Sunday, September 29, 2013

Conservatives want to judge but not act, Liberals want to act but not judge.

This is the second time in ten days where I have come across a researcher whose work I admire and respect, making a claim that is on the face of it either nonsensical or just flat out wrong. It is one of those cases where you have such confidence in their work that you are sure the error must be in your reading and yet the words seem to be clear.

The current case is the article The Parenting Gap by Richard V. Reeves, Isabel Sawhill, and Kimberly Howard, all from the Brookings Institution, and all reputable researchers.

I broadly agree with their thesis, that the role of parents in influencing life outcomes is woefully underestimated and that schools cannot deliver what is expected of them until parents deliver what is needed from them. The crux of inequality lies in the home and the challenge is to determine what can be done and what should be done. It is much easier to beat up schools about results than grasp the thorns of parental performance.

There are some great lines in the article.
Currently, however, parenting policies are the Cinderella of early childhood initiatives, eclipsed by the focus on pre-K education. In part, this is because interventions in parenting are politically unpalatable. Conservatives are comfortable with the notion that parents and families matter, but too often simply blame the parents for whatever goes wrong. They resist the notion that government has a role in promoting good parenting. Judging is fine. Acting is not. Liberals have exactly the opposite problem. They have no qualms about deploying expensive public policies, but are wary of any suggestion that parents—especially poor and/or black parents—are in some way responsible for the constrained life chances of their children. Many liberals instinctively believe that reducing financial poverty is the only worthy social policy goal—and the principal route to reducing other social problems. Poverty reduction is, in and of itself, a vitally important ambition. But raising the abilities of parents is not just about raising their incomes.

Neither the standard conservative nor liberal position will do. Public education, no matter how lavishly funded, can never substitute for good parents. But it is absurd to cast the idea of taking broader responsibility for helping parents as closet communism, as some on the right do. What is needed is a policy agenda and political platform that recognizes the contribution of parenting to mobility and opportunity, and tackles the parenting gap.
They kind of gloss over it, but I think that is a great observation: Conservatives want to judge but not act, Liberals want to act but not judge. That's a Venn diagram with a lot of white space.

But here's the passage that gives me heartburn. They are discussing Home Instruction for Parents of Preschool Youngsters (HIPPY), a program intended to close the parenting gap.
HIPPY has also shown positive results. In our new paper, we estimate the effects of HIPPY on longer-term outcomes of participants. The goal of the program, offered when children are age three to five, is to effectively train parents to be their child’s first teacher. Families receive biweekly home visits from a paraprofessional for 30 weeks out of the year, along with biweekly group meetings. Parents are also given books and toys. A high-quality evaluation of the program found significant improvements in reading and school readiness in first grade. Using a microsimulation model—the Social Genome Model—we predict that HIPPY participants are 3 percent more likely to graduate high school, and 6 percent less likely to become teen parents. These are modest effects, but positive ones, given the importance of the outcomes. High-school graduates make $260,000 more in their working lives. For a program that costs around $3,500 per participant, it’s close to a gold-plated investment.
Set aside the discomfort of using a model to predict outcomes rather than empirically measuring the outcomes. That's a big stone in the path, particularly given the past inability of other similar programs to deliver predicted results, but let's ignore it. What they describe as "significant improvements" yields a 3% increase in the graduation rate. OK, let's stipulate that a 3% improvement constitutes "significant". To be fair, they do change the description to "modest"; I'd still argue that the better words might be negligible and uncertain.

What is arresting is the description of the financial returns on this project as "a gold-plated investment." I would like HIPPY to be successful as well, but there is no financially literate way to describe this as "a gold-plated investment."

Let's assume there are 100 participants in the program. We also have to make assumptions about the graduation rate. The national average is somewhere around 70%. Most at risk urban systems have a graduation rate around 50-60%. I'll assume 60%. 3% more likely to graduate means, I assume, that the graduation rate goes from 60% to 63%.

So what does this mean financially. Taxpayers spent $350,000 (100 X $3,500 per student) in order to improve the graduation rate by 3% or by three individuals. The three individuals are likely to earn $260,000 more each as HS graduates than without a degree. So there is a gross increase in productivity of $780,000.

Who are the winners and losers?

Well, the three individuals are clear winners. They invested nothing and each receive $260,000 in additional income over a lifetime that they would not otherwise have had. That's $5,200 a year - not life changing but not chump change either.

What about the taxpayers and the government? $350,000 was removed from the pockets of taxpayers. If they had kept that money and socked it away in a simple bond paying 5% for fifty years (the working life of a high school graduate, they would at the end of that period have $4,000,000 in the bank instead of the extra $780,000 that the three graduates produced. So from a GDP perspective, we would have been five times better off leaving the money with the taxpayers than investing in this program.

Forget the taxpayers, let's look at it from the Government's perspective. How did they make out? Well, the government spent $350,000 up front. Did the government clear enough extra tax revenues to cover the cost of the program? Let's assume that the HS graduates paid 30% in taxes on the extra $780,000 income they earned (a very generous assumption not taking into account NPV, etc.). So the government spent $350,000 in order to generate $234,000 in extra tax revenue (30% X $780,000). In other words, the government lost $116,000 on this investment.

Is any of this amounting to a "gold-plated investment". I have heard of bad financial planners but don't let Reeves, Sawhill and Howard anywhere near my retirement savings. Bernie Madoffs one and all, it would appear.

The most cynical description of these results would be that a lot of money was spent to achieve very little, and that that little was concentrated among a tiny percentage of the program participants. How do you make a small fortune in government programs intended to make people better off? You start with a big fortune.

If the government had invested $70,000 in a 5% bond and simply given the money to three randomly selected HS graduates, everyone would have been better off from a financial perspective. Taxpayers would have saved $280,000 which would have grown to $3.2 million. Government would have been better off by taking in more tax revenues than the cost of the original bond. The three individuals would have been financially equal (which does not take into account the reduced value that arises from achievement).

There's only one way that I am able to make sense of this assertion on the part of the authors. I think what they might be saying is that for every participant, it costs $3,500 to be enrolled and there is a 3% chance that they will achieve an extra $260,000 in lifetime earnings by graduating. If that is the case then the numbers look like this.

Participant - No cost to be involved but an increase of 3% in graduation rate and the attendant extra income of a graduate of lifetime earnings of $260,000. 3% of that is $7,800 over a lifetime, or 156 dollars a year. Cost of investment zero, return of $7,800, then that is definitely a good return. But only for the participant who makes no investment.

Taxpayer - Lost income of $3,500 which could otherwise have been invested and yielded $40,100. Definitely a losing proposition.

Government - Investment of $3,500 of taxpayer funds. Extra future tax revenues is $2,300 (30% X $7,800) or a program loss of $1,200 per participant. The government would have been better off simply investing $650 at the beginning and paying out the interest proceeds.

Anyway you look at, as far as I can tell, this project has a terrible, negative ROI. In any sort of normal accounting where you take into account compound interest, time value of money, etc. there is just no way that this a responsible investment of scarce societal resources. What am I missing?

Why do the authors claim it is a gold plated investment? I don't know. A gold plated investment is one that returns a net increase in revenues over the cost of the investment while taking into account the time value of money and the risk adjusted value of that return and alternate investments. Perhaps it is tunnel vision. They focused exclusively on the participant and never took into account the costs to the taxpayer and the government. Its the only thing I can come up with.


No comments:

Post a Comment