Saturday, May 17, 2014

Sometimes I Deserve to be Spanked

After posting, "Rays Spring Training Economic Impact Study is Embarrassing," I received a response from the Charlotte Harbor Visitor & Convention Bureau (VCB), rebutting and (fairly) spanking me on several points.  The chief among them was that I used incorrect tax data to show March revenues hadn't spiked in the county.  I was wrong.

Turns out, I was citing tax dollars reported by Charlotte Co. in March, not collected.  Because there's a lag, my data didn't reflect the typical spring break boom, so I apologize.  In actuality, Charlotte County enjoys a big boost in tourism every March (even though it's truly impossible to know how much of it is baseball-related).

The VCB also countered other points.  Including one that claims teams that train close to their regular-season homes (i.e. Rays) don't draw as many out-of-towners as teams like the Red Sox or Twins.

Their response:
Fans of both the home team and the visiting teams attend spring training games and were captured in the study’s findings. For example, on the day the Rays played the Toronto Blue Jays, the study captured a number of Blue Jays fans, including some who had traveled from Canada.
True, but if fans were really making a trip to see their favorite team play one game as the visitors in  Port Charlotte, its unlikely they'd stay the five nights the VCB suggested in its press release.

On my claim that Rays fans from "Hillsborough, Pinellas, Manatee, Sarasota, and Lee counties...were unlikely to spend an overnight," the VCB wrote:
Some 22.7% of the Rays Spring Training game attendees interviewed told us that they were spending the night in Charlotte County. The remainder were day-trippers to the destination and only their spending in Charlotte County for the day was included in the study.
On my conclusion that their economic impact numbers seemed inflated:
Mr. Pransky is characterizing total economic impact, which includes visitors’ direct, indirect, and induced impacts as direct visitor spending. Game attendees who stayed in Lee County or Sarasota County were treated as day-trippers and only their spending in Charlotte was included in the estimation of economic impact. Day-trippers reported average spending in Charlotte of $60 per person. This includes attendees’ spending at the game, as well as their related spending in area restaurants and retail facilities. Overnight Charlotte visitors average spending was reported to be $672 per person. This is the equivalent of $112 per person, per day of spending for the game, food, lodging, other entertainment, retail purchases, etc. The combined direct spending of day-trippers and overnight visitors for the 14 games is estimated to be $13,756,400, or an average of $982,600 per game.
That means Charlotte Co.'s economic impact of $21 million is based on $13.8M of direct spending and $7.2M in "trickle-down" spending based on 1,275 overnight fans per game.

It's not clear how much the study accounts for dollars that immediately leave the county, whether its to out-of-town hotel owners, the out-of-town Rays' owners, or national food chains.

But the even bigger question I have is why does it not (apparently) take into account the fact that some of the overnighters who attended Rays games may have visited Charlotte Co. anyway, even if the Rays weren't there?

The study also appears to neglect the fact that many of the folks who traditionally vacation in Charlotte Co. may now choose to go to a Rays game on a weekday afternoon instead of the mall, the movies, or the golf course.

So while Rays contribute to the hyper-local economy around the ballpark, much of the spending is not new spending, but instead cannibalized spending from elsewhere in the county.  I have reached out to the author of the study regarding these questions on Friday and will post an update as soon as I have one.
UPDATE: I've now since posted more thoughts on the study here after speaking with its author.

The VCB, of course, called out my mistake of using the wrong tax data.  They go on to provide some of their own data from 2013:
Charlotte County’s local option sales tax collection rate is 1%. The {2013} data documents a $790,233 increase in local option sales taxes paid in March 2013 as compared to February 2013. This increase represents $79,023,300 additional taxable transactions in the Charlotte economy in March than February.
If the study is accurate, the Rays are responsible for approx. 1/4 of the county's total tourism success in March.  However, March has always been the best month of the year for Gulf Coast counties like Charlotte Co., even when they don't have spring training.  Looking back at historical numbers, Charlotte Co. enjoyed a March boost in the years prior to the Rays' arrival in 2009 too.

I will quickly admit, I make mistakes.  And I'll own up to them when I do.  My tax stat gaffe was foolish, and the VCB makes some good points in defense of its study. 

There's little question the Rays contribute to Charlotte County's bottom line each spring.  And, because the state paid the majority of the stadium's 2009 renovations, the county got a pretty good deal on landing the Rays.

But at the end of the day, this new information does little to sway my skepticism of economic impact studies paid by agencies or teams.  When they're motivated to show big tourism impact, it's not very difficult to find stats to back it up.

READ: Official response from Charlotte Co. Visitors & Convention Bureau:
Page 1
Page 2

UPDATE: I've now since posted more thoughts on the study here after speaking with its author.


  1. If, of everything you did last week, this is the thing that deserved a spanking, I'm jealous.

    I think I see how you came up with 14%. I won't try to determine how 14 compares to the lower limit of the margin of error that they didn't publish, but in this context, 14 doesn't feel terribly different from 19.5.

    However, their "68, 22.7, 9.3" does feel different from (rosier than) "41.5, 13.8, 5.7."

    It's strange that 22.7% came up twice in one study.
    Is "told us" underlined for hilarious emphasis, or was that originally a link to something? (When I'm so far from home that I choose to stay in a hotel, I'm not good at knowing where county lines are. How did they ask?)

    Maybe the underwhelmingness of most of their responses is part of the reason they sent feedback privately instead of posting it as a comment (which would have made all of this easier for us readers to follow). But it's good that they tried to be thorough.

    I hope they'll explain the methods used to get from $13.8M to $20.9M.

    Unspinning statistics is tedious work. I'm glad you're giving it attention. And those slips are learning opportunities for everyone who notices. Very good work.

  2. The "told us" underline was done by the VCB, presumably for emphasis.

    The $13.8M to $20.9M includes a "multiplier" that economists often use to account for "trickle-down" economy. When a waiter gets more income, he/she is more likely to spend those monies locally too. Therefore, taxes and income are essentially collected twice on that initial revenue, and hence, the multiplier. Not saying it's right, but it's common in economic impact studies.

  3. Yeah, it makes one wonder how they asked. I wouldn't expect someone from out of town to reliably answer which county their hotel was located in. If that was the survey question, then the data are weak, so it's funny for VCB to get sassy about it... especially when they weren't even contradicting your statement.

    I was affirming what you suggested earlier, that the use of a multiplier raises the question of whether the multiplier was chosen appropriately (such as by considering where the money likely goes). If they had simply published the whole study instead of only selected highlights, we wouldn't have to wonder about these details.