Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
Incentives and The Land Of Unintended Consequences, Charity Style…

July 9th, 2009 · 5 Comments

Today’s economics mantra: YOU GET WHAT YOU PAY FOR. Actually, that should be every day’s economics mantra. As someone who writes about incentives, I end up hearing a lot of stories about various compensation and pay for performance schemes in organizations (and in households, relationships, etc., actually). These stories tend to range from the only mildly offensive to the truly heinous. Obviously the heinous stories are the most interesting to me, but I’ll take what I can get. I eventually plan to publish a book on how to design good incentives in organizations, and it’s helpful to have a litany of anecdotes to pull from. Here are a couple to get the party started (I welcome comments and emails with other anecdotes, for the record)…

First, the only mildly offensive:

I posted a while back about how my mother’s semantics lesson resulted in me shoving raisins up my nose. (No, this was not last week. Okay, that one time when I was drunk, but it was funny, ok?) My friend Dave responded with a story about how when he was little he would pour soda out of the cans and bottles in his parents’ house so that he could get the cash for the deposit and/or recycling. This is clearly not profit-maximizing from the parents’ perspective, but it was completely rational thing for Dave to do.

Lesson learned? Smart kids are often huge pains in the ass. Well, yes, but not entirely the point…the actual problem was that the benefits and costs of the soda situation were decoupled for Dave- he got the benefit of the empty bottles and cans without having to think about the cost of buying the soda in the first place. So Dave maximized his payout without thinking of what costs that imposed on his parents. Note that Dave wasn’t really doing anything wrong (unless his parents were prescient enough to anticipate and regulate the behavior), but I am guessing that once his parents got wind of this they put a new rule in place.

Now, the truly heinous:

Maybe you read the above example and thought “Oh, that’s cute, but people figure out how to behave as they get older.” I think the last decade or so of scandals has taught us that that is really not the case. (Enron anyone?) Furthermore, it’s often not even the case that organizations are more clever at designing incentive systems than Dave’s parents were, they just get to try them out on a larger scale.

Consider another friend of mine who works for a consulting firm. (This story is intentionally vague to protect the identities of the not-so-innocent.) This consulting firm is actually a subsidiary of a larger company, and there are bonuses handed down from the parent company to the consulting firm. The bonus is a function of top-line revenue. Now consider that the consulting firm has a non-profit as one of its clients. Normally, non-profits get lower rates because the work done for them is partially a matter of goodwill. However, in this case the rate charged to the non-profit was actually higher than that of a regular client. Why would this be?

The situation makes more sense when you consider that the consulting firm also made a sizeable donation to the non-profit- what was really happening was that the firm was charging an inflated rate in order to put it in the revenue numbers (and thus increase the bonus from the parent company) and then donating most of the charge back. The firm could do so without tax consequences, since the client had non-profit status, and everyone was happy. Or were they?

The parent company would probably not be too happy to know about this, since it’s paying for something that doesn’t create value. That said, the consulting firm is technically playing within the rules of the game. (What’s the saying- “Don’t hate the playa, hate the game?”) The parent company basically did the same thing that the parents above (note the nice parallelism there) did- they decoupled benefits (or revenue) from costs. This gave the consulting firm an incentive to think only about the top-line numbers and ignore the costs that they were putting on the books. This is likely problematic for the parent company, since it stands to reason that they care more about the overall profitability of its subsidiaries and not just the revenue that they can put on the books.

Lesson learned? Even adults know how to, and will, game the system. But the lesson goes beyond that…corollary to lesson learned: Incent what you ultimately want to achieve, not just what is easy to measure. In this case, it is likely that revenue is directly attributable to individual companies, whereas costs are often shared across subsidiaries. As a result, revenue is the easy thing to measure and thus the easy thing to incent. If the parent company tried to incent based on subsidiary profitability, there would probably be a lot of fighting over how shared costs are allocated, among other problems.

So what is a manager to do? Is it better to incent the wrong thing (or part of the right thing) or nothing at all? People usually assume in this case that something is better than nothing, but that is not always true. My take on whether “something” or “nothing” is the better answer depends on how much potential there is to manipulate the system and some much focusing on the “something” detracts from the overall objective.

One final editorial comment: you may be sitting here thinking “yeah, this happens sometimes, but I believe that a lot of people are basically upstanding and ethical individuals and thus wouldn’t take advantage of such a situation.” Granted, in the examples I gave it *should* be at least a little clear to people that they are not abiding by the spirit of the incentive, but that isn’t always the case. It’s quite possible that when a manager says “I am providing you a specific incentive for doing X and nothing else”, the employee honestly interprets that as “X is the only thing that matters to me.” In other words, never underestimate the signalling component of incentives.

I’ll leave you with a related excerpt from Arrested Development, one of my favorite shows ever:

(setup: George Michael and Maeby are working at a frozen banana stand.)

George Michael: Well, this is the cash drawer. My dad’s going to come by at the end of the weekend and the number of bananas has to match the amount of money in here.

Maeby: Oh, so it all has to even out?

George Michael: Exactly.

Maeby: Easy. Banana… (throws a banana into a garbage bin…) Buck. (and takes a buck from the register) Banana… Take a buck. (doing it again as George Michael just watches)

(later that day, at lunch…)

George Michael: You know, I think we might be doubling our losses here. Because, I mean, for every dollar you take, you’re actually taking two dollars because we paid for the bananas.

Maeby: (laughs) Oh, my God, you’re right.

So maybe (maeby?) people aren’t always good at gaming the system, but they certainly do try. 🙂

Tags: Incentives

5 responses so far ↓

  • 1 econgirl // Jul 9, 2009 at 4:30 pm

    Haha, I just now saw this:

    “President Barack Obama believes you get what you pay for — in business, in health care and in teaching.”

  • 2 Rev. Pfloyd // Jul 9, 2009 at 6:33 pm

    Here’s a personal experience on the idea of displaced incentives:

    I used to work in large-scale web-offset printing. Big printing presses that printed millions of ads that you get in the mail and in the newspaper. Anyway, the printing company ran 24/7, running four 12-hour shifts in a swing-shift fashion.

    In the interest of upping production, the company decided that at the end of a 28-day period it would pay the press operator a $200 bonus if he was the top producer that month in terms of just output of product. This seemed like an ill-conceived idea from the get-go. Theoretically you’re simply basing pay on performance but the company failed to see some obvious unintended consequences to this incentive.

    What started happening were two things: First routine maintenance on the machinery was all but eliminated. Since the presses ran 24/7, everyone would skip doing simply things like washing the printing blankets, cleaning up the press, shutting it down to make sure it was greased, and cutting corners on “make-readies”, which are the processes you do in order to switch things like paper sizes, printing plates, and the like to go from one job to the next.

    What was assumed that if Shift A didn’t do this maintenance, Shift B who followed them would have to do it. But then Shift B, feeling like they got the shaft by being saddled with all the maintenance simply refused to do it and hoped it would boomerang back to Shift A when they came back into work the next morning. And then, of course, none of this was communicated to Shift C or D when their swing days came around and then nobody would do any maintenance on the machines.

    The second consequence is that only the press operator got the benefit of the bonus, his crew such as the reel-stand operator (the person responsible for keeping rolls of paper running in the press non-stop) and the “joggers” (the grunt labor that grabbed piles of product off the end of the press, “jogged” them into neat stacks, banded them and stacked them on the skid) would see no windfall from the bonus. What would happen then is the press operator would run the press faster and crack the whip on his employees more who were now suddenly being forced to work harder without any additional compensation. They would then cut corners or do things that would sabotage the operation like the reel-stand operator neglecting to do a splice right, leading to all the paper blowing out of the press in mid-run (which required a lengthy process of re-running the paper back through the press) or the joggers letting the folders or stackers jamming (which required the press to be stopped).

    The consequences of both little maintenance and corner-cutting on the part of the subordinates created a downward spiral in production. Since there wasn’t a minimum rate of production required to snag the bonus, it soon became less of how much work you got done but how much work didn’t get done by other people. Production plummeted and the management couldn’t seem to figure out why. (They weren’t going to listen to advice from THIS lowly reel-stand operator. 😉 )

  • 3 Larry // Jul 10, 2009 at 2:35 pm

    I have a story that kinda fits…

    I worked at a fast food drive through place… you know the ones where people skate out to your car with food and have really good ice cream?

    Well I worked there for about 2 years when I first started college. I worked a few months as a cook… a few months as one of those people on skates and then about a year or so as manager (where I did all of the above and then some).

    The incentive programs where extensive and some where less incentive and more stick. We where required to get our orders out and maintain a certain average time per order. The only exception was that food orders and drink only orders had different requirements but the drinks where included with the food tickets for one count and the second count was just drink only’s.

    We also got “Mystery shopped” which means people pretended to be customers to grade our time and quality of food.

    The Problem:
    I worked in a low volume store. Everything is suppose to be fresh and essentially the other rules required us to basically cook to order except for the busiest hour or two per day. We simply didn’t have the volume to justify pre-cooking food. Now, the time it takes to cook most of the food in the store is between 2:30 and 5:00 minutes for fried food like fries and chicken and cheese sticks ect. Our average time had to be under 3:00 but they pushed for 2:30. Problem that should be apparent by now is that if you are cooking to order which is needed for freshness in that store we can’t do it in the same amount of time as the big stores that have food down all the time. I worked at the bigger store a little as well so I know the difference in timing.

    So what did we do? We cheated. We are in all kinds of competitions all the time graded on time and quality so we spent a lot of energy finding short cuts but even more on clocking orders early (Hitting the timer before they where really ready) and watching for mystery shoppers.

    See we learned the order patterns of the mystery shoppers, we found all the rules and decided that they had very narrow confines for their orders. Eventually we could pick out 5-6 orders a day that could potentially be a mystery shopper out of hundreds. Those 5-6 got special y treatment (we only had 3 shops a months) eventually we started recording those and narrowed down the make color and model of the mystery shop cars.

    Long story short the rules, incentives, and games didn’t increase the customer experience because they where impossible. We where being held to a standard designed for a different scenario and so we got creative. I did very well and my numbers and score where always very high franchise wide… I had low labor costs, low average times, high mystery shop scores and no complaints so corporate loved me…
    Don’t get me wrong, the customers when I was there where treated well and got their food fresh and in a reasonable time but the numbers corporate got where never representative of anything other than my ability to spin the game in my favor.
    They did do a good job of counter incentives though. They focused a lot on labor costs for us managers and time and quality so it was more difficult to cut corners. They also gave inventory and waste prevention bonuses for overall low cost. If they had set independent standards based on volume instead of across the board I’d say they scheme would have been great but when you corner people with incentives they cheat. 🙂

  • 4 econgirl // Jul 10, 2009 at 6:38 pm

    Ha. From Marginal Revolution:

    Andrew: “My friend Dave…would pour soda out of the cans and bottles in his parents’ house so that he could get the cash for the deposit and/or recycling…Note that Dave wasn’t really doing anything wrong (unless his parents were prescient enough to anticipate and regulate the behavior)”

    Of course he was doing something wrong. He was destroying value. Where do economists get this mentality? The older I get, the more I believe Objectivism is righter than bullshonomics, aka economics careerism.

    Jodi: I meant “wrong” to specifically mean “breaking a rule,” and I had hoped that this came across clearly in context. In fact, the objection that you raise is exactly the point that I was trying to make- that a. there isn’t always a correspondence between following the carrot on a stick and maximizing value, and b. you can’t expect people to abide by the spirit of the incentive rather than the rules of the game, either because their ethics don’t compel them to behave in a value-maximizing way for the overall system or because they aren’t aware of what goal the incentive was meant to encourage.

    As a proud bullshonomist, my viewpoint is NOT that people *shouldn’t* feel an obligation to behave in a socially beneficial way, but that bad (read, value-destroying) things can happen when you trust people left to their own devices to make choices that put social value creation ahead of self-interest. The fact of the matter is that there is a distinction between unethical and illegal, yet people seem to be pretty good at blurring the line between the two. I, therefore, take a point of view that tries to deepen that line in the sand.

  • 5 Alex Golubev // Jul 14, 2009 at 10:56 am

    These sorts of agency problems are the rule and not the exception. They’re present in all companies (managers, executives, and board of directors), government, schools, hospitals. you name it. However, you gotta ask yourself why this happens/happened/will happen? I think there are self reinforcing processes with money and power. That’s why we have three branches of government. That’s only working so so, because each branch tries to expand its roles. Judges are ignoring contract law in this environment. Cops redefine crime for statistical purposes (see The Wire season 4). Money breeds power and power trumps everything. Scarface taught us that a long time ago. The next question is whether we CAN change the system. I think the internet CAN help us redefine the power landscape. It’s a democracy machine. Other than that, i think we’re pretty f*d as a society. The democractic illusion of choice only holds up for so long. The reality has to be confronted eventually and our system and we may be wacked out on various resource misallocations (housing, medical costs, government, military, education) to stay competitive. But it’s all a relative race. Other countries have f*d up incentives too. Our networking is their bribery. So yeah, the book sounds like a good idea, but i think you’d need to explore some ways out to make it mroe than just anecdotes.

Leave a Comment