Education Reform - the Fix is In

Note: this is a cross-post from The Realignment Project. Follow us on Facebook!

Introduction:

In any other area of public policy where we see rich hedge-funders providing government agencies with private funds with strings attached, demanding control over choice of administrators and direction of public policy (such as was the case with various foundations and the D.C schools), and financing the opponents of elected officials who disagree with them (such as was the case with the 2010 city council race in New York City), we'd call it corruption by a special interest group.

Yet with education reform, prominent liberals treat it as acceptable and instead turn the traditional liberal analysis of "special interests" on teachers unions instead.

Putting "Data-Based" Reform to the Test:

The power of "special interest groups" tends to be subsumed within debates over campaign finance reform or lobbying, but the greatest power an interest group can have is the power to shape conventional wisdom. While the education "reform" groups have used traditional "special interest group" tactics, most of their power comes from their ability to define what counts as reform.

In this case, they've defined reform as including: high-stakes standardized testing, "accountability" (including the elimination of low-scoring teachers and schools and merit pay), school choice, charter schools/deregulation, and mayoral control. These policy levers are supposed to lead to college attendance and the end of poverty. All other efforts, and any critics (especially teachers' unions) are defined as anti-reform.

Through their wealth, education "reformers" have directed enormous amounts of research in their favored direction, looking for ever-more sophisticated testing measures, but especially looking for confirmation that their agenda is the one true path to improving education. The problem is that the results don't just not add up, they flatly contradict their entire agenda:

High-Stakes Testing Doesn't Work:

Given the manifest problems with current testing (states cheat to get around No Child Left Behind, schools cheat to avoid being closed down, not all subjects are tested, etc.), education "reformers" have hung their hat on value-added testing to restore their preferred model of "accountability." The L.A Times (in)famously ran a searchable database of value-added scores in the LA Unified School District to "shame" administrators and teachers unions. The problem is that even value-added tests don't work and provide wildly inconsistent results. As a study by the EPI found, "across five large urban districts, among teachers who were ranked in the top 20% of effectiveness in the first year, fewer than a third were in that top group the next year, and another third moved all the way down to the bottom 40%."

Underlying all of these difficulties are two basic problems with "data-based decision-making." The first is that human interactions are incredibly complex, especially in education, which means that separating out the effects of teachers from all other potential factors is a difficult process that can't just be adjusted away; a certain amount of imprecision is acceptable as part of an academic enterprise, but not when people's livelihoods are at stake. The second is that measuring in the way that education "reformers" want to use it invokes Campbell's Law; the high stakes of testing creates a systematic incentive to cheat and avoid negative outcomes. In countries that score well on international comparisons, testing is used solely as a diagnostic to evaluate teaching strategies and course materials, not whether someone should be fired or a school should be closed down.

Merit Pay Doesn't Work:

Besides the stick of firing teachers who perform poorly, education "reformers" have put great hopes in merit pay, both as a means of recruiting superior teachers into the profession or enticing superior performing from existing teachers. New York City, under Mayor Bloomberg and Schools Chancellor Joel Klein, put $75 million into a system of merit pay bonuses, only to find that thescheme showed negligible impact on any indicator; similar results were found in a Nashville experiment.

The problem here is that education "reformers" are working from a narrow neoclassical conception of human motivations - pay people more and they work harder, because human beings are rational, self-interested hedonists in  Econ 101. In reality, things are more complex. As the Federal Reserve found out when they studied the empirical results of merit-based pay systems, this kind of incentive only works with simple tasks and in cognitive tasks, has a negative impact on performance

Teaching is an incredibly complex cognitive task, equal parts theatrical performance, anthropological observation, and academic recall. When teachers are distracted by worrying about whether they're going to get a bonus (or lose their job), they can't settle into the moment and devote themselves to teaching. If anything, you'd want pay increases to be so universal and automatic that people didn't have to give them a second thought.

Charter Schools Are a Net Negative:

Stanford University's 2009 study of charter schools found that "17 percent of charter schools reported academic gains that were significantly better than traditional public schools, while 37 percent of charter schools showed gains that were worse than their traditional public school counterparts, with 46 percent of charter schools demonstrating no significant difference." In other words, all other things being equal, if we were to replace every public school with a charter school, academic gains would decline by as much as 20%.

This really shouldn't come as a surprise to anyone who didn't assume that education regulations and union contracts were the culprits for poor educational outcomes in poor districts. That argument never made much sense - all schools have regulations and the majority of schools have unionized workforces, yet outcomes differ greatly from school district to school district.

Most Turnaround Claims Don't Pan Out:

Given all this, it's not a surprise that most turnaround claims peddled by education "reformers" don't actually hold water once they're more thoroughly investigated. The supposed gains that Michelle Rhee claimed on her resume didn't pan out, and likewise testing results out of D.C that are the basis for her current celebrity have been tarnished by claimsof systemicfraud. Likewise, the improvement in New York City test scores that Michael Bloomberg and Joel Klein built their reputations on evaporated when standards were re-calibrated to counter score inflation.

On a basic level, this is an embarrassment for the education "reform" community. If we leave gloating aside, however, there is something we can learn from these scandals and the cheating scandals listed above: looking for short-term "silver bullets" and the Lone Rangers who supposedly have them chambered doesn't make sense when you're dealing with social institutions that are so profoundly shaped by forces of race, class, poverty, public policy on all three levels of government, and geography. As in life, regular sustainable improvements are both more achievable or sustainable then moon-shots.

You'd think, if public policy actually worked in the rational, quasi-scientific way that some progressives wish it did, education "reformers" would be completely discredited both on empirical and credibility grounds. The fact that they aren't, that the "reform" agenda of high-stakes standardized testing, "accountability", school choice, charter schools/deregulation, and mayoral control still drive national, state, and local policy, should suggest an alternative: education "reformers" are a special interest, who've mounted a successful capture of education policy.

An Alternative Agenda:

As I've argued before, as someone who mostly studies employment policy, one of my biggest problem with the education "reform" crowd is their tunnel vision when it comes to education and poverty. Focusing solely on college preparation - in other words, solely on labor supply - is a bad strategy either for workforce preparation, employment, or anti-poverty when you consider the manifest problems of the American labor market.

To begin with, there's the question of whether there's labor demand for the college-educated sufficient to cover large numbers of additional graduates. The evidence doesn't look that good. While those with PhDs, professional degrees, and master's degrees are genuinely doing well, those with bachelor's have an unemployment rate 50% higher, and the 25% of graduates with associate degrees have a 7% unemployment rate and below-average wages (college dropouts have even higher unemployment rates and lower wages). These stats are rosy by comparison to the situation facing recent college graduates, who have less work experience to bolster their resumes:22.4% of them are unemployed.

The situation doesn't look that much better when we consider the question of wages. If we look over time, we can see that college wages have not been immune to the wage stagnation we've seen across the labor market:

 

Uploaded with ImageShack.us While advanced degrees did well in the 90s, and bachelors more modestly well, both have shown either stagnation or gradual declines, not merely in the last recession but in the last thirteen years. College degrees are no longer a shield from adysfunctional labor market. 28% of working recent graduates are working at jobs that don't require college degrees and earning a median salary of less than $16,000 a year; their luckier peers who have jobs that do require college degrees don't fair much better, with a median salary of less than $27,000 a year.

None of this suggests that an education-only strategy will work. Obviously, we still need to focus on education, but we need to focus on boosting the labor market, and onreducing inequality, and lowering the poverty rate as well to ensure that people can earn a living wage whether they've attended college or not. All these things will also help our education system - employed, economically secure parents can provide better for their children and are more likely to have time to provide them help with school, both things that will help their kids do better in school.

At the same time, there's a lot that could be done in the realm of education policy that isn't anywhere on the agenda of the "education reform" movement: de-segregating American schools, especially on class lines, and severing the link between housing prices and school quality; making higher education more affordable; providing top-notch vocational education so that people have more choices than academia or poverty; adopting the Finnish model of teacher recruitment and training; and so on. It's not the case that the critics of the education "reform" community make the case implicitly or explicitly that we should spend less money on education - there are other models of education reform out there.

Conclusion:

The biggest problem we have in making progress on education reform really is the repeated bad faith of education "reformers" and the climate of suspicion and mistrust that's created. Given how they've behaved in the past, given that we see Michelle Rhee collaborating with Tea Party Republicans to strip teachers of their collective bargaining rights, there is no trust between teachers and reformers, and parents are badly split.

In other words, if some sort of grand bargain on education is to be made, "reformers" need to reform themselves.