What Explains Labor’s Declining Share of Revenue in Major League Baseball?

Somehow I had missed this earlier paper by John Charles Bradbury:

Since the early-2000s, the share of revenue going to Major League Baseball players has been diminishing similar to the decline of labor’s share of revenue observed in the US economy. This study examines potential explanations for the decline in baseball, which may result from related factors and provide information relevant to explaining this macroeconomic trend. The results indicate that the value-added from non-player inputs, collective bargaining agreement terms, and related changes in the returns to winning contributed to the decline of players’ share of income. Competition from substitute foreign labor and physical capital are not associated with the decline in labor’s share of income in baseball.

There is also this sentence:

The decline in labor’s revenue share in MLB is consistent with changes in revenue share in the hospitality and leisure industry that experienced a decrease in labor’s share of income from 65.7 percent to 62.1 percent between 1987 and 2011 (Elsby, Hobijn, and Şahin 2013).

Another hypothesis I have heard is that baseball players are not nearly as good at, or as well-suited for, the use of social media, as compared say to the more visible basketball players.  Another (quite speculative) claim is that sabermetrics has commoditized a lot of players and in turn lowered their bargaining power.


Doesn't sabermetrics make players more fairly priced vs market price of peers, but not necessarily cast a judgement on the value of a player itself?

Sabermetrics say a players value is made up a lot by their defensive ability. No one pays to watch defense and sports is entertainment as much as jocks want to infuse it with dignity and distinguish it from movies or WCW. So hitting which puts butts in seats is devalued and a players value is the product of his impact on the team (value A) and his ability to drive interest in the team (value B) so someone like adam Dunn who could rake but plays no defense and as such has a limited impact on his teams performance sees his value A diminish relative to some light hitting center fielder who generates no interest in the team. The net result of this transfer of value between players is a net decrease in their value to the team.

It’s really instructive to look at the list of WAR leaders for a given season and then compare them to the household names. Last year was uncharacteristic in the fact that the correspondence was actually pretty good and still there are a ton of names that inspire only- who?

You're almost as perceptive as a typical sportswriter.

You are a giant bitch.

The latest trend is to downplay defense on the grounds that batters strike out so much these days that defense is less important. So the Dodgers last year would do weird things like have their first basemen switch to centerfield and their left fielder to second base in the middle of the game.

That’s a trend but it’s hardly the trend and it’s certain has effected WAR calculations yet. It would be great if it did positional adjustments are likely the most occluding factor in sabermetrics at this point.

The original Moneyball was almost the opposite of this, in downplaying the role of defense (which Beane thought was overpriced).

The classic Moneyball find is a bad body college hitter with a lot of data to go on.

And then things changed. And moneyball as Steve will explain was a way of laundering steroid usuage via stats. Moneyball at its base was signing jermey Giambi and keeping him supplied with clean needles.

There was a lot of continuity in style between the 1980s Bash Brothers Oakland A's of Jose Canseco and Mark McGwire and the 2000s Moneyball A's.

The "Moneyball" A's, the team built by Billie Beane, has appeared in post-season play nine times during his tenure as general manager of the club that began in 1998, failing to advance to the world series during his tenure.

Ballclubs run sabermetrics on players but players run no such thing on ballclubs. This instantly gives the information advantage to management. More to Tyler's point, I suspect labor's decline is due to 1) baseball has lower margins compared to basketball or football and 2) their top fans are aging out leaving the sport vulnerable to millenial preferences (which ties in to his social media idea).

The players and their agents can analyze too if they want. I recall something about the clubs courting Harper bringing charts showing how his hitting would do in their ballparks. I expect Scott Boras had already had such charts created by his staff.

They can only analyze the data if they get their hands on it. Players are now being asked to wear movement-tracking devices, there are high-speed cameras that capture the ball's velocity and angle off the bat as well as the RPMs of pitchers' spin on the ball.

Those sensors all have data feeds; the key question is where do the data go? To computers owned and operated by the owners, or owned and operated by the players? Do the owners share the data with the players and their agents (my guess is no)?

The article that I linked to suggested that very very few, maybe only one player, has asserted ownership over his own data. What's not clear is how he did it; does his contract say that he will refuse to let himself be videotaped and movement-captured unless he gets the data? That the data feed from his movement-tracking devices will go to his computer and not the team's?

Assuming clubs are all using similar metrics, what sabermetrics does is reduce the variance in evaluation. So when a player hits free agency, a club is less likely to bid more than the rest of the market thinks he's worth. Or, to put it another way, this probably reduces Winner's Curse, wherein it's likely that the winner of an auction (which free agency essentially is) has likely overbid because by definition its valuation of the player was higher than everyone else's.

But shouldn't that *increase* the compensation to players? If the Winner's curse goes away completely (ie every team gets an identical accurate signal) we know what the player will be paid, and this will be higher than the player is paid when the winning bidder adjusts his signal for the Winner's Curse (in expectation.)

Or more likely baseball economics rely dispostionatlely on tv deals with regional sports channels and those channels and the tv deals that are in the back end of a bubble. Owners know it’s coming and are responding by spending cautiously. The amount of time that Time Warner is going to be able to charge every cable user 5 bucks for root Southwest is about up.

Also baseball’s fan base is old and mature enough to be over the jock sniffing stage when you actually get mad about complete strangers not getting paid enough. Basketball’s fan base is almost entirely sniffing Stans and the NFL’s revenue split is pre-negotiated.

Every football player but like six QBs, three receivers, two running backs and 4 defensive players are basically commoditized too. But the tv contracts the NFL signs are far more likely to persist.

I’ll add that the revenue split is overstated because baseball has a huge pension and medical insurance overhead that generally isn’t included in these calculations.

The tv deal thing sounds plausible, anyway.

They're going to need to negotiate a new collective bargaining agreement shortly as well, another reason to play it safe.

The premise of this paper is simply wrong. Labor's share of revenue has not declined. In fact, the MLBPA has come out and admitted this. I will look for the link. You only come up with declining player share when you ignore important inputs into the labor share OR add in revenue unrelated to baseball operations like MLBAM

Thanks for resetting my last paragraph. But the trend is absolutely slightly downward. It just hasn’t dipped below 50 percent yet.

"Or more likely baseball economics rely dispostionatlely on tv deals with regional sports channels and those channels and the tv deals that are in the back end of a bubble. Owners know it’s coming and are responding by spending cautiously. "

This. And I think the behavior of the Nationals, with their proclivity for deferred money in contracts along with the continued uncertainty regional cable deal, is probably a good example of how many teams will start approaching their spending. I can't imagine that the regional cable gold rush will last too many more years.

The speculative claim about the role of sabrmetrics is well-covered here:

There's more data, and more use of data, than ever before in baseball. But the owners rather than the players have access to or ownership of the data. Knowledge is power, and the owner typically knows more than the player does.

Because baseball is the most boring sport just a hair above NASCAR. Basketball has the best last five minutes of any game and football has pretty good action from beginning to end. Soccer is like a lower scoring version of football with the benefit of no commercials.


Soccer on ice, with a ball that can't be seen on TV.

Get a nicer TV, then!

It's true that television coverage is the validation of anything that occurs in the US of A. If it can't be seen on TV it's imaginary.

Basketball is boring. The games are indistinguishable, and end in a free throw shooting contest.

Best last five minutes? Ever watch the last five of a 1-goal hockey game?

It would be interesting to see how wage inequality accross players has changed at the same time.

Speculative: government finds it easier to tax labour income compared to business income. So labour is increasingly being outsourced. This outsourced labour is then measured as some sort of non-wage income. Thus it looks like labour income is declining. A similar argument might explain the delinking of per capita GDP growth and wage growth.

I work for a company that has been around for 75 years or so. We employ half as many people as we used to. At the same time our Purchasing department has become the 2nd largest office function. I can't prove it but it looks like we have avoided labour taxes and regulatory cost by outsourcing functions.

"Another hypothesis I have heard is that baseball players are not nearly as good at, or as well-suited for, the use of social media, as compared say to the more visible basketball players."

Baseball uniforms make fit athletes look the same as any beer drinking fan, so they're not aspirational figures to young kids.

On the same topic, the average age of NFL and NBA players is around 26 years old. For the MLB is 29 YO. I'll just say that age make us less photogenic.

Finally, ~30% of MLB players are foreign. Good luck selling that.

"Baseball uniforms make fit athletes look the same as any beer drinking fan, so they're not aspirational figures to young kids."

Ha! Modest PED enforcement is what makes baseball players look the same as any beer drinking fan. The uniforms aren't helping, but the different drug regimes in the NFL and NBA are most of the problem.

This isn’t even that true of males especially in the age range that applies to pro sports. On average baseball players are probably the best looking atheletes because it’s the one major sport that doesn’t require you to be something of a freak in terms of size.

Interview Bill James.

Possible contributing factors: Teams have gotten better at finding and developing players, raising the level of the replacement level player, and hence the gap between the star level players and the replacement level, and hence the marginal value of the star player in terms of wins.

At the same time, I think revenue streams may have become less correlated with on-field success. In particular, many baseball teams have lucrative local cable deals that are locked in long-term. This reduces the marginal value of the star player in terms of revenue generated.

Unlike the NFL and NBA, major league baseball has a substantial ongoing investment in players that don't play for them and probably never will, minor leaguers. Novice football and basketball players are developed for their professional leagues for free by the bogus US higher education system.

At the same time, major league baseball players are hardly candidates for welfare.

Thank you - I was not aware they gave away all the tickets and beer at minor league stadiums.

Yes, there is much more infrastructure in baseball. Being competitive means tending to the infrastructure/organization/development. That's more obvious than ever.

The NBA is a players league. Little infrastructure/overhead, small pool of workers (400-500 total); ideal collective bargaining conditions for players.

It's all about real estate now. See ATL Braves Battery Park. Clubs are getting more and more of revenue from lease, sale, and profit share in real estate ventures around the park, concessions inside the park, etc. The team is just the anchor.

Baseball owners - through Sabermetrics - have figured out that players start to decline around 28. That means it makes little sense to pay most players who reach free agency (typically around that age or even a little older) a big contract. This has created a problem where a player is underpaid for their best years, and then no longer receives their deserved reward during free agency.

This comment makes the most sense of all.

Labor share of income is just another marxian metric that is meaningless in economics.

All it means is that the marginal revenue product of other inputs has grown faster than that of players. This is not surprising at all. The players are not becoming remarkably better in their game or doing other things to draw attention to the sport. There are diminishing marginal returns to every input, and players are likely near their peaks.

Tyler's explanations sound plausible. It is still like measuring the CO2 content of air in your car tires compared to the CO2 content of air in your seat cushions: scientific, precise, yet meaningless.

MLB led the development of advanced streaming technology through BAMtech. Disney bought most of it for something like $1.5B because it was widely regarded as a leader in the field and an enabler for Disney+.

So this might be more of a case of MLB having more interesting and productive side businesses than comparable leagues.

Two hypothesis:

1) because the stat is based off "revenue," the simple answer is that fixed costs (stadiums, travel, etc) are eating up a larger share.

2) the more fun explanation is George Steinbrenner's death. Baseball players had a great combo going: (i) free agency bid up top talent and (ii) arbitration ratcheted up everyone else using new 'market' values. Now that free-spending George isn't around to push (i), that combo isn't working as well.

Increased reliance on immigrant labor, holding back or driving down the wage level in jobs formerly held more exclusively by Americans?

I'd say that the owners have finally figured out the Winner's Curse when it comes to free agents.

another factor: segmentation
the owners have segmented players into rookies and veterans. In collective bargaining, they can tell veterans that the lower money to rookies will end up going to the veterans (and then in the years after the agreement, replace veterans with more cost-effective rookies). The player's union is controlled by veterans.
Potential solution for player's union: demand a certain fraction of revenue must be spent on players in the next CBA contract.

Other possible causes (none of these are original):

1) Increased revenue sharing de-incentives winning so teams are less likely to bid against each other as much for talent

2) Owners today are more likely to come from a business background and treat the team like a business than in the past where (some) owners were less savvy and more concerned about winning (not that there weren't always cheapskates out there)

3) Along the lines of the sabermetrics idea, teams can more accurately value players today and these valuations are likely to be similar to other teams' valuations (i.e. reduction of "winner's curse").

Tanking is a bigger strategy today than before.

There has been collusion before (e.g., 1987), so it would hardly be surprising if there were some degree of collusion going on now.

For example, tanking could be reduced by introducing a lottery into the baseball draft, as was done in the NBA to reduce tanking. But at present, tanking seems to be benefiting the owners' profits, so it is allowed.

Sabremetrics is the answer: the knowledge and insight provided by management via sabermetrics is generating a larger share of the success of teams than in the past, so individual players are less valuable.

Sabremetrics also makes the game more exciting because it means a larger number of players contribute to teams' success by being matched to their most effective roles - so you don't necessarily need Nelson Cruz at the plate in that 2-out bottom-of-the-ninth at bat.

Interesting to note also that the Angels, long the team of big-money free agents, almost certainly haven't been successful in proportion with their spending.

Is is that baseball players aren’t as good at marketing themselves on social media, or that the environment of mlb is inhibiting them?


In this video, Gary vaynerchuk who is a social media expert and is big in sports representation makes the claim that because mlb has restricted its uploaded content unlike basketball, it has caused it to be inhibited on social media and thus experience a stagnant growth. Highly recommended watch.

Comments for this post are closed