Quantile Regression By Ashley Nissenbaum About the Author

Quantile Regression By: Ashley Nissenbaum

About the Author • Leo H. Kahane • Associate Professor at Providence College • Research • Sport economics, international trade, political science • Editor of Journal of Sports Economics

Previous Research • Golf earnings are highly positively skewed • Schmanske (1992) • Value of the marginal product from putting may be in the range of $500 per hour of practice. • Alexander and Kern (2005) • “Drive for show, putt for dough” • Callan and Thomas (2007) • Skills determine score, which determines rank and thus earnings

Earnings and Skewness • Linear Regression • Focuses on the behavior of the conditional mean of the dependent variable • Most people make under $300 K per event

Reasons for Skewness Payout Structure • Non-linear • Top 50% after the first two rounds: 1 st place receives 18%, 2 nd place receives 10. 8%, 3 rd place receives 6. 8%, 4 th place – 4. 8%, etc • Extraordinary Talented Golfers • Tournament wins are spread across a large number of golfers

Tiger Woods • Won 185 tournaments • 14 professional major tournaments, 71 PGA Tour events • $500 Million net worth • Highest paid athlete from 2001 to 2012 • $132 million from tournaments

Concept of Quantile Regression • Equation for Quantile Regression: • Where: • • • y(i)= real earnings per PGA event Q= Specific quantile associated with the equation Β = Vector of coefficients to be estimated Ε = Error term X(i)= Covariates

Covariates • x(i) = covariates expected to explain golf earnings • Greens in regulation • The percent of time a player was able to hit the green in regulation (greens hit in regulation / holes played x 100). Positive correlation expected. • Putting average • Average number of putts needed to finish a hole per green hit in regulation. Negative correlation expected. • Save percentage • Percentage of time a golfer was able to get the ball in the hole in two shots or less following landing in a greenside sand bunker (regardless of score). Positive correlation expected. • Yards per drive • Average number of yards per measured drive. Positive correlation expected. • Driving accuracy • Percentage of time a tee shot comes to rest in the fairway. Positive correlation expected.

Empirical Results • Simple level OLS (Ordinary Least Squares) regression estimate:

OLS and Quantile Regression Results

Coefficients Graph