You should not hold someone to a lower standard because they are an important voice for something you believe. This is Joe Stiglitz reviewing Naomi Klein and not holding her to a reasonable standard of truth because he agrees with her ideological cause. It’s important to call out people defending things you believe when they are loose with the facts, otherwise they can create a caricature of good ideas in the public mind, and become a strawman for opponents to distract debate with.

This is my problem with John Stossel. Yes, he’s often an eloquent voice for libertarianism, and he promotes those ideas to a broad audience. But while his perspective may differ from the typical TV newsman, his gross oversimplification of complex issues, unfortunately, often does not.

Take a recent piece of his from Reason, where he defends school choice. Like a lot of what Stossel says and does it’s peppered with statements that are distracting oversimplifications:

So when will we permit competition and choice, which works great with everything else?

Seriously, John? Everything else? You couldn’t have said “almost everything” or “many other things”? To a libertarian predisposed to believe in competition and choice, your minds eye will breeze over this sentence without distraction. But to a progressive, these sorts of gross generalizations about the limitations of competition and choice are exactly the extreme form of libertarianism they despise; to them, reading a statement like that is a distracting and off-putting jolt that detracts from the credibility of the rest of the article. To understand how this sentence feels to a progressive, imagine reading an article by a progressive who writes that “governments can always fix market failures and make everyone better off”. At that statement, they’ve lost you for good.

It’s because of glib statements like this that other libertarians have to constantly assure people that they don’t want the police to be privatized, and that they do believe in public goods, and there are limitations to what free markets can achieve.

Another example is his discussion of the Head Start program. He wants the reader to believe that Head Start has been proven to be a failure. I’m no expert on this, but the evidence is certainly more mixed than he portrays. For starters, the study he cites uses a single cohort of students from 2002-2003. With this one year sample Stossel insinuates that the 45-year-old program in it’s entirety has been proven ineffective.

The study Stossel cites criticizes Head Start because it’s impact fades quickly:

The study showed that at the end of one program year, access to Head Start positively influenced children’s school readiness. When measured again at the end of kindergarten and first grade, however, the Head Start children and the control group children were at the same level on many of the measures studied.

Of course, this type of fade-out is understood to be common amongst educational interventions, and also ignores potential longer-term benefits. Stossel’s clear-cut, absolute rebuke contrasts with this recent paper on Head Start by David Deming in the journal Applied Economics:

…some studies find evidence of fade-out for African American participants compared to their more advantaged white peers… if fade-out generalizes to all long-term impacts, the benefits of many of these interventions have been overstated. However, studies of model preschool interventions find dramatic improvements in long-term outcomes among program participants, despite rapid fade-out of test score gains.

In addition to the positive results of Deming’s study, his summation of the literature on long-term gains suggests that they are real:

The best evidence for the long-term impact of Head Start comes from two recent studies…Using different data sources and identification strategies, each finds long-term impacts of Head Start on outcomes such as educational attainment, crime, and mortality…

Now, the quick fade out of short-term gains is an important point. If we are going to spend more money on the program we demand to know if and how they will improve it to prevent the short-run gains from the intervention from being lost. This isn’t what Stossel does though. He instead uses these results to declare that Head Start has been proven a failure over it’s 45 year life. How is someone expected to believe that the rest of his article makes accurate claims?