoittaa meidät pois. Maa oli kumminkin jo hänen omaisuuttaan.
Sinä, minä ja äitisi olisimme muuttuneet samanlaisiksi kerjäläisiksi
kuin kaikki kodittomat mustalaiset. András kertoi rakastavansa sinua
ja haluavansa mennä kanssasi naimisiin. Hän sanoi maiden siten
joutuvan sinun ja lastesi omaisuudeksi. Mitä voinkaan tehdä?
Puukko uhkasi kurkkuani — ja minä suostuin».
Ilonka ei sanonut mitään. Hän tuijotti isäänsä kalpein kasvoin ja

toivottomin, hämmästynein ilmein.
»Hän maksoi tämän talon kiinnityksen, otti maat haltuunsa ja

rupesi hoitamaan niitä niinkuin hän ainoastaan kykenee hoitamaan
maatiloja. Hän omisti kaiken, mutta ei kukaan tiennyt sitä. Hän
neuvotteli kanssani kaikesta ja toimi puolestani kuin jonkunlainen
välittäjä. Joskus unohdankin kokonaan, etten ole näiden maiden
omistaja, ja annan hänelle määräyksiä, jotka hän aina toimittaa
perusteellisesti. Hän kertoi minulle kerran olevansa vain sinun
tilanhoitajasi. Tuolla miehellä on enemmän sydäntä», lisäsi
Bideskuty lyöden nyrkkinsä kovasti pöytään, »kuin kenelläkään
muulla tuntemallani henkilöllä ja…»
»Rakas isä», keskeytti Ilonka, »kerro minulle ainoastaan

tosiseikat.
Älä muserra minua häpeällä enemmän kuin on tarpeellista».
»Et ole milloinkaan kertonut äidillesi, miksi poistuit miehesi luota

eikä Andráskaan ole sanonut minulle mitään. Hääpäivänne
jälkeisenä päivänä lähetti hän minulle erään paperin, jonka luin läpi
hyvin huolellisesti. Tuossa asiakirjassa lahjoittaa hän sinulle koko
Bideskuty’n tilan, pidättäen itselleen vain oikeuden valvoa sen
hoitamista. Katsohan, hän ei oikein luota minuun», lisäsi Bideskuty
hymyillen, »enkä ole hänen mielestään mikään kelvollinen
tilanhoitaja. Mutta hän itse on suurenmoinen, Ilonka!» huudahti hän
innostuneesti. »Näet itse, miten ruhtinaallisessa kunnossa tämä talo
nyt on. Minulla on nyt aina runsaasti rahaa, ja käypää metallirahaa
päällepäätteeksi, paljon viljaa myytäväksi, ja karjani lihoo ja lisäytyy,
kuten kaikki muukin. Minulla ei ole ollut milloinkaan niin paljon
nautoja eikä vasikoita, eikä niin suuria määriä vehnää eikä maissia
myytävänä. Tuo mies tuntee jokaisen maajalan arvon. Hän huolehtii
kaikesta. Minä oleskelen vain täällä kotona, hyväksyn hänen
toimenpiteensä ja korjaan rahat, kun hän on tehnyt hyvän kaupan
puolestani… tarkoitan sinun puolestasi, Ilonka, sillä sinunhan tämä
kaikki on. Oletko milloinkaan halunnut jotakin, jota et ole saanut,
sano oletko»?
»Sitten eivät nuo rahat, joita jaoin köyhille koleeran

raivoamisaikana, tulleetkaan sinulta, vaan häneltä»? kysyi Ilonka
tyynesti.
»Ei, ei oikeastaan häneltäkään, lapseni, sillä omaisuushan kuuluu

sinulle».
»Hänkö on lahjoittanut sen minulle»?
»Niin, hänhän on miehesi».
»Niin», sanoi Ilonka kiihkeästi, kyynelten värisyttämin äänin, »hän

on mieheni. Hän maksoi suunnattomasti huvista saada nimittää
Bideskuty’n kreivin pennitöntä tytärtä vaimokseen. Voi, miten
häpeällistä tämä kaikki onkaan»! lisäsi hän vihaisesti. »Kuinka te
voittekaan, ah, kuinka te voittekaan»?
»En ymmärrä, miksi puhut häpeästä. Lukuunottamatta sitä, että
olet nähtävästi riidellyt miehesi kanssa, ei siinä ole mitään
häpeällistä. Äitisi ja minä olemme riidelleet aikoinamme paljonkin,
mutta hän ei ollut kumminkaan niin itsepäinen, että hän olisi juossut
tiehensä. Tuollaiset riidat unhottuvat kyllä pian».
»Unhottuvat pian! Ah, isä, sinä et tiedä etkä ymmärrä».
Ilonka nyyhkytti nyt kiihkeästi, ja kätki kasvonsa käsiinsä

toistaessaan:
»Ah, millainen ääretön häpeä! Kuinka te voittekaan»?
»En näe tässä minkään toivottomuuden syytä», sanoi Bideskuty

hieman hermostuneesti. »En ymmärrä, mikä teitä naisia oikeastaan
vaivaa, kun te aina kiusaatte muita teoillanne. Halusit tietää ja pakotit
minut kertomaan vastahakoisesti asioista, joita äitisi mielestä sinun
ei ollenkaan olisi pitänyt saada tietää. Sanon vieläkin kerran, etten
huomaa tässä minkäänlaista itkun syytä».
»Ei olekaan, isä», sanoi Ilonka kuivaten nopeasti silmänsä ja tullen

aivan isänsä viereen. »Kuten sanoit, halusin tietää ja nyt olet
kertonut minulle. Olen sinulle hyvin kiitollinen».
»Et suinkaan aio kertoa äidille»? kysyi kreivi levottomasti.
»En», vastasi Ilonka hymyillen kyyneliensä läpi isänsä levottomalle

kasvojen ilmeelle. »En aio puhua asiasta hänelle sanaakaan, sillä
mitäpä se hyödyttäisi. Menen hänen luokseen nyt katkomaan noita
kukkia».
Sanottuaan sen kumartui hän suutelemaan isäänsä.

Contract Options For Buyers and Sellers of Talent in Professional Sports 1St Ed Edition Duane W Rockerbie Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Contract Options For Buyers and Sellers of Talent in Professional Sports 1St Ed Edition Duane W Rockerbie Full Chapter

Uploaded by

Copyright:

Available Formats

Contract Options for Buyers and Sellers

of Talent in Professional Sports 1st ed.

More information about this series at

ISSN 2662-6438 ISSN 2662-6446 (electronic)

Sports economics is a relatively new field of research that is gaining popu-

to them. Instead, sports economists vaguely refer to “bargaining power”

Lethbridge, Canada Duane W Rockerbie

1 On the Rise: Player Compensation and Multi-year

2 The Puzzle of Overpaid and Underpaid Players 23

3 Contract Options for Buyers and Sellers of Talent 53

4 Extensions to the Put Option Model 69

Table 2.1 Free-agent fielding players by playing position,

On the Rise: Player Compensation

Abstract A salary anomaly has been identified in the sports economics

Keywords Real options · Marginal product · Contract lengths · Baseball

© The Author(s) 2020 1

Revenue Payroll Payroll/Revenue

1 https://www.baseball-reference.com/leagues/MLB/misc.shtml taken March 27,

payroll increased from $17.3 million in 1990 (39% of total expenses) to

3 Source: https://sites.google.com/site/rodswebpages/codestaken on December 20,

Index increased by 85.6% between 1990 and 2016), MLB operates on a

6 Source: http://www.inflationdata.com/inflation/Consumer_Price_Index/ taken on

We attempt to clarify the process of bargaining by introducing the

1.2 A Brief History of Multi-year

in the eastern United States. Amateur leagues quickly dissolved as a conse-

Dodgers and accepted one-year contracts of $125,000 and $110,000,

1.3 Early Years of Free Agency

1.4 Free Agency Since 2000

1.5 Multi-year Contracts

1.6 Player Salaries and Marginal

M AX Ri = R(wi (ti ), Ai ) − Z i (ti )ti − FCi (1.1)

Fig. 1.2 Monopsony model of talent acquisition (Source Author’s creation)

19 In addition to the independent variables used in Scully (1974), we include dummy

Each player’s contribution to the team winning percentage is

1.7 Real Options

Rockerbie, D. (2009b). Strategic free agency in baseball. Journal of Sports

The Puzzle of Overpaid and Underpaid Players

Abstract To use real options theory to explain the salary anomaly in

Keywords Marginal revenue product · Surplus · Logistic · Monopsony

© The Author(s) 2020 23

on the subject. More recent contributions include Scully (1989), Brug-

of estimates has been provided by Financial World and Forbes maga-

Table 2.1 Free-agent fielding players by playing position, 2000–2012

Season Catcher 1B 2B 3B SS OF Total

Source Author’s research

2.2 A Revenue Model

R = attendance revenue + television revenue + all other revenues

Total revenue includes revenue that is specific to each club (attendance,

ln[W P/(1 − W P)] =δ + θ O P S + μ(K /B B) + π E R A+

OPS = team slugging percentage (total bases/total at bats) +

Table 2.2 Estimates of the (natural logarithm of the) revenue function

R = MLB club revenue from all sources Taken from http://www.rodneyfort.com

Source For data in regression models

as an independent variable in (2.2). It was omitted here as it lacked statis-

5 A useful reference is Baumer and Zimbalist (2013).

Totally differentiating (2.1) gives

The contribution to the team OPS by a specific hitter is calculated

linear summation of the individual player performances and ignores other

2.3 Estimates of the Revenue

Table 2.3 Estimates of

Note t-statistics appear in parentheses

»Nyt olen kumminkin toista mieltä», sanoi András hieman

»András», sanoi pappi hyvin ystävällisesti, »olet kyllä kärsinyt