A former Michigan office mate (and avid runner) sent me a link to an article about economic impact of the Chicago Marathon. I think it was mostly to convince me that other sports matter besides baseball, but let's be serious ;-). The link is here.
Anyway, I always enjoy taking a look at these articles and the multipliers they use. This one was pretty out of whack (I'll get to that later). There's plenty of literature out there that speaks about the importance of accounting for all the effects and leakage of these events, but these reports keep on coming. Luckily, Allen Sanderson provides us with some common sense.
Now, there are certainly great things about hosting a marathon. There's the possibility that it sparks interest in physical activity (though, this phenomenon is not well understood). It certainly brings some people into the city. People definitely spend money. People seem to have fun--for whatever reason--running or watching others run 26.2 miles. But when we talk about economic impact, we don't really care about the total amount of money spent in the 3 days or so. We care about the additional money spent in those 3 days or so.
A large majority of the spectators of the marathon are likely from Chicago. They would be spending this money in the city anyway. So we need to account for this fact, which reduces the initial impact estimate (pre-multiplier) by a huge amount. We need to know how much non-Chicago native money comes in (a point also made by Sanderson above).
Streets generally get closed for these sorts of events. That could be a net negative on both business and the residents in the city, or those that commute to the city and have to take a longer alternative route. If the marathon wasn't going on, there would also still be plenty of entertainment in Chicago to take its place. So maybe people that would have come for that alternative entertainment don't come during the marathon.
Does it create jobs? Maybe, but it is doubtful that there is any real job creation. Most of the people working these sorts of events tend to be volunteers. And even if there are some new jobs, they'll be very temporary. Marathons may seem to last forever for the runners, but they're really just a weekend event (or less).
What about hotels and restaurants? Well, they're probably liking the event. Most likely, they have some increases in rooms and room rates during this short period, but as the second linked article above notes (and does the math), it is nowhere near the estimation from the first linked article.
What about multipliers? There are reasonable defenses for using a multiplier. After all, when you buy a bagel and coffee at your local store, at least some of that money is likely redistributed and used again within the city (the local employee gets a salary, the owner takes his or her cut, maybe they use the sales to buy coffee grounds from a local coffee grower...not likely in Chicago). If you buy coffee at a place like Starbucks or McDonalds, it is likely that less of this money stays within the city limits to be spent again than a locally owned shop. But in any case, some of that money stays within the city and gets spent again, so we use multipliers.
A huge problem with almost any impact study is that these multipliers are way out of whack. The one above uses a multiplier of 2.29. Or, for every dollar spent, it supposedly results in a total of $2.29 of economic activity within the city limits of Chicago. Conveniently, Mark Rosentraub and I have a forthcoming paper looking at more reasonable multipliers due to leakage using some VERY simple hypothetical examples. Now, keep in mind that our examples could be overly simplistic. But the point is that the multipliers really depend on the size of the region, the types of industry located within the region, and the commute sheds of the region. What do we get as more reasonable levels of a multiplier? About 1.28 or less. Or, for each dollar spent within the city--above and beyond what would have been spent anyway (this is important)--creates an additional 28 cents in economic activity beyond that dollar. That's a far cry from an extra $1.29 in activity.
So if we say that only 25% is NEW spending, then use our multiplier, we get something closer to $25-$30 million in impact. That might be great for the city. But it of course depends on what the costs were to have security there, close the streets, and so on. There could be other negatives (that are not as easily measureable) like residents hating the crowded streets, having to take on more travel costs getting around the city, taking a day off work or leaving early to avoid incoming marathoners on Friday (possibly lost pay for the worker and lost production for the company).
To be sure, these estimates are difficult to make with pinpoint precision. But using more reasonable guidelines at least gets us closer to what is really going on. How about we start doing that...