In  the No Child Left Behind Act (2001), the phrase “scientifically based research” is mentioned 110 times. Not a typo. “Evidence-based practice” and “data-driven decision-making” are variations of the NCLB phrase, popular among educational policymakers, administrators, and researchers. What is common to all of these phrases is the idea that systematic inquiry into a question or problem–either through evaluation or research (or both)–will yield solid data useful to educators in making and implementing policy. Especially so, if the evidence is of student outcomes such as test scores, graduation rates, numbers of dropouts, and percentage of graduates ttending college.

Yet the historical record is rich in evidence that student outcome data have played not a primary but subordinate role in making educational policy. Often, policy choices were (and are) political decisions. This is not a criticism of politics or even ideology in schooling but a recognition that tax-supported public schools are political institutions where stakeholders with competing values vie for resources and influence in making policy and putting it into practice.

There were no research or evaluation studies, for example, that found establishing public schools in the early 19th century was better than educating youth through private academies. No studies persuaded late-19th century district superintendents to import privately-funded kindergartens into public schools. Ditto for introducing desktop computers into schools a century later. Evidence including student outcomes were almost besides the point.

So it is hardly surprising, then, that many others, including myself, have been skeptical of the popular idea that policymakers and teachers should pursue unrelentingly evidence-based policy-making and data-driven instruction. The strong belief persists among educators that when policy and practice are anchored in scientifically researched findings and outcome data, then and only then, rational and effective policy-making and classroom teaching can occur.

Yet that belief is unfounded. Too often, neither research nor evaluation data, much less, test scores guide major policy decisions. Political and practical reasons do.

Consider the case of classroom technology. Politically smart state and local policymakers believe–here is where ideology enters the picture–that buying new tablets loaded with software, deploying them to K-12 classrooms, and watching how the devices engage both teachers and students will produce desirable outcomes; it is considered “best practice” because, well, “we believe in it.” The theory is that student engagement with devices and their software will dramatically alter classroom instruction and lead to improved achievement. The problem, of course (you no doubt have guessed where I am going with this) — is that evidence of this electronic innovation transforming teaching practices and raising test scores is not only sparse but also unpersuasive even when some studies show a small “effect size.”

When the research pantry is nearly empty and evidence for raising student test scores or transforming teaching is sparse, how do  policymakers and administrators justify buying new devices and software?

Here are three reasons that I think explain why decision-makers allocate scarce budget dollars for new technologies.

First, keeping up with the rest of the changing world. Call it “modernization” or recasting schools as less like museums and more like fast-paced companies using technology in daily work. No more jokes about educators being technological slow-pokes. Use of new technologies is considered modern, being with-it, even an unadulterated “good” that all children and youth in age-graded schools should embrace.

Second, because new technologies are highly valued in the culture, school boards and their superintendents feel strong pressures to keep up with usage in other sectors–both public and private–undergoing technological changes. If those leaders do not act, they fear that taxpayers and voters will lose confidence in public schools. And public confidence is like money in the bank since tax-supported public schools are politically and fiscally dependent on taxpayers’ good will.

And there is a less obvious third reason for school leaders to purchase new technologies: increase efficiency in students taking tests and scoring the results. Schools have to have computers because most U.S. students take tests online. The Los Angeles Unified School District’s fiasco with iPads a decade ago was triggered by demands to implement required standardized testing from adoption of the Common Core standards. Just as the move from quill pens to pencils to testing-by-computer required no research studies but were done on grounds of cost-saving efficiency, so it was when the LAUSD School Board and Superintendent authorized buying iPads.

Note that the three reasons I offer are political–not in any negative sense–but ones that are practical and realistic in the world that policymakers and practitioners inhabit. Research findings and student outcome data to support the promises that school leaders make for the “good” that high-tech purchases will achieve, are simply not there. And that pattern of pursuing innovations without much evidence or data to support those decisions that school boards and superintendents make is now plain to see.