Ahh, the joys of small samples. Home runs are up, Kansas City is a contender and Howie Kendrick somehow pulled a Vice Versa with Albert Pujols. Thankfully, there are some statistics we do not have to wait until May to analyze. The average Major League salary is up 1.2-percent from last season, and, unlike Jorge Posada’s 162-homer pace, it’s for real.
This offseason, I attempted to model baseball’s salary inflation through regression analysis. For a quick review, here were the three main takeaways:
1) Salary inflation in baseball can be explained decently well using only attendance and the unemployment rate as independent variables.
2) Unemployment rate is the best economic indicator to include in the model, but it is unclear if including it as a raw variable or transformed into percent change from the previous year makes the best model.
3) The two models had very different projections for the 2011 season.
Now that rosters are finalized, we know the average MLB salary in 2011 is $3,318,855.20, representing a 1.2-percent increase from 2010. The first question that comes to mind is, “Which model made the better prediction?”
The first model, which includes the unemployment rate as a “raw” variable, predicted an inflation rate of 5.9-percent for 2011. The second model, which includes unemployment as the difference from the previous year, predicted a salary deflation of 6.4-percent. It would have been impossible for both models to be right, but truthfully, they both pretty much missed the mark.
It is important to remember that these models are intended to be directional. They project the market for baseball talent based on the cash in teams’ pockets (attendance) and the economic outlook (unemployment rate). On a year-to-year basis, expensive multi-year contracts have something to do with the stated factors, but also have a lot to do with the quality of the free agent class, television deals, the collective bargaining agreement and the economic situation of each individual team. The predictions may be very wrong for a given year, but they should provide educated outlooks for where the market is headed on the whole.
That being said, the first model did a better job predicting the actual inflation rate. While this is certainly a feather in Model 1’s hat, one data point is not enough sample to definitively say which model is better.
Although the season is early, there is enough data from early ticket sales to make some conclusions about what attendance will look like this season. Eric Fisher at the Sports Business Journal recently quoted Commissioner Selig that “after drawing 70.06 million fans in 2010… this year’s total will likely fall somewhere between 75-million and 78-million.” These projections translate to an attendance boom between three and seven percent for 2011.
The next step is getting an unemployment estimate for the rest of this year. The current rate is 9.2-percent and The Congressional Budget Office predicts it will stay at that level into the fourth quarter of this year. The average unemployment rate in 2010 was 9.6-percent, so the projection would mean a decrease of 4.5-percent from 2010. Note that this is a percent change, not a difference in percentage points.
Plugging attendance and unemployment projections into the models, it becomes possible to forecast salary inflation for the 2012 season:
Again, the models give two very different views of where the market is headed, both with valid arguments:
– Model 1 says that the economy is turning around. Unemployment is headed in the right direction, ticket sales are up, and we can expect somewhere between 7.9-percent and 9.3-percent salary inflation in 2012 depending on just how high ticket sales go. Heck, teams that invested heavily in the 2011 offseason may have gotten a good deal on long-term contracts considering where the market is headed.
– Model 2 says that the unemployment rate is still extremely high, the country is in a recession, and that has to catch up to baseball at some point. Depending on the final attendance figures, the market should experience deflation in 2012 somewhere between 2.5-percent and 3.9-percent. Teams who spent big in the 2011 offseason will be regretting it soon.
The actual market for talent next year will probably not be predicted exactly by either of these models. However, depending on if you are an optimist or pessimist, you can make an educated guess for what the market will look like. In a more general sense, these two models provide a ceiling and basement for salary inflation in the near future.
The most pressing application for this analysis is for the massive contract which will be negotiated for Pujols next offseason. As detailed here, a 5-percent salary inflation rate would yield an approximate $267 million market value for a 10-year contract. If Model 1 is correct, the market will continue to rise at an average of, perhaps, 7-percent in the next 10 years, then Pujols’ value is 10/$294. If Model 2 is correct, there is a market correction coming, and salaries will rise at an average of 3-percent in the next decade, then Pujols’ value should be around 10/$244.
Forecasting the salary inflation rate becomes very important when a couple points one way or the other means $50-million dollars. The teams gunning for Pujols will definitely have to do some inflation estimation before offering a contract. It is likely that the highest bidder for El Hombre is the team which is the most bullish on the salary market.