In the past two years, raising the quality of the nation’s teaching force has become a priority that cuts across partisan lines. Driven by news of shortages in certain subjects (such as math, science, and special education) and in rural and inner-city schools, state legislatures have earmarked billions of dollars for salary increases and teacher training. Governors as diverse as George Pataki of New York, Roy Barnes of Georgia, and Gray Davis of California have led the push to strengthen certification requirements, to design innovative recruitment incentives, and even, at least in Barnes’ case, to loosen tenure rules. The movement to “professionalize” teaching has gained momentum as well. Many states now offer bonuses of $5,000–$10,000 to teachers who go through the National Board for Professional Teaching Standards’ board certification process, and the American Federation of Teachers (AFT) made a splash by coming out in support of aptitude testing for new teachers.

ILLUSTRATION BY DAN VASCONCELLOS.

What is strange about the debate over teacher policy, however, is that it has taken place largely within the confines of the existing public school system. Hardly anyone talks about how the growing movement toward parental choice and competition, in the form of vouchers and charter schools, will affect the teaching profession. Only the teachers unions seem to hold a clear, if unspoken, opinion on this question. Though they never say so, the unions obviously think that school choice would harm teachers. Both the National Education Association (NEA) and the AFT stand resolutely opposed to vouchers, and their attitudes toward charter schools have ranged from tepid support to outright hostility. While these unions don’t craft their agendas solely with the interests of teachers in mind, they seldom—if ever—oppose policies that would benefit their members.

Yet economic theory suggests that school choice would change the teaching profession in ways that would fulfill many of the reform movement’s goals. In short, theory predicts that schools that faced stronger competition would favor teachers who raised the schools’ ability to attract students. These schools presumably would strive to attract and retain teachers who were especially talented or hard-working or who possessed rare skills. In turn, you would expect their tolerance for less-effective teachers to wane. You would expect, in fact, that teaching would be transformed into a true profession, where workers are rewarded not only on the basis of seniority but also on the basis of their skills and performance.

Making a Profession a Profession

In the current system, teachers unions tend to compress wages within a school district so that teachers with the same seniority and the same degree are likely to receive similar (if not identical) salaries. Even nonunionized districts tend to adopt salary scales that resemble union scales. So teachers who excel at their jobs or teach hard-to-staff subjects, such as math and science, are paid the same as if they were mediocre or could be replaced easily.

This is one reason why teaching, as it stands, is not as attractive to candidates with high aptitude, a strong work ethic, or math and science skills. The evidence suggests that such people are less likely to start teaching than their lower-aptitude peers (even when those being compared are all certified to teach) and less likely to remain in teaching. This situation has worsened over time, as other professions such as management, law, and medicine have opened their doors to women. Women with high aptitude or math and science skills have chosen such professions over teaching, perhaps because these professions do reward workers based on their skills and performance. By making teaching a more market-oriented profession, school choice might also make it a more attractive career to people who would thrive in such an environment.

Why might teachers welcome such a shift? In a word, professionalism. Both the NEA and the AFT support so-called professionalization agendas that aim to make the route into teaching look more like the routes into the legal and medical professions. Much as lawyers and doctors ensure quality within their professions through bar examinations and board certification, the AFT and the NEA call for higher certification standards for new teachers (but rarely for current teachers), apprenticeships, peer review, and rewards for teachers who earn additional credentials, such as National Board certification. The movement also includes proposals to give teachers more control over the profession through independent, state-level standard-setting boards and to give the National Council for the Accreditation of Teacher Education, the body that accredits education schools, more control over the number and origin of undergraduate degrees granted in education each year.

In most professions, however, credentials such as board certification merely maintain minimum standards. The size of a professional’s client list or how much a professional earns usually depends on how the market values her services, since professionals typically compete in a market setting as employees of a competitive firm, in self-employment, or in partnership. The teachers unions, however, want to use credentials to restrict the supply of teachers while opposing competition, making their range of policies more characteristic of a craft union than of a profession. Professionalization in this sense is a misnomer for the current movement. Yet the name and rhetoric of the movement suggest that many teachers do want teaching to become more professional. Economic theory suggests that school choice would make it so, in the sense that it would raise demand for highly talented and skilled teachers who receive rewards closely linked to their performance.

Just a Theory?

Without evidence, of course, this is just a theory. And without full-fledged school choice, it is a tough theory to test, though not impossible.

Comparisons between public and private school teachers already give us some insight into how choice might affect the teaching profession. After all, the revenues of private schools depend on their ability to attract tuition-paying students. They thus have an added incentive to hire teachers who will help them do so. What do they look for? In a comprehensive comparison of private and public school teachers, economists Dale Ballou of the University of Massachusetts Amherst and Michael Podgursky of the University of Missouri–Columbia found that, in making hiring decisions, private schools value high aptitude more than public schools do. They also found that salaries in private schools correspond to aptitude and scarce skills (such as math and science skills) more than they do in public schools.

Private schools, however, are not a perfect guide to what teachers will experience under a system of school choice. Parents who pay private-school tuition must also continue to pay taxes that support public schools. Private schools thus face financial constraints that a true school of choice would not. As a result, they typically pay teacher salaries that are about 60 percent of local public school salaries.

To get a broader picture of how choice affects teachers, I used data both from traditional forms of school choice (choice among public schools through choice of residence and choice among private schools) and from charter schools. My analytical strategy relied on a simple economic argument: If schools that face stronger competition prefer certain characteristics in their teachers, then they should hire more teachers who possess those characteristics and pay them a higher wage. In economic terms, they would show greater demand for teachers with those characteristics; they would be willing to pay a higher price in order to buy more of them.

My findings suggest that school choice would change the teaching profession in significant ways. Schools that face tougher competition have greater demand for teachers who attended well-regarded colleges, majored in subject areas (especially in math and science), and who put in more effort and show more independence. These schools are more likely to hire such teachers and to pay them higher wages than they would earn in schools that face less competition. In general, they also have less demand for certification and master’s degrees. They pay teachers who hold such credentials less than similarly educated teachers earn in schools that are less choice-driven.

Making Choices

Before I discuss the results of this study, let me briefly explain the measures of choice I used and where some of the data came from.

Within the present K–12 education system, parents and students can make three types of choices. First, they can choose their public schools by choosing where to live (this is called Tiebout choice, after the economist Charles Tiebout, who called attention to its importance). Some metropolitan areas, however, contain more school districts than others, allowing for more choices among public schools. Compare the cities of Miami and Boston, for instance. In Miami, one school district encompasses the entire metropolitan area. Parents have almost no choice among public schools. Boston, by contrast, has 70 districts within a 30-minute commute of downtown and even more in the surrounding metropolitan area. In an area with so many choices, a town’s reputation and real-estate values depend heavily on the quality of its local school district, giving residents added incentive to ensure that their schools compare favorably with those in other towns.

Various factors besides the degree of choice might affect a district’s demand for certain qualities in its teachers. For instance, a school district in an area with high employment in the technology field might show greater demand for teachers who possess skills in math and science. But it also might be a school district that happens to face strong competition. Without controlling for such a coincidence, any comparison would seem to show that competition affected the demand for math and science skills in this district, when, in fact, another factor contributed to the district’s demand as well. Likewise, teachers in an area where the population is more educated tend to have attended more selective colleges. You must adjust the data to take account of this, or they may give you the false impression that competition raises the demand for high-aptitude teachers. Other factors, such as household income, ethnic makeup, and union membership among teachers, can also affect a district’s demand for certain qualities. After controlling for all these factors, I compared teachers in areas where parents have more choices among public schools with teachers in areas where they have fewer.

Second, parents can choose to send their kids to private school. But again, some metropolitan areas have more low-tuition private-school options than others, mainly due to historical variations in religious populations. Private schools in areas with large, long-standing religious populations tend to have higher endowments and more established donor bases. This allows the schools to charge a tuition that is lower than the actual cost of educating a child, thereby making it easier for parents to choose a private school. In some metropolitan areas, up to 15 percent of the elementary school population is enrolled in private schools where the tuition is about two-thirds of the per-pupil expenditure. (Religiously affiliated schools typically charge $1,600 in tuition while spending about $2,300 per student.) Other metropolitan areas have less than 1 percent of their elementary school populations enrolled in such schools. They may have seats available in schools where the tuition is higher and there are no tuition subsidies, but this hinders a parent’s ability to choose a private school. After again controlling for factors other than competition that might affect a private school’s demand for certain teachers, I compared teachers in areas where parents have more choices among private schools with teachers in areas where they have fewer.

Finally, and most recently, parents can choose to send their kids to a charter school. I compared teachers in charter schools with those in private and public schools in the same regions. Here you must control for the fact that the existence of a charter school may reflect factors in an area that also determine the area’s demand for certain teacher characteristics. I also tested for composition effects, meaning that a charter school might enter an area and favor the same teacher characteristics that the public schools want. After the charter entered the area, you might see teachers with that characteristic slowly draining away from the public schools. This is unlikely, given that a charter school has little incentive to move into an area where schools are already doing what it plans to do, but I tested for it anyway. For the comparison among charter, public, and private school teachers, I assumed that charter and private schools face more competition than public schools, since a greater share of charter and private schools get funding only if they attract students.

For both public and private schools, I used data from the Schools and Staffing Survey (SASS), a comprehensive survey of public and private school teachers and administrators. I examined variables such as salary, years of teaching experience, whether teachers planned to stay in teaching for the next few years, the number of hours per week (on top of required hours) that teachers spent on activities related to their students’ academic progress, and the number of hours that they spent on their students’ extracurricular activities. I also looked at where they earned their baccalaureate degrees, whether they were math or science majors, the number of undergraduate and graduate courses they took in math and science, and whether they earned their degrees in a field of the arts and sciences (rather than in education).

The SASS does not cover charter schools, so I had to conduct a survey of my own. I distributed surveys to one administrator and two randomly selected teachers in every charter school that was open as of October 1998. The survey questions were taken from the SASS. Data from the survey will be described in a forthcoming monograph.

Results

Remember that my test for whether school choice raises demand for certain teacher characteristics is two-fold: 1) whether a school that faces stronger competition hires teachers with more of a certain characteristic; and 2) whether that characteristic earns a premium in an environment of greater school choice. Remember also that I performed two different comparisons: 1) a within-sector comparison of teachers in both the public and private sectors; and 2) an across-sector comparison of teachers in public, private, and charter schools.

The first question I asked was whether competition raises demand for teachers who possess high aptitude. This is important to know because research has shown that teachers’ aptitude, as measured by scores on standardized tests, significantly affects student achievement. The SASS, unfortunately, does not have a measure of the aptitude of individual teachers, since only some states require teachers to take proficiency tests. Even if they did, teachers would not have taken comparable tests anyway. The SASS does, however, list which college teachers attended. This is a good measure of the quality of their education and, to some extent, of their aptitude. Colleges take account of students’ aptitude, high-school grades, extracurricular activities, and character in making acceptance decisions, so the selectivity of a teacher’s college summarizes a complex set of information on aptitude, work ethic, leadership, creativity, and so on.

The most obvious way to use a teacher’s college as a proxy for aptitude would be to assign that teacher the college’s average SAT score. But most teachers come from colleges that range from nonselective to somewhat selective, where SAT scores are poorly defined because so many students do not take admissions tests. Their average SAT scores would not be an accurate proxy for an individual teacher’s aptitude. So, instead of SAT scores, I used the rating each college received in the 1982 Barron’s Profiles of American Colleges. I chose the 1982 version because it represented a middle ground: Older teachers would have graduated before 1982, younger teachers after. In any case, the rankings are fairly stable from year to year, so the choice of year doesn’t make much difference. I assigned the teachers from each school a number that corresponds to that school’s Barron’s rating, from 9 for Barron’s top rating, “most competitive,” to 1 for its lowest, “nonselective,” which is reserved for schools that will accept nearly anyone with a high school diploma or GED.

My first finding was not so surprising. Several scholars have shown that teachers are drawn mainly from the bottom half of the college-going population’s aptitude distribution. Likewise, most of the teachers in the SASS sample came from schools rated between 4 (“competitive plus”) and 1 (nonselective).

My findings also suggest, however, that enhanced choice and competition could help to remedy this situation and thereby bolster the skills of the nation’s teaching corps. I found that teachers who work in areas with higher Tiebout and private school choice attended more selective colleges. Teachers in areas of maximal Tiebout choice attended colleges that were ranked 0.4 levels higher than in areas of minimal Tiebout choice. Areas with a high degree of private school choice have teachers who attended colleges that were ranked 0.1 levels higher than areas with minimal private school choice.

Because most teachers attended schools that rated in the 1–4 range, I was also interested in whether competition raised demand for teachers who went to schools rated 4 (competitive plus) or higher. It appears to. For instance, about 20 percent of public school teachers went to such schools, compared with 36 percent of charter school teachers and 36 percent of private school teachers (see Figure 1). In public schools, teachers from such colleges earn 3 percent more than their peers. In charter schools, they are paid 6 percent more, suggesting that charter schools have greater demand for teachers who have graduated from colleges that rate at least “competitive plus.” Their need to attract students in order to receive any funding seems to drive their hiring of higher-aptitude teachers.

Similarly, a teacher in an area with many public school choices (maximal Tiebout choice) is 8 percent more likely than a teacher in an area of minimal Tiebout choice to have attended a college that rates at least “competitive plus.” Teachers who attended such colleges also earn 8 percent more than their peers. A high degree of private school choice boosts the likelihood that a teacher attended a college rated at least “competitive plus” by 5 percent. The degree of private school choice does not have a meaningful effect on the wage for this characteristic.

Skills in Short Supply

My second question was motivated by the evidence that schools face a shortage of teachers who have math and science skills (only 7 percent of teachers in the United States were math or science majors). Indeed, schools need teachers who have baccalaureate degrees in any field of the arts and sciences (as opposed to education) because they have subject knowledge that is useful in grades 7 through 12.

Again, my findings suggest that enhanced competition and choice may help to ameliorate these shortages. Stronger competitive pressures appear to raise the demand for teachers who have math and science skills. In an area of maximal Tiebout choice, a teacher is 15 percent more likely to have majored in math or science than a teacher in an area of minimal Tiebout choice. A teacher in an area with a high degree of private school choice is 10 percent more likely to have majored in math or science than a teacher in an area with minimal private school choice. Teachers who majored in math or science earn 16 percent more if they work in an area with maximal Tiebout choice than they would if they worked in an area with minimal Tiebout choice. In private schools, math or science majors earn 14 percent more if they work in an area with a high degree of private-school choice.

In charter schools, 10 percent of teachers are math and science majors, versus 8 percent of public school teachers and 7 percent of private school teachers. In the public sector, math and science majors earn about 4 percent more; in charters, 8 percent.

The results also suggest that schools facing more competition are better at retaining such teachers. In general, math and science teachers are more likely to leave teaching.They have 1.2 fewer years of experience and are 6 percent less likely to say that they plan to keep teaching than other teachers. But in an area of maximal Tiebout choice and a high degree of private school choice, math and science teachers have only .4 fewer years of experience and are no less likely to say that they plan to continue teaching. This may be a result of such teachers’ being paid more in districts that face more choice-based incentives.

Schools that face stronger competition also appear to favor teachers who majored in subject areas. Of teachers in charter schools, 56 percent majored in a field of the arts and sciences (as opposed to education), compared to 37 percent of public school teachers and 42 percent of private school teachers (see Figure 1).

Effort and Independence

I also wanted to know whether school choice increased demand for teachers who put in more effort and show more independence. I used data that measure the number of hours teachers spend on instructional tasks (such as tutoring) and on noninstructional tasks (such as directing a school play) beyond the hours required in their contracts. The measure of independence is a teacher’s assessment of the control that she exercises over her teaching (particularly teaching methods and the organization of material).

In general, schools that face more Tiebout choice demand more effort and independence from their teachers, while the effect of private school choice is unclear. Teachers in an area with maximal Tiebout choice spend two extra hours per week on instructional work and one extra hour per week on noninstructional work than teachers in an area of minimal Tiebout choice. A high degree of Tiebout choice also raises the wage paid for every extra instructional hour per week by .7 percent and for every extra noninstructional hour per week by .3 percent.

Charter school teachers work 13 extra instructional hours per week, versus 9 extra hours for public school teachers and 9 extra hours for private school teachers. The difference between the charter and public school teachers is highly statistically significant. In public schools, teachers are not compensated for the extra instructional hours they work; in charter schools, they are paid 5 percent more for an extra 13 hours of work. In public schools, teachers claim their level of control over their teaching is 4.8 on a 6-point scale, while charter and private school teachers rate their level of control at 5.6 on a 6-point scale.

The professionalization movement among teachers includes a sharp focus on certification and on the importance of holding a master’s degree, so I was interested in how choice affects schools’ demand for such credentials. It is important to note that many teachers-union contracts specify salary increases for master’s degrees and for being certified. So examining the data shows whether choice motivates schools to make such salary increases larger than they would be in the absence of choice.

Overall, about 44 percent of public school teachers, 41 percent of charter school teachers, and 28 percent of private school teachers have master’s degrees (see Figure 2). The public sector pays teachers who hold master’s degrees about 25 percent more; in charter schools, they are paid about 20 percent more. Furthermore, nearly all public school teachers are certified, while only 87 percent of charter school and 65 percent of private school teachers are. In private schools, teachers who hold certification are paid 1.4 percent less than uncertified teachers; charter schools give a similarly negative premium, but this finding is not statistically significant. This suggests that private schools may slightly prefer uncertified teachers and that charter schools probably have less demand for them than public schools do.

Another way to look at this question is whether teachers actually teach in the areas for which they are certified. Most states put few restrictions on teachers—once certified—taking on assignments that are relatively remote from the fields in which they were certified. About 97 percent of public school teachers claim to be certified in their teaching area, while only 83 percent of charter school and 54 percent of private school teachers do (see Figure 2). Public school teachers who teach in their areas of certification earn a substantial wage premium, 9 percent, compared with a premium that is not meaningfully different from zero for charter teachers and a 2 percent premium for private school teachers. All in all, it appears that public schools have a greater demand for degrees and certification than private or charter schools.

Broadly speaking, my findings suggest that enhanced competition and choice raise the demand for high aptitude, skills in math and science, subject-area expertise, effort, and perhaps independence among teachers. Choice also seems to lower schools’ demand for certification and master’s degrees. These findings further suggest that school choice has the potential to create a professional environment for teachers in which more motivated and skilled teachers earn higher pay for such qualities. For new teachers, this could quickly change the profession in areas where a growing number of choice schools offer a large share of the new teaching positions. It would take longer for veteran teachers to feel the pinch. They would notice a change only when their schools began to feel competitive pressures and, as a result, began to demand teachers whose characteristics attract parents. Some teachers, obviously, would dislike such changes. Less able or less motivated incumbent teachers might find themselves earning smaller salary increases than some of their peers. Such teachers might be more likely to leave the teaching profession. That would be a welcome reversal of the current pattern, where the most able teachers are also the most likely to exit early.

–Caroline M. Hoxby is an associate professor of economics at Harvard University and a visiting fellow at the Hoover Institution, Stanford University.