Monday, January 2, 2012

Paying for Your Kid's College:The Principal-Agent problem

The financing of my daughter (3) and her sister's (due April 2012) education, and in particular college education is never too far from my mind. What is the best way to save and pay for college? This is a tough question given that there exists a lot of uncertainty. For starters I do not know IF either of them will decide to attend college. It is entirely possible that they may not have the aptitude or desire to continue studies beyond secondary school. Maybe they will seek a liberal arts education in the Marine Corps instead of at St. John's College. Even knowing what their career decisions will be at age 18, the saving-financing strategy is not very clear at all. Sure we could save but every dollar that gets put away for college is one less dollar for consumption today or for retirement tomorrow.

For a quick refresher as to why financing college is a difficult strategic task aside from the simple fact of scraping together a lot of money see The Atlantic article on how savers are punished when it comes to financial aid and a Fidelity website to assess the huge differential in cost between public versus private education. Then consult a CPA, a CFA, and an from an economist.

Not to dismiss those questions and issues but I think the more important question is "Should I pay for my kid's college?" and if so how much and in what way. The answer to those questions are fundamental and are prerequisites to untangling the mess above. This is where economics comes in handy.

What I want to know, before laying out big money for the holy grail of middle class aspirations, is my ROI, return-on-investment. And I don't mean that in the strictest sense as in : how much money will I make off my kid for financing their college? Is the money is well spent? Are they better prepared and equipped at age 22 to deal with the world than they were at 18? Or will they be more dependent on me afterwards or god-forbid will I be dependent on them? How can I prevent my child from wasting four years of their life and gobs of my money? Will they go to university and learn to think, learn a skill, learn about the world? Or will they learn how to waste time, get wasted, and waste an opportunity?

Before signing checks or signing for a loan wouldn't it be great to have knowledge of my child's intentions for the 4 years of college? If we could accurately predict that Johnny was going to spend four years drinking, not attending class, partying, attending football games, and in the end not graduate then we'd rather spend the money drinking and attending football games ourselves. At least then we'd get something out of it. Instead if our accurate predictions told us that Johnny was going to try hard in college, grow as a person, learn a lot, and graduate then we'd be disposed to help him achieve that goal in any university we could afford.

(Before assuming that we parents know our children well enough to know if they are going to be studious and successful or stupid and irresponsible at college consider the statistics compiled by Complete College America. Granted the statistic that only 60% of full time college students complete their 4 year bachelor degree within 8 years is a little misleading. It is not a random 60%. Better students with higher College Board scores are more likely to do better and not surprising children from wealthier families will do better, on average. And not all students who fail to complete college do so because they partied their way out. Lots of people for lots of good reasons take a while to get there.)

This is pretty much what economists (and lawyers) call the Principal-Agent problem. It is a problem where the Principal (Boss) wants to hire somebody to do something (Agent) that requires effort, includes some uncertainty, and is costly for the Principal to monitor the Agent. The classic example is the Landowner hiring a Farmhand to work the land. If the Landowner pays the Agent a fixed amount, say 5 dollars per hour, or 50 dollars per day, or 1000 dollars total to grow and harvest the crop there is always a danger that the person will slack off because they get paid regardless of what is produced. The Landowner may be hesitant to pay wages on that scale because he would have to monitor the worker constantly. On the other hand the Landowner could offer a salary based on the amount of crop yielded. On the face of things this is not too bad of an idea because ultimately this is what the Landowner is going to sell for his profit. More is better. But because uncertainty is involved the crop yield is not fully predictable year to year. So if there is a drought or floods or whatever the Farmhand might not get paid after putting in all the work. He may refuse to work under that type of contract. 

In this case the Principal is Dad and the Agent is my Child. The Dad wants to structure the financing of Child's college in such a way to avoid paying for a 4 year lifestyle of hedonism. Or in risk management parlance, I want to make myself robust against the Black Swan slacker.

Here is what I came up with:

Assumption1: Agent-Child is either A) a studious person or B) a slacker with a short time horizon or C) a slacker with a longer time horizon (I'll explain the difference but rest assured it matters)

Assumption2: Principal-Dad does not know if he has an A-type, B-type or C-type Child and won't know until four years later.

Assumption3: The total four year costs to attend a private school is roughly three times that of an in-state college. To make this concrete lets use 50,000 and 150,000. So assume the four year total cost of sending my child to Florida State University is 50,000 and to a fancy New England liberal arts college...say Wesleyan College in CT, is 150,000 (hint: it's actually much higher).

Assumption4: Principal-Dad either has enough financial resources already or can finance the cost of college and repay it over time.

In this problem I cannot know in advance if the Agent-Child is A, B, or C. And in reality there is no difference between B and C if the Agent-Child goes to college. What I hope to do is dissuade B (the slacker who wants instant gratification) from attending college at all. And what I'd like to do is "nudge" C (slacker who wants to spread the partying out over 4 years) into FSU, the cheaper option. If I can structure a contract in such a way as to allow A the most freedom to attend the school she wants then I will have succeeded.

Here is my solution: Offer the following contract to Agent-Child...
Option1: Principal-Dad pays 20,000 dollars to Agent-Child for not attending college
Option2: Principal-Dad pays 90% of in-state university costs/Agent-Child pays 10% of in-state university costs
Option3: Principal-Dad pays 85% of private university costs/Agent-Child pays 15% of private university costs

This contract uses two economic principals....the first is that everything that is "free" is overconsumed. By making people put skin in the game they will value the good more than if they received it for free or could consume as much as they want. I want the Agent-Child to share in the risk of this adventure but take on the majority of the costs myself and not make college inaccessible. The second is to present to the Agent-Child a significant enough of a price differential so they can reveal their true self through their choice, hence a signal. While some A types might choose the cheaper option because of personal preference the idea is that ALL C types will choose the less expensive option.

Lets go through each hypothetical:
If the Agent-Child is a B type slacker with a short time horizon then they will want instant gratification and will gladly take the 20,000 and burn through it on a trip to Australia, Europe or whatever. Why is this good for the Principal-Dad? Consider the alternative.....if there were no option 1 then the B-type slacker would end up costing me at best 50,000 x 90% = 45,000 dollars and at worst 150,000 x 85%= 127,500 dollars. It will save me at least 25,000 dollars (45,000 - 20,000) to just write a check and be done with it.

On the other hand, if the Agent-Child is a C type slacker who is looking forward to a four year Wed-Sun pub crawl routine she won't be as easily dissuaded by cash up front. She intends to go to college for a long-term good time so I need to present her with two prices to signal that there is a big difference in cost, to me and to her. Hopefully, she will either see that her share of private school is higher than public school (15% vice 10%) or if she can do the math and see that the absolute cost is also higher (150,000 x 15%=22,500 vice 50,000 x 10%=5,000). I'm praying that having to take on 22,500 dollars of loans just to party will persuade her to take summer jobs to pay 5,000 and party at an in-state school. The benefit to me.....simple. Persuading a type-C slacker to go to an in-state school will save me (150,000 x 85%= 127,500) - (50,000 x 90%= 45,000) = 82,500.

The A type studious Agent-Child who cares deeply about getting something out of the educational experience of college still has a bargain either way. Their education is heavily subsidized by the Principal-Dad and the wage differential between college graduate and high school graduate should easily compensate for the large 22,500 dollars of debt during their working life if they go to Wesleyan or 5,000 could be paid off rapidly either through summer work or shortly after graduation if the FSU option is exercised. If the Principal-Dad feels guilty after the fact for loading so much debt onto their child he could, once the diploma is in hand and transcripts scrutinized, decide to ex-post raise the "scholarship" to 100% and reward a good effort.

Beware...unless the Agent-Child is an only child the younger and older offspring will demand "equal" treatment and loan forbearance a la Occupy Wall Street. The trick only works once and then all credibility is lost. If your 30 something adult child moves back in to the house and refuses to leave or take showers because they also want Dad to pay off their loans then all this will have been a waste of time.

This is of course an over-simplification. There are A, B and C types out in the world but the vast middle is a mix of all three. Many students are very happy to trade off a little schoolwork for a Th night bender but not to the point where they would fail or get an unacceptably low grade. I think of this problem often because I went to an expensive liberal arts university and witnessed the unpredictability of my peer's outcomes. There were many kids from well-off families who were exceptionally driven and who sucked the marrow out of the educational offerings at Chicago. There were other kids from top .01% families who couldn't be bothered to lift a finger except to reach for their bong. I noticed the same diversity of behavior from lower and middle class students as well.....lots of bright stars and underachievers. The only difference between what was going on at the University of Chicago and the University of Illinois was the cost. Time is a lot cheaper at the later than the former institution. It made me think hard about how to create the right incentives for my children with regards to their education and adult futures. I'll report back in 14.5 years to see if any of this worked. 

No comments:

Post a Comment