Teacher Unions and School Board Elections. Terry M. Moe Hoover Institution Stanford University PEPG 03-05

Transcription

1 Teacher Unions and School Board Elections Terry M. Moe Hoover Institution Stanford University PEPG Preliminary draft Please do not cite without permission Prepared for the conference, School Board Politics Kennedy School of Government, Harvard University, October 15-17, 2003

2 In the folklore of American education, school boards are bastions of local democracy responsive to local citizens, representative of community interests, concerned about children and the public welfare. It is not clear, however, that school boards have ever measured up to this ideal. During the early years of the twentieth century, they were often under the thumb of party machines. And even later, as Progressive reforms weakened the parties and attempted to promote democracy, governance of the schools shifted to newly powerful groups business, middle class activists, education administrators whose interests were not necessarily congruent with those of their communities either (Peterson, 1985; Tyack, 1974). Scholars have so rarely studied school board politics that we may never have a good sense of what happened in those days. But one thing is clear. During the 1960s and 1970s, the balance of power within American education underwent a dramatic shift again generating a modern politics of education that does pose a serious challenge for democratic governance. This transformation was brought about by the unionization of teachers. Before the 1960s, states did not have collective bargaining laws for public employees. Few teachers belonged to unions or engaged in collective bargaining, and teachers had little political power. All this changed during the next two decades, as fierce battles over collective bargaining rights were fought throughout the country, and the teachers and their unions emerged triumphant. By the early 1980s, virtually all school districts of any size (outside parts of the South) were unionized, and the two major teacher unions the National Education Association (NEA) and the American Federation of Teachers (AFT) had become the de facto political leaders of the education establishment (Moe, 2001). The teacher unions have all the requisites of political power. They are supremely well organized at the local, state, and national levels. They have a guaranteed, continuous source of 1

3 money (member dues), which they can convert into campaign contributions, lobbying, and political advertising. And most important, they have millions of members spread almost uniformly across the country, along with armies of activists in virtually every political district who are willing to make phone calls, ring door bells, distribute literature, and do what it takes to elect their friends and defeat their enemies (Lieberman, 1997). No other political group in the country can claim such a formidable combination of weapons, whether in education or in any other field of public policy. Indeed, in a recent study of the political systems of all 50 states, the teacher unions emerged as the single most powerful interest group in the nation, outdistancing bankers, insurance companies, trial lawyers, farm groups, and all others (Thomas and Hrebnar, 1999). From a democratic standpoint, the rise of the teacher unions is troubling because it threatens to give one special interest group disproportionate influence in the politics of education, while giving short shrift to other social groups particularly broadly based ones, like parents and taxpayers. It is also troubling because teachers are not just any social group. They are employees of the education system. And as such they have vested interests in job security, higher wages and fringe benefits, cushy pensions, restrictive job rules, bigger budgets, higher taxes, lax accountability, limits on parental choice, and lots of other things that may often conflict with what is good for children, schools, and the public interest. This power of the agent is not unique to education. It is a generic problem that is inherent to democracy. In any democratic system, there is a hierarchy of authority in which elected officials make policies on behalf of their constituents and then hire agents to see that those policies are carried out. But precisely because the authorities are elected, the agents whose vested interests give them strong incentives to get involved in politics can organize for 2

4 political action and attempt to play a powerful role in electing the very authorities who will serve as their democratic superiors, and who are supposed to represent the public. To the extent that the agents are successful, the rightful hierarchy of democratic authority is essentially turned on its head, with the authorities doing the bidding of the agents rather than the other way around and government run for the benefit of vested interests rather than society at large. The teacher unions are the most notable example of public bureaucrats getting organized for the exercise of political power. But they are just the tip of a very big iceberg. Public sector unions now organize virtually every sector of governmental activity not just in the United States, but in most every developed democracy and they are extremely active in politics (Moe, 2003). How much clout do they really have? We don t know for sure. Thus far, public sector unions have rarely been studied as political actors. And the teacher unions are no exception. In the more popular literature, which is essentially at the fringe of social science, there are a few books and articles that provide insight into what the teacher unions do in politics and with what effects (Lieberman, 1997; Brimelow, 2003). But there is very little in the way of systematic research on the unions, and for that matter very little on the politics of education as a whole. One of my purposes here is simply to provide some basic facts on the role of teacher unions in school board elections. In a recent paper, I used data on electoral outcomes to show that the teacher unions are quite successful at getting their favored candidates elected to local school boards (Moe, 2003). Here, using a data set derived from interviews with candidates, I take a closer look at the political process itself: at what the unions do in elections, what kinds of candidates they support, and what it all means for the composition of school boards. My second purpose is to offer some theoretical arguments rooted in very simple rational-choice logic about what we should expect, and then to see how they square with the 3

5 data. The arguments essentially come down to this: there are good reasons for thinking that teacher unions will ordinarily be the most potent political force in school board elections, but also that their power will be constrained in important, identifiable ways that vary across contexts. Thus, while school board politics should often be tilted in the unions favor a negative for local democracy there are built-in limits on what they are able to achieve. The data, as I will show, largely bear this out. THE SURVEY This study focuses on school board politics in the state of California, where interviews were conducted with 504 school board candidates from more than 250 school districts. The interviews were carried out intermittently over a period of several years, beginning in 2000 and continuing through 2003, with districts and candidates chosen randomly from their larger populations. Why California? The obvious alternative is a national cross-section. But with a focus on elections, we would need information from each district on who the candidates are, who won and lost, who the incumbents are, and the like, and we would need to get phone numbers for candidates in order to make the needed calls. In principle, such a thing is possible on a nationwide basis. But because states and localities often don t publish election information for school boards, the information can only be gathered by dealing with each school district individually (and perhaps the county and state election offices). I leave it to other researchers to do that. In California, statewide information on school board elections is readily available through the League of Women Voters. Which is a big help. And California is an attractive subject of 4

6 study anyway. It is the nation s most populous state, enrolling more than 10% of the nation s school children. And it is geographically large and demographically diverse, yielding a population of school districts that is enormously varied urban and rural, big and small, conservative and liberal, white and minority, you name it. While the California Teachers Association (an affiliate of the NEA) is noted for its political power at the state level, teacher unions in virtually all states are noted for their state-level power too, and there is no reason to think that local politics in California is much different than anywhere else on average. Indeed, if there is one factor that might set California apart, it is Proposition 13, which drastically limits local property taxes and thus limits the financial discretion of local school districts. If anything, this would seem to weaken union incentives to get sympathizers elected to local school boards, suggesting that unions may actually be less active and powerful in California s local politics than elsewhere. On the whole, though, I would expect the basic political patterns that turn up in California to be fairly indicative of what happens nationwide. Now let me provide a few details about the California data. The sample of districts was stratified by enrollment to ensure that roughly equal numbers of small, medium, and large districts would be included in the analysis. While the vast majority of California s school districts are quite small (in terms of enrollment), the vast majority of California s school-age children are enrolled in much larger districts--indeed, 12% are enrolled in Los Angeles Unified alone. These larger districts needed to be oversampled in order to provide a solid basis for generalization and comparison, and for understanding politics in the kinds of districts that matter most. As the guiding purpose here was to examine the role of teacher unions in school board elections, the focus was naturally on those districts that actually held elections in recent years. 5

7 When districts failed to hold elections, it was because there was no competition for seats on their school boards. This absence of competition is unusual in the larger districts, but it is not uncommon in the smaller ones. Indeed, some of the latter even have a hard time finding people who are willing to serve. As a result, the smaller districts that made it into our sample may be somewhat more competitive on average than the entire population of smaller districts statewide. This is not a problem, but it is something to keep in mind. Once the districts were selected to be part of the sample, an effort was made to contact and interview candidates who had run in their elections. The object was to gather information on district politics from people who are inside participants and capable of providing knowledgeable accounts. Within any given district, of course, different candidates can be expected to offer at least somewhat different assessments of how things work and what is actually going on. And it is reasonable to worry that the differences could possibly be systematic e.g., that winners may offer different perspectives on power and politics than losers do. To try to reduce potential biases and arrive at more reliable estimates, I made an effort to interview at least two candidates from each district, and to choose one winner and one loser. Because it proved difficult to get accurate, up-do-date phone numbers on these people, and because about a quarter of the elections involved just two candidates vying for one seat, I often (in 43% of the cases) had to settle for one interview per district. The sample as a whole is about evenly split between winners and losers, though, ensuring diversity in the aggregate. EXPECTATIONS The scholarly literature on school board politics does not offer a coherent view of what we should expect or why. But on the whole, its basic thrust seems to be that school board politics is complicated and diverse, with numerous groups competing for power; and that if any 6

8 one group tends to play a predominant role, it is the business community (e.g., Wirt and Kirst, 2001; Wong, 1992). In my view, there are two problems with this line of thinking. One is that it is probably wrong, or at least very misleading. The other is that emphasizing the complexity of politics does not really help us explain anything it is an easy out that substitutes for clear theoretical ideas about what is actually happening. Things look a lot less complex when we understand them. Indeed, it is usually because we have allowed ourselves to think simply and clearly about a few essentials that we gain real insight into what is going on. This said, let me set out some simple ideas about school board politics and teacher unions that I think ought to be central to our thinking. Any effort to understand the teacher unions is wise to begin by recognizing that their survival and well-being as organizations are rooted in collective bargaining. It is through collective bargaining that they get their members, and through their members that they get most of their resources. Their fundamental interests as organizations, then, have to do with gaining collective bargaining rights, and with keeping and increasing their members and resources through successful negotiations with local school boards: negotiations that yield the job security, material benefits, and working conditions that teachers value, and that produce all manner of related policies bigger budgets and smaller class sizes, for example that make unions and teachers better off (Moe, 2001; Lieberman, 1997). To put it mildly, local democracy presents them with attractive opportunities for pursuing and promoting these fundamental interests. The fact that school boards are elected means that, in stark contrast to the private sector, where unions find themselves up against independent groups of managers, the teacher unions can actually participate in choosing or even literally choose the management groups they will be bargaining with. This being so, it goes without saying that 7

9 the teacher unions should ordinarily have strong incentives to take advantage of these opportunities. And strong incentives, therefore, to use the framework of local democracy to gain and exercise political power. But how powerful should we expect them to be? The answer is that they should often be quite powerful indeed, because the political cards are basically stacked in their favor. Unlike parents, taxpayers, and many other broad constituencies, teachers are a well-organized group. With the great help (in most states) of legislation that promotes union organizing and collective bargaining through the force of law, they have already overcome the kinds of collective action problems that tend to prevent many groups from participating effectively in politics (Olson, 1965). The unions also have a guaranteed source of money to finance campaign activities and support candidates. They often (in the larger districts) have paid staff to coordinate political activities. They get political assistance from experienced operatives hired by the state and national unions with which they are affiliated. And they can count on having a certain percentage of members who are willing to be political activists and do the trench-work of campaigning. For these and related reasons, the unions have major organizational advantages in the democratic process that should translate into electoral power. And this is all the more likely given that school board elections tend to be low-interest, low-turnout affairs in which organized action can be especially consequential. Other things being equal, then, we should expect the teacher unions to be powerful political forces in school board elections. They are a constant presence in school districts of any size, and they have so much of what it takes and for reasons I ll discuss below, so much more of it than other groups do that any effort to develop a coherent perspective on district-level politics needs to start with them. They are the ones to beat. 8

10 Nonetheless, this is just a baseline, and other things are not really equal. Conditions vary across districts, and there are clearly conditions under which the unions might have a difficult time exercising power. In some cases, in fact, they might not be very powerful or active at all. In the sections below, I want to extend the perspective I ve begun to develop here by discussing, and then bringing evidence to bear, on four basic types of conditions each of them a potential constraint that need to be taken into account for a better-rounded understanding of district politics and union power. They are: the size of the union, diversity and competition, political culture, and incumbency. UNION AND DISTRICT SIZE The size of the union may seem a mundane consideration, but it can be a very important one for politics. The smaller the unions, the less likely they are to have the organizational advantages I outlined above. Small unions should typically have less money, fewer (or no) paid staff, fewer activists, and less organized capacity for political action. And because they are obviously located in small districts, the aggregate stakes that serve to motivate political action should tend to be low. So there is reason to expect, for size reasons alone (and ignoring other factors), that union activity and power will be at their height in the largest districts, and will drop off as the districts and their unions become smaller. What do the data have to say about this? The survey asks candidates whether the unions in their districts engage in specific kinds of activities during school board elections: supporting candidates, giving them money, recruiting people to run, mobilizing union members to vote, mobilizing other citizens to vote, making phone calls for favored candidates, campaigning door- 9

11 to-door, and providing mailings and favorable publicity. The results, aggregated to the district level, are described in Table 1 1. The relationship between union political activity and the size of the school district (and the local union) is quite dramatic. In districts with enrollments of less than 1000, teacher unions often seem to play little or no role at least overtly in school board elections. They regularly support candidates for office in less than 25% of these smaller districts, and most other political activities are even less common. They make campaign-related phone calls in just 15% of the cases, for example, they engage in door-to-door campaigning in 15%, and they providing mailings and other publicity in 18%. As districts get larger, however, the unions themselves become larger and more politically capable as organizations and they become much more active along all dimensions, to the point that, in the biggest districts with the biggest unions, every mode of political action we ve explored here (except perhaps for recruiting candidates) becomes the norm. The unions regularly support candidates for office in 92% of these districts. They make phone calls in 97%, they campaign door-to-door in 68%, and they providing mailings and publicity in 75%. Large teacher unions are almost always thoroughly political organizations, geared for electoral action. Small unions aren t. Given these differences, it is only natural to expect that the teacher unions will be much more powerful in the larger districts than in the smaller ones, and thus more successful at getting 1 As there are often multiple candidates per district, we need to move from individual responses to summary scores for each district. One way to do this a method I will rely upon throughout this paper is to quantify candidate responses, average them within districts, and then categorize districts according to their average scores. In this case, candidate responses for any given dimension of union activity are assigned a 1 for yes, a 0 for no, and averaged within each district. Districts are then coded as having that type of union activity if their average is greater than.5, and they are categorized as not having it if their average is less than.5. For a relatively small number of districts, the average is exactly.5, and they can t be placed into one of these categories. They are labeled as mixed. 10

12 sympathetic candidates elected to office. So let s take a first look at (perceived) union influence and see what the data seem to suggest. As one way of getting at influence, the survey asked candidates a simple question: In general, how important would you say the teachers union is in your district s school board elections? Their responses are summarized in the bottom portion of Table 1. According to candidates, the teacher unions are often electorally important even in the smaller districts, where we know they are not very active. In districts with less than 1000 students, unions are regarded as important in 43% of the districts and as unimportant in 46%--a rather high level of importance, given the unions starkly low level of activity. In districts with students, moreover districts that are still quite tiny, with unions seemingly inactive the unions are regarded as important in 57% of the cases and unimportant in just 31%. Something is clearly going on in these districts that is not being picked up by activity levels alone. Nonetheless, it is also true that assessments of union importance grow with the size of the district. And in the largest districts, the unions are almost always regarded as important (82%) and rarely regarded as unimportant (11%). Big districts have powerful unions or so it seems. DIVERSITY AND COMPETITION The greater the number and variety of social groups that populate a district, and the more these groups are organized for political action, the more the unions will face competitors and the less their power should be. If school districts are like other geographical units, where heterogeneity typically grows with size, then these conditions are more likely to prevail in larger districts. So we need to moderate our initial expectation that the unions may be especially powerful in those contexts. A pluralistic struggle among groups may put some limits on what they can achieve. 11

13 As I implied earlier, it is important not to get carried away with this argument. Many broad constituencies are unlikely achieve effective political organization due to inevitable collective action problems. Parents, although they presumably care about education and the schools, are typically only organized through the local PTA (if then); but the PTA is a parentteacher organization, not solely a parent organization, and research has shown that it rarely takes positions in opposition to the teacher unions (Haar, 2002). Moreover, virtually all other groups that manage to get organized business groups, community groups, ethnic groups, religious groups are not focused on education. They have broader social, economic, and political concerns, and are typically less motivated and less continuously active on educational issues than the teacher unions are. The unions, meantime, are not just any old group in an apparently pluralist system. They have vested interests in public education, and they are totally focused on educational issues. In some districts at some times times of crisis or scandal or deep frustration, say business and perhaps other groups may take it upon themselves to make serious investments in educational politics. But under most conditions, it is reasonable to think that the unions have big advantages over all these groups and are well situated to shape outcomes in the usual course of events. Even so, it is worth recognizing that diversity and competition may moderate their power somewhat in the larger districts. The corollary is also worth emphasizing for the smaller districts: although the unions there have less organizational capacity because of their size, they should also tend to face fewer organized competitors for power. Because of the pluralism factor, then, there may not actually be a yawning power gap between large and small unions and it could be a mistake to think of the bigger districts as somehow being in another league when it 12

14 comes to union power. The unions may do quite well in many of the smaller districts, all things considered. If they don t do well in these districts, it could be that there is still another factor at work: that some and perhaps many small districts may function as genuine communities as reflected in strong social networks, personal knowledge of the candidates, and norms of community service (Putnam, 2000). In these sorts of closely knit societies, which are only likely to exist in small districts (if there), it could be difficult for special interests like unions to be overtly active and influential. To the extent that such communities are present, then, the power gap between the large and the small would widen, countering the effects (toward convergence) of diversity and competition. The survey does not provide data on community, but it does give us opportunities to explore certain aspects of pluralism. Let s begin by looking at the competitiveness of elections. Candidates were asked to characterize their districts school board elections in one of two ways: usually they are vigorously contested with a lot of campaign activity, or there is usually little competition or campaign activity. Their responses, displayed Table 2, suggest that the electoral context changes radically as the districts increase in size. Elections are seldom vigorously contested in the smaller districts. Of districts with less than 1000 students, only 13% are regarded as having competitive elections. When we turn to districts with between 1000 and 2500 students, the corresponding figure is just 14%. And for districts with enrollments of 2500 to 5000, it is 17%. As district size continues to increase, however, the level of competition and campaigning increases dramatically to the point that, in the largest districts, 61% are regarded as having highly competitive elections, as compared to just 18% with little competition. 13

15 We have at least one good indication, then, that the unions in larger districts do indeed face more competition than they do in smaller districts, as we should expect. But does this competition really limit union power very much? And what about the smaller districts does the lack of competition mean that all groups, including the unions, are without influence? Or does it mean that there are fewer obstacles in the way of union control? We cannot answer these questions with certainty, of course, but the survey does shed some additional light on them. For instance, it asks candidates whether, compared to all other organized groups in their districts, the teacher unions are the most influential. The district-level results, set out in middle portion of Table 2, are interesting on two counts. First, while the unions are regarded as more influential in the larger districts, the differences across contexts are not very impressive or consistent and the unions show surprising strength in the smaller districts, where they play little overt role in campaigns. Second, the teacher unions in the larger districts do not come across a consistently dominant political force. They are rated as the single most influential group in 50% of the largest districts and in 42% of the next-largest. These figures, I should emphasize, are hardly indicative of the kind of pluralistic politics that education scholars seem to envision. Any pluralism here is clearly tilted in favor of one special interest group: the teacher unions. But the data do suggest that, in about half the larger districts, other groups are regarded as either more powerful than the unions or as competing with them on a more-or-less equal footing. This pluralistic counterbalance to the unions, however, is weaker than it might appear. Candidates who said the unions are not the most influential group in district politics were asked to indicate what other organized groups are more influential, and their responses are enlightening. As Table 3 shows, the one type of group that education scholars would surely expect to rise to the top here business is singled out for its influence in just 22% of the large 14

16 districts and 11% of the smaller districts. Remember, we are dealing here solely with the districts in which the unions are said not to be dominant, so these figures are especially telling about the weak role of business. With perhaps one exception (parents in small districts), the other types of possible competitors party and ideological groups, religious groups and churches, newspapers, other unions, prominent individuals, community groups, ethnic groups do no better than business, and usually do worse. If the teacher unions are being effectively counterbalanced, then it is happening through the action of different types of groups in different districts. In some districts, community groups may be especially influential. In others, prominent individuals. In still others, religious groups. But there is nothing systematic going on here. There is no dependable, consistent counterweight to union power. In small districts, parents come closest to providing such a counterweight. They are the stand-outs. But they are regarded as more influential than the teacher unions in just 34% of such districts, which hardly means that they can be counted upon in this respect. And as I said earlier, there is no reason to think that parents are effectively organized outside the PTA, and thus that they can truly counterbalance the unions. If we go back and revisit Table 2, a closer look at the numbers raises even more questions about the real potency of pluralism. In the first place, the groups most often singled out in large districts (and sometimes in small districts) as more influential than the teacher unions are other unions meaning, most often, the unions representing classified (nonteaching) employees of the school districts. These unions are allies of the teacher unions, and they are clearly not a counterbalancing force. In the second place, as the table shows, many candidates were not able to name any group at all that is actually more influential than the teacher unions, so there is 15

17 reason to think that the teacher unions are in fact the most influential organized group in those districts. If we reclassify the problem districts the ones in which no other group was mentioned, or in which only other unions were mentioned then things begin to look quite different. As the adjusted figures at the bottom of the table show, the unions appear to be the major political force in 68% of districts overall, with other groups judged more important in just 18%--and interestingly, about the same ratio prevails for districts of all sizes. This reclassification makes some assumptions and may err on the side of attributing too much power to the unions. (The multivariate analysis will speak to this issue later on.) I offer it here for purposes of comparison, knowing that the truth may lie somewhere in-between the two classifications. Even so, there is good reason for doubting that the diversity of groups often observed in school politics is providing a genuinely pluralistic politics. This is not to claim that the teacher unions have an easy time getting their way, nor that other groups don t count. But it does appear that the kind of pluralism operating in local elections tends to be tilted in favor of the teacher unions, and that this is the case for both small and large districts. Before moving on, I want to add one more finding of relevance to the pluralism issue. As I said, conventional accounts of school board politics have highlighted the key role of business which suggests that, if there is an effective counterweight to the unions, this is the most reasonable place to look for it. We already know that, in our data, business is usually not mentioned as an influential force in board elections. But we might also ask: who are the people who actually get elected to the board and in particular, are they important business people who would provide a counterweight to the unions, and perhaps give a business tilt to board decision making? 16

18 The impression one gets from the literature is that board members often are prominent representatives of the business community, and that this does indeed give school boards a tilt toward business (e.g., Wirt and Kirst, 2001). The evidence for such a conclusion is pretty slim, however, and a look at our own data suggests a very different conclusion indeed. Consider the information in Table 4, which describes the occupations of candidates who were either incumbents or won their elections. For the issue at hand, the findings are roughly the same for large and small districts, so I will simply focus on the former. In the larger districts, 24% of board members either work for the public schools or are employees of other educational institutions (e.g., community colleges). Another 28% have no job at all; usually they are retired or are housewives. So some 52% of all board members are clearly not prominent business people (except possibly for retired ones). The remaining 48% might plausibly be categorized as coming from the business community but this would be quite misleading. In a market economy, it is inevitable that most people in any school district, and probably most board members, will work in the private sector and thus for business. But if we look at the actual jobs of the people who get elected, it becomes clear that any conclusion that they somehow represent business interests, or serve as some sort of coherent group or pluralistic counterweight, would be a big mistake. In the larger districts, 18% of all board members who constitute well over a third of all alleged business members are people who hold ordinary jobs and cannot be considered business leaders. A sample: auto mechanic, forester, x-ray technician, insurance broker, postal worker. Almost none of the other categories of business members can be regarded as representatives of the business community either. 10% have professional occupations; and although there are some attorneys who may fit the bill, most professionals are psychologists, 17

19 optometrists, ministers, and the like, and they don t fit the bill at all. 5% are self-employed (beauty salon owner, craftsman, merchant). 3% are from the construction industry (mainly engineers and contractors, who probably care about getting a share of building contracts, but not about representing business interests generally). And 4% are consultants of one type or another (possibly interested in getting consulting business and making contacts within the district). The board members who seem to come closest to the stereotype are those who have managerial-level jobs of one kind or another. But there are precious few of these, just 4% of the total, and most of them are hardly what we can consider leaders of the business community. They hold jobs like: supervisor in a hospital, human relations director, purchasing manager, and director of accounting. There may possibly be a few business leaders in the category labeled general business these are people who gave businessman as their occupation, or some such thing but there aren t many of these (2%), and there is no reason to think they are actually prominent business types. So far as I can tell, in fact, there appear to be no CEOs or true corporate executives in the entire lot. All in all, business makes a very poor showing here. There is no evidence that business is a coherent, organized force in school board politics, except perhaps in unusual cases, and no evidence that its longtime reputation for local power is well deserved. Any notion that it is serving as an effective pluralistic counterweight to the unions is probably overstated, at least most of the time. 2 2 It is easy to think of particular cases in which business has played a huge role in school board politics. In both Los Angeles and San Diego, for example, the business community has been very active in promoting reform and taking on the unions, and similar examples can be found in other states. But these high-profile cases seem to be the exception rather than the rule. 18

20 UNION SIZE, PLURALISM, AND UNION SATISFACTION Let s go back and look more closely now at the smaller districts. There is clear evidence that the unions are not very active in these contexts and that elections are usually not competitive. Politics comes across as downright serene. Yet there is also evidence that the unions are surprisingly influential. How can the unions be influential if they aren t doing much of anything and electoral politics is so uneventful? Here is one possibility. As in most states, the teacher unions in California grew dramatically during the 1960s and 1970s. During this period, they gained collective bargaining rights in virtually all school districts, and by 1980 or so were established political powers. What had occurred was nothing less than a revolution within the education system, and it couldn t help but change the politics of education at all levels: local, state, and national. At the local level, the unions came onto the scene as newcomers facing a system controlled by other social interests; and they had strong incentives to take control for themselves by getting sympathetic people elected to local school boards and keeping them there. While there is no good history to tell us exactly what the unions did in local politics and how successful they were, it is reasonable to speculate given the initial conditions (control by other social interests) and the incentives (to get control) that battles were fought over local school boards in districts of all sizes, a period of political upheaval ensued, and a new political equilibrium emerged that gave greater (and in some places, substantial) weight to the interests of the powerful newcomers. If this theory is roughly correct, then what we are witnessing today is an outgrowth of political battles that were fought more than twenty years ago. In districts where the unions were largely successful, there may be no political battles at all in today s world. The districts are simply in an equilibrium satisfactory to the unions, everyone basically accepts the new reality, 19

21 and politics is peaceful. This could be what has happened in many of the smaller districts, where politics is less likely to be diverse and competitive anyway, and unions rarely have to deal with organized opposition. In the large districts, on the other hand, political environments are inherently more diverse and competitive, and early union attempts to establish control even if successful at some point in time may be threatened more often by organized opposition. The serenity that prevails in many smaller districts, then, may often (but not always) hide the fact that teacher unions are satisfied with what they are getting and simply have no need to be more active than they are. In many cases their inactivity may be a sign of power, not of weakness. Our data cannot test this notion definitively, of course, but they do shed some interesting light on it. Candidates were asked whether the teacher unions tend to be satisfied with the school boards in their districts or would prefer to see members elected that are more pro-union. Absent the scenario I just set out, it might seem that unions would always prefer to see more pro-union members elected to their school boards. But the data in Table 5 show that this is not the case. Overall, candidate responses suggest that the unions are actually satisfied with their school boards in some 39% of the districts statewide, and this percentage varies systematically with the size of the district. It is in the small districts that unions are most likely to be satisfied. In districts with less than 1000 students, for example, where the teacher unions are almost totally inactive in political campaigns, the unions are satisfied 53% of the time By contrast, they are satisfied in 32% of the largest districts, where they are extremely active. The smaller districts, of course, are contexts that have less competition and unions that are less active and important in the electoral process. So it should come as no surprise that union satisfaction is positively associated with these sorts of contexts. If we look inside the small districts themselves, however, these same relationships hold up. I won t lay all these numbers 20

22 out in another table, but suffice it to say that, in districts with fewer than 5000 students, the teacher unions tend to be (1) more satisfied in contexts where they are not very active--they are satisfied in 50% of the small districts with little union political activity, but in 33% of the small districts with relatively high levels of union political activity. (2) more satisfied when they are not regarded as important players in electoral politics they are satisfied in 54% of the small districts where they are regarded as unimportant, but in 42% of the small districts where they are regarded as important. (3) more satisfied when there is little competition or campaigning on anyone s part they are satisfied in 55% of the small districts that are noncompetitive, but in 36% of the small districts that are competitive. Here again, the evidence suggests that the peace and serenity so characteristic of small districts are not indications that the unions are weak there, nor that they aren t getting what they want. Indeed, there may very well be nothing going on in many of these districts, and no overt union activity, precisely because the unions are happy with the status quo there a status quo that was essentially determined years ago. This said, it is worth noting that something else could be going on here too: namely, that when unions find themselves in these sorts of settings which could be settings in which a sense of community is high they have lower expectations about what they should be getting from their school boards. They are not powerful, just more easily satisfied. There is no evidence that this is so, and we cannot put it to the test; but it is plausible and cannot be dismissed. 21

23 POLITICAL CULTURE Different districts may have different political cultures, as measured by a variety of social characteristics: party and ideology, ethnicity, income and education, religion, and perhaps others. Group competition aside, if the electoral system is reasonably democratic in responding to the general public and reflecting its popular culture and this is a big if that I ve left open to question here these differences in public values and orientations should affect the kinds of candidates and platforms that attract public support, the kinds of people that get elected, and thus the ability of unions to get what they want. Three implications for union power stand out, which I ll illustrate with reference to the dimension of political culture that I happen to have information on: party and ideology. The first implication is that some districts those that are Democratic and liberal, in this case will have cultures that are more sympathetic to what the unions want, and they will be more likely to elect union sympathizers to office. Union power should be easier to exercise in culturally friendly environments. The second is that, in environments that are unfriendly Republican and conservative ones the unions do not have to sit by while hostile candidates are elected and their own candidates go down in flames. They can adapt by supporting candidates who are sufficiently reflective of the local culture to be electable sufficiently Republican and conservative, in some sense but who, on the specific issues that unions care about, are more sympathetic to union positions than the other candidates are. In this way, the unions can gain as much as possible and be as powerful as possible under difficult circumstances. The third is that the content of union power will be very different across these contexts. In friendly contexts, the unions may have high win rates and very sympathetic boards. In 22

24 unfriendly contexts, however, their win rates may also be high due to the compromises they have made in endorsing culturally electable candidates, but their victories may generate school boards that are much less sympathetic to their cause. They can be powerful, in the sense of winning elections, yet not very successful in getting what they want. This is the constraint of political culture. Now let s take a look at the data. If we categorize districts as Republican, mixedpartisan, or Democrat based on the party registration of the citizens who live there, we find that union endorsements do indeed vary in a systematic way across political cultures. 3 See Table 6. In school districts that are Republican and (presumably) conservative, 55% of the candidates endorsed by the teacher union are Republicans and 56% are conservatives figures that would otherwise be something of a shock, given the teacher unions almost exclusive support of Democrats in state and national politics (Lieberman, 1997). As the school district populations get more liberal and Democratic, so do the candidates that the unions endorse although, interestingly, the unions continue to endorse fair numbers of Republicans (33%) and conservatives (38%) even in heavily Democratic environments. These findings are perhaps a signal that, in local politics, the usual labels may not mean as much as they do in state and national politics. In any event, what ultimately counts for the unions is not the labels, but how sympathetic these candidates are to union interests. To get at this, the survey asks candidates a number of specific questions about collective bargaining, having to do with its effects on various aspects of public education costs, academic performance, school organization, conflict, teacher professionalism, and teacher quality. As a 3 The party registration data for each school district were gathered from county election offices. I am scoring districts as Republican if at least 55% of the voters who register with one of the two major parties are Republican, Democrat if at least 55% are Democrat, and mixed-partisan if the party balance is somewhere in-between. 23

25 follow-up, it then asks them for a summary evaluation: In general, what is your attitude toward collective bargaining in public education? The results, set out in Table 7 for candidates of different parties, have a lot to say about unions and their political environments. (1) The vast majority of school board candidates, 66%, have positive overall attitudes toward collective bargaining. Even among Republicans indeed, even among Republicans who are not endorsed by the unions the majority take a positive approach to this most crucial of union concerns. Thus, districts that are heavily Republican and conservative may well be unfriendly territory for unions relative to districts that are heavily Democratic and liberal. But they are not necessarily unfriendly in an absolute sense and the unions are not as constrained as they might appear to be. (2) Candidates are more critical of collective bargaining when assessing its specific effects. Almost two-thirds think collective bargaining leads to higher costs, and more than 40% think that it leads to greater conflict and to more rigid organizations. But these aspects of schooling are several steps removed from what happens in the classroom, and even supporters of collective bargaining might agree that it has such effects. The more telling finding is that candidates are quite positive on the two items that are most directly related to how much kids learn namely, academic performance and teacher quality. Only 22% think that collective bargaining has negative consequences for academic performance, and only 20% think it threatens teacher quality. (3) Party does make a difference for candidate attitudes. If we compare all Republicans to all Democrats, disregarding endorsements, Republicans are consistently less supportive of collective bargaining than Democrats are. This is true for their general evaluations 57% 24

26 positive for Republicans, 75% positive for Democrats but it is also true for each of the more specific items. Just 36% of Republicans think collective bargaining leads to better academic performance, for example, but 50% of Democrats do. Similarly, 32% of Republicans think collective bargaining leads to more cooperation within the district, but 41% of Democrats see it this way. These and other differences are not huge a reflection of the fact that most candidates for school board are supportive of collective bargaining. But still, party does matter. (And so, I should add, does ideology. If we were to break these results down by ideology instead of party, all of the same patterns would emerge.) (4) Most important: the unions do appear to use endorsements strategically to promote the candidacies of people who are sympathetic to collective bargaining. Endorsed Republicans are consistently more favorable toward collective bargaining than unendorsed Republicans, and endorsed Democrats are consistently more favorable toward collective bargaining than unendorsed Democrats. Moreover, it is important to note, as a measure of how much the unions are able to achieve (potentially) through their endorsements, that the Republicans they endorse tend to be more positive toward collective bargaining than the Democrats they don t endorse. Indeed, the endorsed Republicans are in some ways (on issues of organization, conflict, professionalism) just as positive as the endorsed Democrats. By using endorsements to support candidates who are at once compatible with the local culture (electable) and relatively sympathetic to collective bargaining, then, the unions can take significant steps especially in districts filled with Republicans and conservatives toward loosening the constraints of political culture. If this assessment is basically correct, and if the unions are strategically adapting to their constraining environments, then even in (relatively) unfriendly districts they may well win 25