Labor Markets in Professional Sports

Labor Markets in Professional Sports An Overview Smoothly functioning markets  Human Capital  Monopsony and Free Agency  Salary Arbitration  Superstars and Winner-take-all  When to turn pro?  Average Salaries in Pro Sports MLB 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 142,663 154,066 156,480 170,976 289,365 329,557 472,734 598,895 710,995 700,557 863,617 1,457,703 1,489,102 1,348,403 1,621,099 2,079,469 2,410,543 2,486,609 NFL 199,611 203,361 214,572 258,949 289,724 268,220 307,330 507,248 496,380 490,214 621,477 741,867 859,100 971,589 1,158,896 1,224,340 1,365,036 1,333,333 NHL 121,714 203,361 249,057 285,508 266,546 247,587 234,902 214,535 248,190 274,648 304,957 495,475 716,341 1,073,925 1,353,258 1,801,889 1,879,550 1,830,000 NBA ----402,716 389,721 414,993 499,976 646,328 814,362 1,083,971 1,481,041 2,166,869 2,407,903 2,661,983 4,607,317 4,725,125 4,900,000 Source: Table 8.1, Leeds and von Allmen (3e). Adjusted for inflation to 2004 dollars. Average Salaries in Pro Sports $5,500,000 $5,000,000 $4,500,000 $4,000,000 $3,500,000 $3,000,000 NBA MLB $2,500,000 $2,000,000 $1,500,000 $1,000,000 $500,000 $0 NHL NFL 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 What would Babe Ruth earn today?  Ruth earned $80,000 for the 1930 New York Yankees   1930 CPI = 16.7 2006 CPI = 201.6  Ruth’s 1930 salary in 2006 dollars is: (80,000)(201.6/16.7) = $965,749 Labor Market: Competitive Model $ S1 w1 D1 L1 Labor Labor Supply  Income-leisure tradeoff Substitution Effect (work effort rises)  Wage increase causes: $ S IE > SE w* Income Effect (work effort falls) Backward-bending labor supply curve SE > IE Labor Labor Demand   Profit-max decision by employers Hiring Rule: hire until MRP = w $ MRP = MP * MR MP = ∆Q/ ∆L w1 w2 MRP L1 L2 Labor Reflects DMR MR = ∆TR/ ∆Q = P Applications    Decrease in television revenues due to fan preference for drama shows Increase in the number of players available A minimum salary set above the equilibrium wage $ S1 w1 D1 L1 Labor Human Capital Theory Human Capital Productivity Earnings  General Training  Increases MP to all employers Increases MP to specific firm  Specific Training  Who Pays for Training? $ MRP1 = untrained worker MRP1 – T = trainee’s net productivity MRP2 = trained worker Benefit MRP2 MRP1 Cost MRP1 - T Training period Hiring Rule: MRP = w t1 time MRP2 w2 MRP1   w1  T 1 r 1 r GT: worker pays in form of lower training wage ST: worker and firm share costs Labor Market Restrictions Monopsony  Reserve Clause  Arbitration  Salary Caps  Player Draft  Monoposony  Sole buyer of labor  Enables employer to exert market power by paying lower wages $ MC S MRPm  Monopsonist hires until MRP = MC and sets wage off S curve  Lm < Lc  wm < wc < MRPm wc wm D = MRP Lm Lc Labor Reserve Clause  National Association of Professional Base Ball Players (1871-1875)  Players prohibited from signing with another team during season Gentlemen’s agreement: clubs protected 5 players 1889: clubs "reserve" the right to renew the contract for one additional year of service on the same terms… all players 3 year contracts, no reserve clause Profit-sharing between team and players  National League of Professional Baseball Clubs (1876-present)    William Hulbert, Chicago franchise Players’ League (1890)    Federal League (1914-1915) Free Agency  MLB: 1976  After 6 years of service After 5 years of service After 4 years of service After 4 years of service Restrictions: • Right of First Refusal • Compensation requirements • Salary caps  NBA: 1983   NHL: 1993   NFL: 1994  Coase Theorem Revisited [Rottenberg’s Invariance Principle]  Reserve Clause vs. Free Agency Initial assignment of property rights does not affect efficiency of resource allocation  Only difference is who gets larger share of pie  Final Offer Arbitration   MLB 1972 Arbitrator must select either team’s or player’s final offer—No compromise!   must base decisions on info regarding player performance and salaries of comparable players can not consider financial condition of team Wage WT WA WP  Overpaying a player leads to further overpaying down the road Mean Salaries of FOA-eligible Players, 1986-1991 FOA Arbitration Awards Year 1986-1991 Negotiated Settlements $689 (576) $1,061 (121) $763 (111) $698 (97) $564 (76) $497 (69) $384 (102) Total $759 (132) $1,341 (17) $904 (24) $765 (12) $672 (18) $670 (26) $485 (35) Player Wins $867 (59) $1,634 (6) $900 (14) $929 (7) $622 (7) $933 (10) $572 (15) Team Wins $671 (73) $1,181 (11) $909 (10) $535 (5) $699 (11) $507 (16) $420 (20) Total Sample $703 (708) $1,095 (138) $788 (135) $705 (109) $585 (94) $544 (95) $410 (137) 1991 1990 1989 1988 1987 1986 Source: Burgess and Marburger (1993), Table 1, p553. Other Restrictions  Salary Cap  NBA • Larry Bird exception “soft cap” • Hard caps in 1999  NFL • Rozelle Rule • Hard caps in 1994  Player Draft  Allocation of new players by reverse order finish • NBA: 7  2 rounds • NFL: 12  7 rounds  Coase Theorem applies Estimating Player’s MRP  Scully (1974): two-stage model using 1968-69 data PCTWIN = f(PRODUCTIVITY) REV = g(PCTWIN) Scully’s Results TSA = Team Slugging Average TSW = Team Strikeout – Walk Ratio NL = National League CONT = Contender OUT = Out of contention SMSA = Market Population MARGA = Differences in Fan Interest STD = Stadium Age BBPCT = % Black Players PCTWIN = 37.24 + 0.92 TSA + 0.90 TSW – 38.57 NL + 43.78 CONT – 75.64 OUT REV = -1,735,890 + 10,330 PCTWIN + 494,585 SMSA + 512 MARGA + 580,913 NL - 762,248 STD – 58,523 BBPCT 1 point increase in TSA raises PCTWIN by 0.92 1 point increase in PCTWIN raises REV by $10,330 MRP per point = MP x MR = (0.92)(10,300) = $9,504 MRP = ($9,504)(340)(1/12) = $270,000 1970 Average salary was $30,000 Avg Hitter SA: .340 1/12 of team’s offense Results   Scully (1974): Players paid 10-20% of MRP Krautman (1999)    Apprentice: 27% of MRP Journeyman: 85% of MRP Free agents Evidence of monopsonistic exploitation? Alternative Explanation: Low salaries of younger players may reflect general training Example: The Mark McGwire Show    During McGwire’s record-breaking run at the home run record in 1998, attendance in St. Louis increased by 1.5 million. Even if McGwire was only half of the reason, just the gate portion of his MRP that year was around $15 million! McGwire earned $8.9 million that year. Economics of Superstars  Rank order tournaments: golf, tennis, auto racing   relative productivity matters rather than absolute difficult to measure absolute effort (MRP) when many factors are involved $ MC MR1 Increasing MC of effort requires large difference between first and second place for optimal effort. MC′ MR1′ MR2 E1 Effort Economics of Superstars   Forbes Top 50 Celebrities and CEOs Downside to Tournaments?   Tonya Harding NASCAR You are offered a banned performanceenhancing substance that comes with two guarantees: 1. You will not be caught. 2. You will win every competition you enter for the next five years, and then you will die from the side effects of the substances. Would you take it? When to Turn Pro?   The majority of student athletes want to play professionally, but only select few ever will. Some basic statistics:    1 million high-school football players - roughly 150 will make it to the NFL (odds: 1 in 7,000) About 500,000 high-school basketball players roughly 50 to the NBA (odds: 1 in 10,000) Less than 3% of all college seniors will play one year in professional basketball When to Turn Pro?   Why would a player choose to leave early? Consider the MB and MC of staying in school: Marginal Benefit of waiting the extra year is: MB = (1 + g)S0 [where S0 is the pro salary and g is the growth rate in the salary] Marginal Cost of waiting is the foregone salary plus the sacrifice on the use of that salary: MC = (1 + r)S0 [where r is the interest rate] When to Turn Pro?  As usual, the player is best off when MB = MC (1 + g)S0 = (1 + r)S0   Player should stay in school as long as g > r Player should turn pro when r > g Sample Problem Suppose a junior could earn a salary of $500,000 by declaring himself eligible for the draft. If he waits until his senior year he can make $600,000. If the interest rate is 8% should he stay the extra year? g = (600,000 – 500,000)/500,000 = 0.20 or 20% Assume the pro league plans to institute a rookie salary cap of $500,000 at the end of the player’s senior year. Should the player play his senior year? g = (500,000 – 500,000)/500,000 = 0.00 or 0% Now consider that the player has a 12% chance of having a career ending injury in his senior year and thus having a median income of 35,000 per year. Would he consider going pro or not? g = (532,000 – 500,000)/500,000 = 0.064 or 6.4% Labor Unions and Labor Relations The Decline of Labor Unions Unionization Rates: 1973-2005 0.45 Public Sector 0.40 0.35 Percent Unionized 0.30 0.25 0.20 0.15 Private Sector 0.10 0.05 0.00 1971 1976 1981 1986 Year 1991 1996 2001 2006 The Decline of Labor Unions  Structural Hypothesis Global competition  Growth of service sector  Changing demographics  Managerial Opposition  Government Substitution  Economics of Labor Unions $  SU unemployment Free Market: wN, LN wU SN  Union Outcome: wU, LU  Unemployment  Inefficiency (DWL) wN DWL D LU LN Labor Bilateral Monopoly  Union behaves as monopolist:    $ MC S Sets employment where MR = S Sets wage off D curve WU , L U wU wM  Employer behaves as monopsonist:    Sets employment where D = MC Sets wage off S curve WM, LM WU – WM = Range of Indeterminacy MR LU LM D Labor Bargaining and Strikes   Each of major sports had a work stoppage during 1990s (when overall labor strife was pretty tame) Why resort to a strike/lockout?     Irrational behavior? Excessive optimism? Excessive uncertainty? Political gamesmanship? Contract Zone Strike fund Alternative jobs Union threat point = WU Low Wages Acceptable to Employer Acceptable to Union Contract High Wages WE = employer threat point Strike insurance Replacement workers Zone 1972 Baseball Strike   Main issue was player pension and health benefits (or power play?) 1912 Detroit Tigers Ty Uncertainty Cobb vs Ban Johnson • Owners were over-optimistic (believed players' threat point was lower than it was) • MLBPA was optimistic due to Commissioner’s behavior  Strike lasted 13 days (including 9 days at the start of the season) • Owners lost $5m in revenues • Players demonstrated solidarity 1994 Baseball Strike     8th work stoppage since 1972 Main issues: salary cap, free agency, arbitration, small-market/large-market Uncertainty: no MLB commissioner 232 day strike: World Series was canceled   Owners lost $1b Players lost $230m   NLRB declared ULP by owners Old contract terms were re-instated 2002 contract settled without a strike! 1987 NFL Strike  Main issue: free agency (Rozelle Rule) • 1977-1988: 125 free agents per year • Only 2 players changed teams!   Uncertainty: Gene Upshaw and demise of USFL Strike lasted 4 weeks • Replacement players cost $1000 per game; teams profits rose by more than $100k per game • Players lost $80m in salary  NFLPA disbanded in 1988 to sue on antitrust grounds 1998-99 NBA Lockout     Main issue: hard salary cap; revenue sharing Uncertainties: lackluster attendance; turmoil within NBPA; rising power of agents (stars vs benchers) 191 day lockout  50 game season Outcome: Individual player salary cap; players guaranteed 55% of league revenues; limit on raises for “Larry Bird” free agents NBA would be paid TV contract money even though games weren’t played. Arbitrator ruled NBA did not have to pay Players with guaranteed contracts NHL: 2004-05 Lockout     Whole season canceled Main issues: cost certainty (linking salaries to league revenues) Uncertainty: league losses Outcome: $39m salary cap; salaries at no less than 54% league revenues; maximum player salary at 20% of cap; salaries rolled back by 24% Bob Goodenow Gary Bettmen Revenue sharing; luxury tax; 5% pay cut Revenue sharing; luxury tax; 24% pay cut $52m salary cap linked to league revenues $49m salary cap linked to league revenues $40m salary cap linked to league revenues $42.5m salary cap linked to league revenues Discrimination Theory and Measurement First Black Pro Athletes . . . in the modern era Jackie Robinson 1947 Brooklyn Dodgers Larry Doby 1947 Cleveland Indians Kenny Washington 1946 Los Angeles Rams Earl Lloyd 1950 Washington Nationals Willie O’Ree 1958 Boston Bruins Becker’s Taste for Discrimination Model  Becker (1957) “rational choice” model   People are WTP a price to discriminate Source of prejudice: • Employers • Employees • Customers Gary Becker Employer Discrimination  Hiring Rule: w = MRP  Workers with same MRP will be paid same wage  Assume:   MRPB = MRPW d = discrimination coefficient “Psychic cost”  Perceived wage of black player: w*B = (1+d)wW Example: wW = $20 d = 0.20 w*B = (1.20)(20) = $24 In a picture… $ Black wage as perceived by discriminating firm w*B = $24 wW = $20 wB = $16.67 MRPB = MRPW Black wage if firm hires same number of black works as white workers LB Conclusion: > Owners must pay for the right to discriminate in the form of lower profits. > Competitive markets force discriminators out of the market. LW Players Employment if blacks are paid same wage as whites: w = $20 Monopoly Power  Baseball has legal cartel  Policed by commissioner  Bill Veeck foiled in 1943  Jackie Robinson reintegrated baseball Integrated teams tended to dominate  Dodgers, Giants, Indians, & Braves • Red Sox & Phillies last to integrate   Great Celtic teams built on integration Moses ―Fleetwood‖ Walker 1884 A.A. Bud Fowler 1878 Employee Discrimination   Early whites didn’t want to work with blacks  Feel psychic cost  Demand higher pay to work with blacks What would employer do?  Segregation vs Discrimination  Dodgers protested Robinson’s presence Frank Grant 1889 Inventor of Shin Guards Customer Discrimination   Employers punished for tolerance? George Preston Marshall & NFL’s Redskins  Last NFL team to integrate: 1962 • “Burgundy, Gold, and Caucasian”   "We'll start signing Negroes when the Harlem Globetrotters start signing whites." Southern focus Forced by U.S. government • Facility on government land  Nardinelli and Simon (1990)  Examined baseball card prices for black and white players  PB < PW by about 10% Statistical Discrimination  The use of group averages to judge individual productivity levels  Profit-maximizing strategy to reduce cost of hiring English-speaking Canadian players French-speaking Canadian players F E MRPE MRPF productivity Measuring Discrimination $ $500 = WW White Wage Function Black Wage Function $260 = W*B $200 = WB SAB SAW Slugging Average WW – WB = observed wage gap W*B – WB = explained wage gap = 500 – 200 = 300 = 260 – 200 = 60 WW – W*B = unexplained wage gap = 500 – 260 = 240 Economic Findings on Pay Discrimination    There is evidence that pay discrimination existed in pro team sports in the past. But by the mid-1990s, pay discrimination is pretty much gone. Only a negligible premium for the very best white players in the NBA appears to remain. Interestingly, in the NHL, there appears to be pay discrimination against French-speaking players. Puzzling Numbers From Track  Blacks ~ 12% of world’s population  95% of best 100m times are due to runners of West African ancestry • All 40 finalists in last 5 Olympics   North Africa dominates middle distance East Africa dominates distance Astonishing Kenya     More than half of top 100 5K & 10K times Top 60 times in steeplechase Won every World Cross Country Championship since 1986 38 Olympic track medals since 1964  13 golds in men’s races  Only U.S. (with 10X as many people) has more More on Kenya   75% of Kenya’s runners from the Kalenjin tribe in the highlands near Lake Victoria 50% of these from Nandi district  1.8% of Kenya’s population  20% of winners of major distance runs   Not due to better facilities Other nations also have altitude Based on the results achieved by the finest Kalenjin runners, Berkeley anthropologist Vincent Sarich, the co-founder of the field of genetic anthropology, statistically estimates that the average Kalenjin could outrun 90 percent of the rest of the human race. Kip Keino "We are natural athletes. Most of our runners are naturally born athletes. We have only to work on improvement, the technical part of it. We feel that running is our blood." Gender Discrimination  Harder to measure    Men & women seldom in same venue Often don’t play same sport Even same sport may vary • Tennis, figure skating, & “women’s basketball”  Direct competition?  Jockeys & auto racing & golf Gymnasts and figure skaters  Women not always victims  Title IX     Part of 1972 Education Amendments to Civil Rights Act Mandated equal access & opportunities for women in federally funded education programs 3 ways to comply  Funding proportional to enrollment  Show history of expansion  Interests of students accommodated Few programs in compliance  But NCAA certifies all Impacts of Title IX  Spurred rapid growth in women’s sports   1971: 294,000 HS girls 2002: 2,800,000 HS girls  Zero-sum game?  Cut men’s programs rather than expanding women’s College Athletics And the Amateur Ideal “When I entered coaching 43 years ago, athletics was a component of education. Today it is about making money and winning big. There is a professional intensity creeping into intercollegiate sports.” – Dale Brown, LSU Basketball Coach (1972-1997) “Cheating against the NCAA rules will continue-- indeed, increase-- because it is the profit-maximizing strategy for nearly all universities and the income-maximizing strategy for coaches.” - Economist Roger Noll. Overview     College sports is similar in many economic ways to pro sports, but the relationship between the athletics department and the university deserves careful attention Conferences and the NCAA play an important role in limiting competition, negotiating TV broadcasts, and managing competitive balance. NCAA player rules have dramatic impacts on the economic welfare of college athletes. Colleges enjoy special tax and antitrust status for much the same reason as pro owners. 5/29/2008 Amateurism & the Olympic Ideal  Ancient Olympics (776 BC-393 AD)   Even central myth hypocritical Winners well rewarded by home cities Pierre de Coubertin • Sought integration of education & athletics • “Mens sana in corpore sano”  Modern Olympics (1896-present)   Amateurism reflects class snobbery  Laborers not considered amateurs American Scene  Commercialism & Corruption always present   1st competition: 1852 Harvard v. Yale in crew • Sponsored at resort by railroad 2nd competition brought first eligibility scandal • Harvard’s coxswain had already graduated!  Second sport: Football   Rutgers v. Princeton (1869): First academic scandal • 4 Rutgers players were flunking math University of Michigan (1894) • 7 of 11 starters were not registered students NCAA as “Incidental Cartel”  Restricts movement   Prevents “tramp athletes” • Students – not hired guns Monopsony power • Players have little mobility • Drives down pay NCAA Recruiting Game OSU High High UM Low Dominant Strategy? Low 20 90 75 50 50 90 20 75 Equilibrium? Cooperative Optimum? Athletic Scholarships  NCAA forbade them until 1956   NCAA rules often ignored “Seven Sinners” They would be easier to police if out in the open Analogous to drugs or prostitution  The NCAA’s justification of scholarships   The “Student Athlete”  Scholarships created a problem    “Pay for Play” Students were effectively employees Sought workmen’s compensation for injury Disavows desire for pay Colleges do not have to provide workmen’s compensation In football >$500k/yr In basketball >$800k/yr  “Student athlete” is a legal term    Stars worth more than tuition (Brown 1993; 1994)   The Value of an Education  An athlete who…graduates is overpaid” Joe Paterno  Do athletes get an education? On average athletes graduate at the same rate as non-athletes  Average masks wide differences  Long and Caudill (1991) Male college athletes earn more than non-athletes Graduation Rates Vary by Sport and Quality   Money sports worse    Especially football & men’s basketball Only 4 schools had graduation rates over 50% 7 schools had rates at or below 33% 9 schools graduated 50% or fewer Only 31% of its football players graduated Only 10% of its basketball players graduated Consider 2006 men’s basketball’s Sweet 16 2005 football’s top 20 was little better    Texas was at or near top of both polls BUT   Women do Better  6 of women’s Sweet 16 teams had higher rates than best men’s team   Stanford & Duke > 90% Only Maryland <50% • • • • Duke: 100% v. 40% Boston College: 86% v. 31% UConn: 67% v. 27% LSU: 62% v. 20%  Head-to-head with men:  Women athletes graduate at rates similar to all women Why do Some Sports do Worse?  Some athletes less prepared  SATs, HS rank, HS gpa lower • True for basketball & - especially – football • Not so for softball or golf  Is dropping out a rational investment?   Go to Florida State to get to NFL? Go to Harvard to become a physicist? Academic Standards  Preserve academic integrity  Don’t recruit students who cannot read Established powers keep out new entrants Competitors cannot pay athletes more Now cannot take weaker students either  Creates Barrier to entry    History of Standards  Nothing uniform until 1965   1.600 Rule To play needed projected 1.600 gpa Ostensibly higher standards Actually needed C+ average in high school • Could take any courses • Worst abuses came under this rule • The sad case of Chris Washburn  1973: Replaced 1.600 with 2.00 rule   Proposition 48 (1983)  Provisions   Needed SAT=700 & GPA=2.00 in 11 core courses If not: no scholarship in 1st year & cannot play Disproportionately affected black athletes • SATs for blacks average 200 points lower • Are SATs a valid predictor of college performance?  Was Prop 48 Racist?   Still – graduation rates rose for whites and blacks Could receive aid if pass one criterion  A concession: Partial Qualifiers  Proposition 42 (1989)   Meant to eliminate partial qualifiers Loophole restored – and then some   Under Prop 48 scholarship “counted” Under 42 doesn’t count against limit Proposition 16 (1992)  Created sliding scale  Lower gpa permitted if SATs higher & vice versa    Clearinghouse evaluated individual courses Allows partial qualifiers to practice Challenged in court   Students claimed disparate racial impact • Won initial case Verdict overturned on technicality • NCAA does not disburse federal funds Latest Revision (2003)  Eases initial restrictions   14 core courses (up from 13) Sliding scale • 2.0 core GPA requires 1010 SAT • 3.55 core GPA requires 400 SAT  No Partial Qualifier status Need 40% of degree requirement after 2nd year Need 60% of degree requirement after 3rd year Need 80 % of degree requirement after 4th year  Stiffens progress requirements    Academic Progress Rates (APR)  School scored for student progress   1 point if athlete stays enrolled 1 point for staying academically eligible Consider Big State U’s basketball team • 52 possible points (13 players *2 points*2 semesters) • If one player is ineligible in spring – lose 1 point • APR=100*(51/52)=981  Computes % of total possible points   If its score falls below 925, BSU could lose scholarships Entry Barrier or Academic Standards?  Small schools  May be unable to compete with larger schools  Faculty fear Greater pressure to pass  Proliferation of garbage classes  Profitability of Specific Programs at Division I-A Schools (measured in $1000s) Sport All Men’s Sports Football Men’s Basketball Women’s Basketball 1997 3,300 3,200 1,600 - 500 1999 4,000 3,700 1,600 - 600 2001 4,900 4,700 1,600 - 700 2003 6,100 5,920 2,020 - 775 All Women’s Sports -2,300 -2,400 -3,200 -3,600 Source: Table 11.8, Leeds and Von Allmen, 2008 NCAA Financial Database Football Coaches Salary Database March Madness   NCAA has 11-year, $6 billion contract w/CBS Tourney revenue now $425 million/year   $389m in TV rights $35m from ticket sales, etc. $132.6m distributed according to program size • Number of sports offered • Number of athletes on scholarship.  ~60% goes to Division I conferences & schools   $132.6m distributed according to performance • Conference gets 1 "unit" per member game • Each unit worth ~$177,000. March Madness  ~60% of the performance distributions   Went to 6 major conferences • ACC, Big East, Big 10, Big 12, Pac-10, & SEC Big East gets most: $16.7 million   9 small conferences each got $784,182. Most conferences divide revenue evenly Non-Profit vs Profit Seeking Principle-agent problem  Growing payroll costs for sports programs 