Against unionization for graduate students at Cornell

In one week from now, graduate students at Cornell will vote to decide whether they want to be represented by Cornell Graduate Students United (CGSU), in which case this group will gain the exclusive right to negotiate the terms and conditions of employment for every graduate student. To be clear, this means that if more than 50% of the graduate students vote for unionization, CGSU will negotiate the contract of every graduate student and every graduate student will have to pay dues to the group, even if they voted against unionization. To be more accurate, only graduate students who are part of the bargaining unit are concerned, but as we shall see that’s almost half of the graduate students at Cornell. If you want to know more about the process, you can read this FAQ, which the graduate school put together. You can also check CGSU’s website, which I linked to above, as well as the website of At What Cost, a group of students who oppose unionization and seek to provide more information to students before they vote. Obviously, each of these groups has its own agenda, so I encourage you to read everything critically. In what follows, I argue that 1) CGSU is making the case in favor of unionization dishonestly by cherry-picking the evidence to support their claims that students would benefit from it, 2) the groups to which CGSU is affiliated and to which most of our dues would go if we vote in favor of unionization are deeply immoral organizations that have a negative effect on education in the US and we should not fund them with dues taken from our wages and 3) even if people disagree with me about that, it’s wrong to use the law to force your colleagues to pay dues to organizations they don’t want to be associated with.

I have several friends who are involved with CGSU, so I bear them no ill will, but I believe they are wrong and that students should vote against unionization. (This post is primarily intended for graduate students at Cornell, but I will discuss various issues that are of interest to a lot of other people. So you may want to read it even if you’re not a graduate student at Cornell, especially if you care about education and topics such as the role of teachers unions.) First, I think what CGSU says on its website is extremely misleading, but I don’t have time to go through everything in details, so I just want to point out that, despite what CGSU would have you believe, we don’t really know what is going to happen if the graduate students vote in favor of unionization. In particular, we don’t really know whether the graduate students stand to gain from this decision, in terms of compensation and work conditions. The reason we don’t know is that, up until 2016, the National Labor Relations Board (NLRB) didn’t allow graduate students at private universities to unionize, so we don’t really have enough data to know what is likely to happen. In fact, because the NLRB has yet to determine what the rules are for graduate students exactly, we don’t even know for sure what would be up for negotiation in the collective bargaining process if graduate students voted in favor of unionization.

A common misconception among people who support CGSU is that current work conditions and remuneration can only improve with collective bargaining, but that is simply false and anyone who says otherwise is either lying or doesn’t know what he is talking about. What the outcome of the negotiation would be if the graduate students voted in favor of unionization would depend on a lot of factors that are simply impossible to predict. The only thing we know for sure is that graduate students will have to pay several hundred dollars in dues every year. I’ve heard several people say that, if the terms offered by the university during the negotiations didn’t improve the compensation of graduate students (taking into account the dues they will have to pay if they decide to unionize), they would not accept it. But the truth is that, depending on the circumstances, they may not have a choice. Even if the union doesn’t accept the terms offered by the university, if the NLRB — which is going to be filled with people appointed by Trump, so they probably won’t open their sessions with a recitation of passages from the Communist Manifesto — rules that Cornell negotiated in good faith, the university can just implement its last offer whether the graduate students agree with the terms or not. So what’s going to happen will depend on a lot of factors that are difficult to predict, such as how united the graduate students will be, how willing they will be to forgo a substantial amount of money by going on strike if they don’t like the terms offered by Cornell, how willing to engage in a showdown the university is, how the NLRB would rule if no agreement were found between the union and the university, etc.

The claim that collective bargaining can only result in a net gain for graduate students, because they would not accept a contract that doesn’t result in a net gain, is made by CGSU on its website. However, as I have just explained, that is purely and simply false. In fact, even CGSU acknowledges this in passing, for although it says that “there are only things to be gained from a contract” (emphasis in the original), it also says in the same paragraph that “benefits such as stipend increases, better access to healthcare such as dental and vision, and other benefits will likely outweigh the cost of dues” (emphasis mine). This shows that CGSU is aware that there is no guarantee, but they’d rather you didn’t know that. This is exactly the kind of things I had in mind when I said that you should read everything about the election, including this post, carefully before you decide how to vote.

So far, the only private university where graduate students have been able to unionize and where collective bargaining led to a contract is NYU, because the university agreed to recognize a union in 2013, even before the NLRB ruled that graduate students at private universities had the right to collectively bargain. On its website, CGSU claims that “since NYU’s first contract in 2002, [the graduate students] have won [a] 38% increase in minimum stipends”. First, that’s a very misleading claim because, after a contract was collectively bargained in 2002, NYU decided to stop recognizing the union in 2005 following a decision by the NLRB the previous year to no longer allow graduate students to collectively bargain, which itself reversed a decision it had made in 2000. (If you feel like the NLRB is constantly reversing itself on this, that’s because it is, which in turn is because the composition of the Board depends on which party controls the White House. Indeed, since Trump won the election, it would not be surprising if the NLRB changed its mind again on whether graduate students at private universities have the right to collectively bargain, because Trump is unlikely to fill it with pro-union leftists.) After NYU decided to recognize the union again in 2013, it wasn’t until 2015 that a contract was negotiated through collective bargaining. Thus, between 2002 and 2017, the terms of employment of graduate students at NYU were only determined through collective bargaining during approximately 5 years. Are you starting to feel like the people at CGSU are trying to play you? Good, because that’s exactly what they’re trying to do. But don’t go anywhere, because there is more.

Indeed, according to the Inside Higher Ed article I linked to above, the union at NYU only represent 1,200 graduate students. Since there were 20,749 graduate students at NYU, that’s not even 6% of them because the contract excludes a lot of graduate students, especially in the sciences. At Cornell, on the other hand, CGSU would represent every graduate student who is a teaching assistant, a graduate assistant, a research assistant or a graduate research assistant. According to the graduate school, that’s about 2,200 people, out of a total of 5,265 graduate students, i. e. more than 40% of them. So it’s really not clear that we can infer anything from the terms obtained by graduate students through collective bargaining at NYU, because the situation is completely different. (Part of this, I suspect, is because there are more part-time graduate students at NYU. But I wasn’t able to find how many graduate students at Cornell are part-time, so I don’t know that for sure. Even if that’s right, however, it would only go to show how different the situation at NYU is from that which prevails at Cornell.) In particular, since the collective bargaining agreement at NYU concerns a much smaller share of the graduate students than would be the case at Cornell, the university could presumably afford to be more generous.

Moreover, recall that NYU recognized the right of graduate students to collectively bargain on its own in 2013, before the NLRB forced private universities to grant the right to collectively bargain to graduate students in 2016. Cornell, on the other hand, has been fighting against this from the beginning and even filed a brief in court a year ago with several other prominent private universities to prevent graduate students from being able to collectively bargain. This suggests that, whereas NYU was at least open to negotiate with a graduate students union, Cornell is not, which in turn means that we can’t infer much from what graduate students were able to obtain through collective bargaining at NYU about the terms graduate students at Cornell would be able to obtain if they decided to vote in favor of unionization. But the administration can’t tell you that it doesn’t have any intention to make concessions to the graduate students, because under the law, this could be considered a sign that it doesn’t intend to bargain in good faith if CGSU wins the vote, which constitutes a legal risk. On the other hand, CGSU can promise anything it wants, including things it knows perfectly well are never going to happen, because for a union this doesn’t involve any legal risk. Finally, note that a 38% increase in minimum stipends since 2002, which is what graduate students at NYU gained since 2002 according to CGSU, is not very impressive anyway. Indeed, that’s barely more than what they would have gained if their stipends had just tracked inflation during that period, based on the CPI. In fact, since I suspect that inflation was higher in NYC than at the national level (which is what the CPI measures), it’s likely that it didn’t even compensate them for the loss of purchasing power caused by inflation.

CGSU also makes a lot of what graduate students at Oregon State University were able to gain since 2001 when they first negotiated a contract through collective bargaining. Again, this is completely dishonest, since they obviously cherry-picked this particular university to make unionization more attractive and we have no reason to think that unionization was responsible for those gains. I found this study which looks at several dozens public universities and uses a multilevel regression model to estimate the effect of graduate student unionization on stipends. It suggests that, while unionization increases stipends at public universities, this is compensated and perhaps more than compensated by reductions in tuition remission and increases in other fees, so that net compensation does not increase and may even be reduced. (I have only skimmed through the paper, so there may be something wrong with the data and/or the model, but nothing jumped out at me.) Moreover, according to that study, unionization has no effect on the probability of obtaining health benefits for students and their dependents. (Which is one of the main arguments used by CGSU to convince graduate students to vote in favor of unionization.) Finally, Oregon State University is a public university, which implies many significant differences with Cornell. In particular, the contract is subject to state labor law, whereas since Cornell is a private university federal labor law would prevail if graduate students decided to vote in favor of unionization. So again we really can’t infer much from the experience of unionization at public universities and, even if we could, things aren’t nearly as nice as CGSU pretends by cherry-picking one public university where graduate students are represented by a union.

I could say more about what CGSU says on its website to convince graduate students to vote in favor of unionization, but I think what I have already said should be enough, so I’m going to leave things at that on this front. At this point, you should be asking yourself why, if CGSU really has only your best interest at heart, it’s making the case in favor of unionization in such a dishonest manner. Personally, when I realize that someone is trying to bullshit me, my first reaction is not to entrust them with negotiating the terms of my employment on my behalf in exchange for dues taken from my wages, but perhaps that’s just because I have a suspicious nature… Now let’s talk about what I consider the most problematic consequences a vote in favor of unionization by the graduate students at Cornell would have. So far, I have only discussed some of the arguments that CGSU has been using to convince graduate students to vote in favor of unionization, but now I want to briefly explain why I think they should vote against it. I have one argument directed specifically at CGSU and another that is more general and is critical of the way collective bargaining works in the US.

The problem I have with CGSU specifically is that it chose to be affiliated with the American Federation of Teachers (AFT) and New York United Teachers (NYSUT), which are some of the largest teachers unions in the US and, in my opinion, are morally bankrupt. In fact, if you ask me, teachers unions are the eleventh plague of Egypt. The only reason they are not mentioned in the Bible is that the authors of the book of Exodus thought it would give a bad name to God if they recorded that he unleashed teachers unions on the Egyptians. It was one thing to kill their firstborn sons, but forcing them to send their kids to schools controlled by teachers unions was considered too much, so they figured it was preferable not to mention it. (Obviously, I’m joking, but not as much as you might think. In fact, this is my best effort to be civil, so as not to shock people’s sensibilities when I talk about teachers unions.) The fact that CGSU is affiliated to the AFT and NYSUT means that, if graduate students vote in favor of unionization, everyone will pay dues to both the AFT and NYSUT, even if they voted against unionization. According to CGSU, everyone will pay at least $397.68/year in dues to the AFT and NYSUT ($606.36/year for graduate students who make $34,000/year or more), but this doesn’t even include dues paid to CGSU itself. We have no way of knowing exactly how much these will be, since it will depend on the outcome of the collective bargaining process. The graduate students union at NYU takes 2% in dues, so if the same thing happened at Cornell (but remember that we have no reason to assume it would, although presumably the amount would be in the same ballpark), someone who makes the minimum stipend of $25,152/year would have to pay $503.04/year in dues.

As you can see, the bulk of your dues would go to the AFT and NYSUT, which as I have already noted are horrible organizations. Since you probably have never heard of the AFT and NYSUT, let me tell you a few things about them, to support my claim that they are morally bankrupt. Let me start with some anecdotes, which I think are pretty revealing. For several decades, teachers at public schools in Buffalo benefited from a contract that included a cosmetic surgery rider, which covered procedures such as face lifts, tummy tucks, breast implants, etc. During the last year of the contract in question, $5 million had to be set aside to cover the cost of the rider, even though the school district was facing a $10.5 million deficit. The Buffalo Teachers Federation (BFT), which as CGSU is affiliated to both the AFT and NYSUT, must have thought that breast implants were essential for teaching, because when the School Board tried to end the rider and use the money to help cover the deficit, it sued and obtained a temporary injunction from a court to prevent it.

The BFT argued that it was opposed to the rider, which is probably why it had never proposed to end it on its own, but didn’t want the Board to unilaterally change their contract. In fact, they used that as a bargaining chip to obtain a significant raise and several other advantages a few months later, in exchange for giving up the rider and making a few other largely insignificant concessions. It’s not as if the teachers in Buffalo’s public schools were forced to live in abject poverty before that either, since their median income was already $50,886/year before the new contract, which is 68% more than the median personal income in the US according to the Fed. (Of course, this doesn’t include the various benefits, such as free breast implants.) It’s not the just students and taxpayers that were harmed, but even some of the teachers. According to this article in The Atlantic, a few years ago, the School Board offered to avoid 100 layoffs in exchange for suspending the cosmetic surgery rider for a year, but the union declined. The same union which, you will recall, claims that it was totally opposed to the rider in question… Nor were this an isolated example, for horror stories of this kind about teachers unions abound, as you can easily find out for yourself if you do a little research.

A few days ago, the Board of Regents in New York State dropped the requirement to pass a literacy test  for people who want to become teachers, on the ground that it was discriminatory toward minorities. The proponents of that decision, including teachers unions such as NYSUT (one of the organizations to which CGSU is affiliated), claim that the test was redundant since candidates were already required to have a Bachelor. Indeed, it was so redundant that, in 2014, only 68% of candidates passed it. Of course, aspiring teachers will still have to pass other tests in order to be certified, but a much larger share of the applicants pass them, which means that presumably a significant number of people who previously couldn’t get certified because they were not able to pass a literacy test will be able to teach. So much for the alleged redundancy of the literacy test in question. But perhaps the proponents of that change don’t know what “redundant” means because they also couldn’t pass a literacy test… Of course, none of the illiterate teachers who obtain their certification thanks to this decision will end up in schools where rich kids go, but you can bet that plenty of them will end up in classrooms full of poor and minority students, who will have teachers unions to thank for this great opportunity.

Now, these are just anecdotes, but I could write a whole book full of them and, perhaps more importantly, there is more than anecdotal evidence to back up the claim that teachers unions such as the AFT and NYSUT have a negative effect on education in the US. Here is the abstract of a review of the literature on the effect of teachers unions published in 2015:

In this paper we consider more than three decades of research on teachers’ unions in the United States. We focus on unions’ role as potential rent-seekers in the K-12 educational landscape, and specifically how teachers’ unions impact district and student outcomes. We review important methodological improvements in the identification of union impacts and the measurement of contract restrictiveness that characterize a number of recent studies. We generally find that the preponderance of empirical evidence suggests that teacher unionization and union strength are associated with increases in district expenditures and teacher salaries, particularly salaries for experienced teachers. The evidence for union-related differences in student outcomes is mixed, but suggestive of insignificant or modestly negative union effects. Taken together, these patterns are consistent with a rent-seeking hypothesis. We conclude by discussing other important union activities, most notably in the political arena, and by noting that recent changes in state laws pertaining to teachers and teacher unions may provide context for new directions in scholarship.

In other words, teachers unions significantly increase expenditures (at the expense of the taxpayers), yet have insignificant or perhaps even somewhat negative effects on student performance. Thus, in terms of cost-effectiveness, teachers unions have a substantial negative effect on education. You should read the whole thing if you’re interested in that issue. Note that this is a review of the literature, so it doesn’t rely on a single study but takes into account the whole literature.

But it’s even worse than you think, because as this paper acknowledges in section 6, the studies reviewed weren’t concerned with the effect teachers unions have on state and federal policy about education, which in my opinion is by far how they do the most damage. According to the Center for Responsive Politics, a nonpartisan group which tracks money spent on lobbying by various organizations, teachers unions (mostly the AFT and the NEA) have spent more than $148 million on federal political lobbying since 1990. The AFT alone, to which CGSU is affiliated, spent more than $92 million during that period, which makes it one of the largest political contributors in the US. In fact, the AFT spent $27 million more on federal politics during that period than Chevron, Exxon Mobil and the NRA combined. Moreover, this doesn’t even include the money spent on state politics, which is where most of the action happens on education because in the US it’s mostly a state business. Although they like to present themselves as David against the corporate Goliaths of the world, the fact of the matter is that teachers unions are among the most powerful lobbies in the US. They use that influence to systematically oppose any meaningful reform of the education system. Of course, if you ask them, they will tell you that they are totally in favor of reforms. But if you look at what they call “reforms”, you will see that it always consists in pouring taxpayer money into failing systems, from which teachers unions derive their power. Poor kids and minorities are disproportionately harmed by the fact that teachers unions systematically oppose any effort to reform education in the US. This post is already getting long and I have work to do, so I won’t get into the details, but I encourage you to do your own research if you don’t believe me.

If the graduate students at Cornell vote in favor of unionization, this is how their dues will be used, since most of them will go to the AFT and NYSUT. Personally, I don’t want anything to do with those organizations, because I think they are deeply immoral and I’d rather cut my arm than give them even one dollar. But perhaps you have done your own research and disagree with me. (Although, if you’re going to disagree with me about this, you have to do your homework. It would be irresponsible for you to vote in favor of unionization without even knowing whether your dues are going to be used to support immoral policies.) This brings me to the other, more principled reason why I think you should vote against unionization. It’s fine if you disagree with me about the role of teachers unions in the US and think it’s actually positive. Reasonable people can disagree about that kind of things. (Although, to be honest, I think the case against teachers unions is pretty overwhelming.)

The problem is that, if the graduate students vote in favor of unionization, even people like me who strongly disapprove of the AFT and NYSUT will be forced to pay dues that will go to these organizations and will be used to support policies they consider deeply immoral. That’s because the National Labor Relations Act (NLRA) says that, if more than 50% of the voters in a bargaining unit vote for unionization, then everyone is represented by the union and forced to pay dues to it even if they voted against it. As I have explained, the AFT and the NYSUT are eminently political organizations (almost every dollar they spend to support candidates running for office are Democrats, although they also spend a lot of money against policies supported by Democratic politicians, as long as they perceive them as a threat to their rent), so by voting in favor of unionization, you would essentially force even people who strongly disapprove of these organizations and their activity to fund them by using the law to automatically extract dues from their wages. This is why I consider the NLRA a deeply unjust law that should be repealed as soon as possible. If you’re in favor of unionization, just ask yourself if you would be okay if other graduate students were using the law to force you to fund the NRA and, indirectly through its political contributions, the Republican party. I don’t think there are many people who support the CGSU who could look me in the eyes and tell me that they would be okay with that if the situation were reversed. But if they would not accept that kind of treatment for themselves, then they should not try to impose it on others.

I imagine that some people will reply that, in a democracy, even people who didn’t vote for the person who won the election have to pay taxes that are used in a way they sometimes find morally repugnant. As someone who has to pay federal taxes in the US, which are then used to finance wars I find completely immoral, I’m not going to deny that. But I don’t think you can really compare that with collective bargaining under the NLRA. Almost everyone agrees that we need a government that collect taxes and, however imperfect that system is, most people also agree that majority vote is the best system to decide how much taxes are collected and how they are going to be used. On the other hand, we don’t have to force workers to pay dues to a union even if they don’t want to, it’s just something that some people try to do because the law is unjust and allows them to do so as long as they manage to convince more than 50% of the people eligible to vote in a bargaining unit.

Now, don’t get me wrong, I have nothing against unions per se and the idea that workers should collectively organize to negotiate better terms of employment with their employers. (In the case of graduate students at Cornell specifically, I also disagree with the administration’s argument that graduate students aren’t really workers, so they are not entitled to collectively bargain. I think collective bargaining as it currently exists under US law is wrong, but I don’t think the fact that we are graduate students makes any difference.) On the contrary, I’m all for that, but I don’t think it should be possible for unions to force workers who don’t want to be represented by them to pay them dues and let them negotiate their terms of employment on their behalf. People will say that, if workers can opt out of the union in their workplace, they will be free riding, i. e. enjoy the advantages negotiated by the union without paying dues. But this is confused, because I’m not advocating right-to-work laws, of the sort that exist in many states and that Republicans defend. The only reason why workers would be able to free ride is because the NLRA gives the union the right to represent everyone, including those who don’t want to be represented by it. What I’m saying is that the NLRA should be repealed, so that this is no longer possible.

In my opinion, a union should only be able to bargain on behalf of the workers who agreed to be represented by it, but not anyone else. People will say that, if a union only represents the people who agreed to be represented by it, it will have less bargaining power to negotiate with the employer. This is absolutely true but that’s exactly as it should be. A union should only have as much bargaining power as workers allow it to have by freely choosing to be represented by it. If it does a good job and convince more workers that it’s in their best interest to let it represent them, then it will have a lot of bargaining power because it will represent a large share of the workforce, but otherwise it won’t and there is nothing wrong with that. Under such a system, unions would have a strong incentive to act in the best interest of workers, since otherwise people would just rescind their membership and stop paying dues. By contrast, under the NLRA, people have to pay dues even if they’re not satisfied with the union and, once they have voted in favor of unionization, it’s very difficult to reverse that decision and get rid of the union. Of course, graduate students at Cornell can’t decide to repeal the NLRA, but they can decide not to use that unjust law to force their fellow students who don’t like the AFT and NYSUT to pay them dues even though they don’t want to. All they have to do is vote against unionization in the upcoming election.

Anyway, this post is getting really long and I want people to read it, so I will stop writing. Before I do so, however, I want to make one last observation. As you can see if you browse my blog, I have various controversial views about a lot of issues, which you may or may not agree with. If a lot of people share this post, as I hope they will, I have no doubt that supporters of CGSU will try to discredit me by bring up my views on other, unrelated topics. But even if you don’t like some of the views I defend on other topics, it doesn’t mean that I’m wrong about unionization. I have written a long and carefully documented post to defend my view about the upcoming election, so you should only care about my arguments and the evidence I provide. By all means, regard anything I say with suspicion and verify if you’re skeptical (I’m not worried), as you should do with anyone else. I have provided at least one source for every single factual claim I make, so you just have to follow the links to check that I’m not misrepresenting anything. But don’t let yourself be distracted by people who, instead of replying to my arguments, commit the ad hominem fallacy against me. If people share this post and the supporters of CGSU don’t reply to my arguments but engage in that kind of tactics instead, then I suggest that you consider the possibility that perhaps it’s because they can’t…

34 thoughts

  1. ” It’s not as if the teachers in Buffalo were forced to live in abject poverty before that either, since the average teacher in the district already earned approximately $52,000/year before the new contract according to The Atlantic, which is 72% more than the median personal income in the US according to the Fed.”

    This is triply misleading.
    1. You are comparing a median to an average. Average US income is over 50K/year.
    2. You are comparing full time employees with at least a bachelor’s degree to a group that includes part time 15 year old paper boys. Median full-time employed with a BA or more income in the US is $71k . If average is similarly 50% higher, than that would make relevant number about $106k.
    3. NY incomes are higher than national average by about 10%.

    1. 1) It would make no sense to compare the average personal income of teachers in Buffalo’s public schools with the average personal income in the US, because the latter is artificially driven up (compared to the median personal income) by the massive inequalities that characterize the income distribution in the US, whereas the former is not because, in part as a result of union activity, the distribution of wages in education is very egalitarian. Thus, according to data from March 2016 (before the raise), the median income of teachers in the district of Buffalo was $50,886/year. But I’ll edit my post just for the sake of accuracy.

      2) My only intention was only to debunk the miserabilist account of the income of teachers in public schools, so I think it was appropriate to make the comparison to the median personal income in the US. But if you insist on making an apple-to-apple comparison, it would probably make more sense to compare the average personal income of teachers in public schools with that of teachers in private schools, which at the national level is about $36,000/year. In other words, the average teacher in Buffalo’s public schools makes about 39% more than the average teacher in a private school nationally, and this doesn’t even include benefits like free breast implants…

      3) Actually, while median personal income in New York State is higher than in the US, it seems to be lower in Buffalo. (This is about household income and not personal income, since I wasn’t able to find data about that, but presumably it also applies to personal income or, at least, there isn’t a substantial difference between the national and local personal income.)

  2. “A common misconception among people who support CGSU is that current work conditions and remuneration can only improve with collective bargaining, but that is simply false and anyone who says otherwise is either lying or doesn’t know what he is talking about.”

    Very often the union DOES lose in a first contract, as anyone with labor relations and collective bargaining experience would attest to if they are being honest. The problem with the rules about representation elections is that the union is essentially allowed to make a ridiculous claim such as “your conditions will definitely improve under a first contract” while management is restricted from saying “we will seek to reduce wages/otherwise change conditions in a way you will perceive as negative” even if it is true because it would be an indication that they are not going to bargain in good faith.

  3. According to the literature review you’ve cited (the only piece of non-anecdotal evidence on the general effects of teachers unions mentioned in your post), the evidence of union effects on student outcomes is either insignificant or modestly negative. On the other hand, there is good reason to believe that improving educational outcomes has little to reducing poverty. Furthermore, as I’ve been arguing during our (rather, ahem, one-sided) discussion of Sanandaji’s bad book, the most effective way to reduce poverty is with social democratic transfers. It’s no coincidence that social democracies have far higher union density than the United States (and teachers unions are especially powerful, for example in Finland). It is exceedingly likely that in order to make generous redistributive transfers politically likely, high union density is necessary. Therefore, high union density is politically necessary for poverty reduction (aside from whatever direct economic effects it might or might not have).

    Now, as you appear to concede, the effects of unions on student educational outcomes is either nil or slightly negative. Even if it is slightly negative, the modest negative effect on poor students would be more than compensated for by more generous transfers to poor families, especially because education has little to do with poverty. But generous transfers are politically unlikely unless unions are very powerful. So anyone who cares about poor students should want unions to become very powerful in the United States. The fact that graduate student wages will go to NYSUT, the AFT, and the AFL-CIO as dues, is a feature, not a bug.

    You further claim that the NLRA “forces” people to hand their wages over to organizations with which they disagree, and that this is “unjust”. I find these claims surprising coming from a consequentialist. Certainly, for you, whether someone is “forced” to do something is morally irrelevant. What matters is whether the consequences of forcing people to do that thing are good or bad.

    You consider the analogy with taxes, but you dismiss this by observing that “Almost everyone agrees that we need a government that collect taxes and, however imperfect that system is, most people also agree that majority vote is the best system to decide how much taxes are collected and how they are going to be used.” But surely it makes no difference whether “Almost every one agrees…” or “most people… agree….” What matters is whether they are right to agree. For a consequentialist, if the result of using majority rule to “force” people to pay taxes is good, then we should do it; if it’s bad, we shouldn’t. Similarly, if it’s good to “force” people to pay union dues, then that should be done. If it’s bad, then it shouldn’t. I’ve argued above that it’s good.

    You ask readers to consider the following thought experiment:

    If you’re in favor of unionization, just ask yourself if you would be okay if other graduate students were using the law to force you to fund the NRA and, indirectly through its political contributions, the Republican party.

    But given what I’ve said above, the thought behind your request reduces to this tweet.

    I should further explain why I’ve been putting scare quotes around “force” and its variants. It’s because I’m not at all convinced that paying union dues as a condition of one’s employment is “force” in any coherent sense. Suppose my employer decides to reduce my wages, and donates the money saved to the NRA. Have I been “forced” to donate to the NRA? Conservatives and libertarians say no, and the reason invariably given is that I am part of a voluntary contract with my employer, and if I don’t like it I can always quit my job. The typical left response is this argument is to question how “voluntary” such a relationship is in an economic system in which workers don’t own the means of production and must work for a living, or something along those lines.

    It’s not that I disagree with such responses, it’s that I think they are missing a much more obvious point: You can also quit your job if you don’t like paying union dues. If union dues are “forced”, so are wage cuts. If wage cuts aren’t “forced”, neither are union dues. You can’t have it both ways.

    The same hold for a host of other policies that reactionaries like to claim involve “force”. Are you an employer who doesn’t like being “forced” to pay your workers the minimum wage? Liquidate your business. Don’t like taxes? Quit your job and live in the forest. It’s no different from asking a worker to quit if she doesn’t like her pay or working conditions.

    1. According to the literature review you’ve cited (the only piece of non-anecdotal evidence on the general effects of teachers unions mentioned in your post), the evidence of union effects on student outcomes is either insignificant or modestly negative. On the other hand, there is good reason to believe that improving educational outcomes has little to reducing poverty.

      I agree that both liberals and conservatives probably overestimate how much education can do to reduce poverty, but there is no doubt that it helps and the post you mention, which is statistically incompetent, doesn’t show otherwise. Basically, it points out that even though the general level of education has significantly increased between 1991 and 2014, poverty has increased during the same period. But of course this doesn’t show that education doesn’t help to reduce poverty, since plenty of other things happened during that period that could have affected the prevalence of poverty, such as the largest recession in almost a century, to cite only the most obvious.

      Furthermore, as I’ve been arguing during our (rather, ahem, one-sided) discussion of Sanandaji’s bad book, the most effective way to reduce poverty is with social democratic transfers. It’s no coincidence that social democracies have far higher union density than the United States (and teachers unions are especially powerful, for example in Finland). It is exceedingly likely that in order to make generous redistributive transfers politically likely, high union density is necessary. Therefore, high union density is politically necessary for poverty reduction (aside from whatever direct economic effects it might or might not have).

      Of course, you haven’t shown that social democratic transfers were the most effective way to reduce poverty, but we’re not going to solve that debate here so I’ll just leave it at that. (As I already told you, I will reply to you on the debate about Sanandaji’s claim relative to poverty, especially since you’re being snarky ^_^ But I’m completely swamped with work at the moment because of job market stuff, dissertation-related work and grading, so I’m not sure when I will.) On the other hand, it’s patently false that without strong unions, large social transfers are unlikely to happen. As I’m sure you know, unions have become significantly less strong in practically every developed countries in the past few decades, yet in virtually all of them public social expenditures have increased considerably during the same period. At the very least, this shows that, despite what you say, high union density is not a necessary condition for poverty reduction even if we accept your claim that social democratic policies are the most efficient way of reducing poverty, which we shouldn’t.

      Now, as you appear to concede, the effects of unions on student educational outcomes is either nil or slightly negative.

      I most certainly don’t concede that. As I explain in my post, the review of the literature I mention only analyzed studies that bear on the most direct/least indirect effects of teachers unions on educational outcomes, but as I also say in my post I think unions do far more damage through political lobbying, whose effects the studies in question for the most part couldn’t possibly detect since that’s not what they are designed to do. I guess one could argue that political lobbying by teachers unions doesn’t have a negative effect on education, but it’s hard to see how anyone who is familiar with the political activities of teachers unions could seriously defend that view. In any case, we are not going to solve that question here, because that’s another huge debate. It’s also worth noting that even if we only consider the most direct/least indirect effects of teachers unions on educational outcomes, the literature suggests that they are slightly negative but also that teachers unions have a massive negative effect on costs, which means that in terms of cost-efficiency the literature actually suggests that teachers unions have a very substantial negative effect on educational outcomes even if we ignore the effect of their political activities. (The question of cost is something that is systematically ignored by pro-union people in discussions about education.)

      Even if it is slightly negative, the modest negative effect on poor students would be more than compensated for by more generous transfers to poor families, especially because education has little to do with poverty. But generous transfers are politically unlikely unless unions are very powerful. So anyone who cares about poor students should want unions to become very powerful in the United States. The fact that graduate student wages will go to NYSUT, the AFT, and the AFL-CIO as dues, is a feature, not a bug.

      That’s only true if we accept the claim that unions have a substantial positive effect on poverty, which I and a lot of other people who are perfectly reasonable don’t. To be sure, we could be wrong about this, but it’s hardly obvious.

      You further claim that the NLRA “forces” people to hand their wages over to organizations with which they disagree, and that this is “unjust”. I find these claims surprising coming from a consequentialist. Certainly, for you, whether someone is “forced” to do something is morally irrelevant. What matters is whether the consequences of forcing people to do that thing are good or bad.
      You consider the analogy with taxes, but you dismiss this by observing that “Almost everyone agrees that we need a government that collect taxes and, however imperfect that system is, most people also agree that majority vote is the best system to decide how much taxes are collected and how they are going to be used.” But surely it makes no difference whether “Almost every one agrees…” or “most people… agree….” What matters is whether they are right to agree. For a consequentialist, if the result of using majority rule to “force” people to pay taxes is good, then we should do it; if it’s bad, we shouldn’t. Similarly, if it’s good to “force” people to pay union dues, then that should be done. If it’s bad, then it shouldn’t. I’ve argued above that it’s good.

      You seem to be assuming a very simplistic form of consequentialism. As I’m sure you know, consequentialists can and many of them do support norms even if their application can be suboptimal in particular cases, if they think that on the whole they are likely to have good consequences. One important norm, in my opinion, is that as far as possible the politico-legal order should be ideologically neutral. It would take a very long development to make precise what I mean by neutrality, so I will not endeavor to do that here, but the basic idea is that as far as possible individuals should not be forced to support political causes they disagree with. One of the many reasons why someone might think that such a norm is justified on consequentialist grounds is that people are fallible and, as a result, the probability that democracy will produce optimal results is maximized if the politico-legal order is as ideologically neutral as possible. Of course, on a consequentialist picture, this desideratum must be balanced with others, so a consequentialist could also object to this norm. It’s clearly a very complicated debate, which again we aren’t going to solve here, but what is certain is that the claims I made in my post are only surprising from a consequentialist if you assume a very unsophisticated form of consequentialism. I used that language in my post because I didn’t write it for a philosophical audience and, moreover, I suspect that most people, insofar as they have a view about that, are not consequentialist. Of course, I think they are wrong, but I also recognize that I could be mistaken about that.

      I should further explain why I’ve been putting scare quotes around “force” and its variants. It’s because I’m not at all convinced that paying union dues as a condition of one’s employment is “force” in any coherent sense. Suppose my employer decides to reduce my wages, and donates the money saved to the NRA. Have I been “forced” to donate to the NRA? Conservatives and libertarians say no, and the reason invariably given is that I am part of a voluntary contract with my employer, and if I don’t like it I can always quit my job. The typical left response is this argument is to question how “voluntary” such a relationship is in an economic system in which workers don’t own the means of production and must work for a living, or something along those lines.
      It’s not that I disagree with such responses, it’s that I think they are missing a much more obvious point: You can also quit your job if you don’t like paying union dues. If union dues are “forced”, so are wage cuts. If wage cuts aren’t “forced”, neither are union dues. You can’t have it both ways.
      The same hold for a host of other policies that reactionaries like to claim involve “force”. Are you an employer who doesn’t like being “forced” to pay your workers the minimum wage? Liquidate your business. Don’t like taxes? Quit your job and live in the forest. It’s no different from asking a worker to quit if she doesn’t like her pay or working conditions.

      I think this is the most interesting point you make, and it’s certainly a fair one, so I want to say something about it. There are at least two differences between the way in which corporations spend money on politics and the way in which unions use dues to do the same thing which I think are relevant to the point you’re making. First, despite a widespread belief to the contrary among liberals, as far as political contributions are concerned, corporations contribute to both major parties more or less evenly. This means that the political activity of corporations is more ideologically neutral than that of unions, whose political contributions overwhelmingly go to the Democratic party. (Note that I’m not saying that the political activity of corporations is ideologically neutral, which it most certainly is not for reasons that should be obvious, but only that it’s more ideologically neutral in the relevant sense than the political activity of unions.) Thus, given how corporations and unions spend money on politics, the fact that corporations spend money on politics does not violate the norm I briefly discussed above as much as the fact that unions do the same thing by using dues they collect from people who disagree with them.

      Moreover, and perhaps more importantly, there is no way of preventing corporations from spending money on politics and violating the norm of neutrality I was talking about without undermining other norms/institutions which I think are justified on consequentialist grounds, whereas it’s very easy to prevent unions from collecting dues from workers who don’t want to be represented by them without doing so. In particular, you could impose more drastic legal restrictions on how corporations can spend money on politics than what currently exists in the US, but they would just find other, typically much less transparent ways to exert their influence on politics, as I can guarantee you they do in countries such as France where the laws about money and politics are stricter. We could discuss at length various possible legal schemes to reduce the ways in which corporations can influence politics, but I think that, in the end, none of them could achieve significant results without requiring that we give up some essential aspects of the liberal and capitalist institutions which I think are justified on consequentialist grounds. (Note that I’m not saying that there are no legal changes we could implement to reduce the extent to which the political activity of corporations violates the norm of neutrality I briefly discussed above, but I don’t think any such changes could make a substantial difference without violating other norms and institutions which are more important.)

      Of course, none of that is obvious and there are various points on which you could reasonably disagree, so I’m not claiming that it’s the end of the story. All I have tried to do here is argue that it wasn’t obvious, even from a consequentialist point of view, that there was no difference between the way in which corporations spend money on politics and the way in which unions use dues they collect even from unwilling people to do the same thing. I have done so by trying to sketch how a consequentialist could argue that the former is morally acceptable but not the former. But it would take a very long and technical post to make the case in a satisfactory manner and I wasn’t going to do that in my post. Nor do I intend to do that anytime soon because I’m totally swamped with work and, to the extent that I have time to write on the blog, there are other things I want to talk about. However, it’s certainly true that, unlike the fact that CGSU argues that graduate students will benefit from voting in favor of unionization in a very misleading way (which is absolutely indisputable) and, to a lesser extent, the claim that teachers unions have a negative effect on education (which I think is certainly true if we take into account the issue of costs), this is a very complicated philosophical question and I think our exchange should have made that clear.

      1. I agree that both liberals and conservatives probably overestimate how much education can do to reduce poverty, but there is no doubt that it helps.

        The question isn’t whether it helps at all (I didn’t say it didn’t, just that it had “little to do” with poverty). The question is whether it helps enough to justify focusing efforts on “educational reform” (assuming that the “reforms” in question actually will improve education) is worth the opportunity cost of doing other stuff, like increasing transfers. If your educational”reform” are going to undermine unions and thereby make transfers less politically feasible, then there is going to be a trade-off between transfers and educational “reform”. Even assuming that neoliberal education policies improve education (they don’t) and that education matters substantially to poverty (it doesn’t), it could still be the case that those educational policies create more poverty relative to the alternatives, if the alternatives, like transfers, are better at cutting poverty (they are), and neoliberal educational policies are make those alternatives more politically difficult to implement (they do).

        Basically, it points out that even though the general level of education has significantly increased between 1991 and 2014

        Not just overall, but among the poor specifically.

        poverty has increased during the same period. But of course this doesn’t show that education doesn’t help to reduce poverty, since plenty of other things happened during that period that could have affected the prevalence of poverty, such as largest recession in almost a century, to cite only the most obvious.

        It certainly would be nice to know the counter actual of how much poverty would have fallen, if at all, during this period, in the absence of the recession. But we can have the next best thing by looking at the poverty trends before the recession, from 1991-2007, during which the educational composition of the poor also monotonically improved.

        As you can see, poverty went from 14.2% in 1991 to 12.5% in 2007, a fall of just 1.7 percentage points or 12 percent. But even that is doubtfully due to improved education, since the education of the poor improved continuously through these years, yet almost all of the reduction took place in the 90s, which was an economically prosperous time for the country. There was a very modest recession (it may not have even technically been a recession) in the early 2000s that caused a slight uptick in poverty, but even during the good years of 2003-2007 poverty essentially flatlined (2003: 12.5; 2007: 12.5).

        The reduction is larger if you look at the anchored supplemental poverty measure, but even there, the entirety of the poverty reduction took place in the 90s, and the poverty rate flatlined in 2000s (Figure 2, p. 12).

        As I’m sure you know, unions have become significantly less strong in practically every developed countries in the past few decades, yet in virtually all of them public social expenditures have increased considerably during the same period. At the very least, this shows that, despite what you say, high union density is not a necessary condition for poverty reduction

        You’re here guilty of the same “statistical incompetence” you accused Bruenig of, except it’s worse here, since Bruenig includes many years before the recession, whereas you don’t. In addition to and related to the statistical confusion, there’s also a conceptual confusion about what it means to “increase social expenditures”. The document you link to compares social expenditures in 2007 (before the recession) to social expenditures in 2014 (after the recession). Of course spending on transfers increased after the recession: more people needed them and were eligible for them! That doesn’t mean that transfers per person who received them have been getting more generous, or that eligibility has been expanding. If anything, they’ve been getting less generous, and eligibility has been restricted in OECD countries.

        The document itself makes this point on p. 2.

        In an economic downturn, social spending-to-GDP ratios usually increase as public spending goes up to address greater need for social support, while simultaneously economic growth falters (GDP as in the denominator). At the onset of the Great Recession both these features contributed to a rapid increase in public social spending-to-GDP ratios on average across the OECD

        But of course that doesn’t mean that the welfare programs have gotten more generous (which is what I was talking about), it just means that more people qualify for the previously existing programs. Those previously existing programs were won through the organized political power of the working class, and as long as that power continues to decline, those programs won’t get more generous or otherwise improve, though more people might receive them if the economy slows down. But of course, a mere expansion of the welfare roles, without an increase in welfare generosity per recipient or an expansion of eligibility, won’t do much besides hold down poverty somewhat in the face of a severe recession.

        If you doubt that labor is necessary for transfer generosit, take a look at the graph on p. 2 of the very document you linked to. It shows that the social spending to GDP ratio in the OECD and the US stayed essentially flat from the end of the early 90s recession to the beginning of the late 2000s recession. This was during a period of severe setbacks to organized labor, during which the political parties traditionally linked to labor retreated far from aggressive redistribution. If you think these phenomena aren’t linked, you’re kidding yourself.

        I guess one could argue that political lobbying by teachers unions doesn’t have a negative effect on education, but it’s hard to see how anyone who is familiar with the political activities of teachers unions could seriously defend that view.

        It’s part of my argument that the lobbying efforts of unions in general are on the whole good, not necessarily with regard to education but certainly with regard to poverty and inequality. With regard to education, teachers unions, their political efforts, as far as educational policy goes, are devoted primarily to blocking vouchers and charter schools. I’ve never seen much evidence that either of these policies have much of an impact on educational outcomes for either good or ill. Therefore, they should be opposed because they are a waste of time, and also because they hurt an important partner in a progressive political coalition: teachers unions.

        the literature suggests that they are slightly negative but also that teachers unions have a massive negative effect on costs, which means that in terms of cost-efficiency the literature actually suggests that teachers unions have a very substantial negative effect on educational outcomes even if we ignore the effect of their political activities.

        First, the literature suggests the direct educational effects are slightly negative or insignificant. Second, part of my argument is that teachers unions are cost effective. Not necessarily with regard to education specifically, but the benefits unionized teachers get are a necessary part of building and holding together a broad social democratic coalition that will include enough of the educated middle classes to survive. Middle class professionals like teachers won’t support the high middle class tax rates necessary to sustain social democracy unless they get something out of it. Support for unions, even and especially those representing relatively well off workers, is part of the social democratic bargain. It’s the same reason why universal programs are to be preferred to means-tested ones, and why things like tuition-free college are necessary even though the benefits mainly go to the top and the middle.

        As I’m sure you know, consequentialists can and many of them do support norms even if their application can be suboptimal in particular cases, if they think that on the whole they are likely to have good consequences. One important norm, in my opinion, is that as far as possible the politico-legal order should be ideologically neutral…. the basic idea is that as far as possible individuals should not be forced to support political causes they disagree with.

        I have doubts about the coherence of rule-consequentialism, but leave those aside. It seems at the very least that there is going to be a conflict of rules. Suppose one of your rules is “reduce poverty”, and another is “don’t force people to support political groups they disagree with”, where the latter is interpreted to prohibit unions from “forcibly” collecting dues from their members. Suppose also, as I’ve been arguing, that there is a pragmatic and political conflict between the first and the so-interpreted second rule. Then you’re going to have to choose. For what it’s worth, I’ll take fewer impoverished kids over some weird procedural obsession with “political neutrality” any day.

        But I’m not sure that making people pay union dues as a condition of their employment is even a violation of the political neutrality norm, as you’ve articulated it. That norm, as you describe it, prohibits forcing people to support causes they don’t like. But why is paying union dues as a condition of employment forced any more than accepting a wage cut as a condition of employment forced? You attempt to answer this question later on, but you fail to do so, because you go off on an irrelevant tangent about the pragmatic difference between preventing political donations by unions and preventing political donations by firms. But I wasn’t interested in the “donation” part of the scenario, I was interested in the “force” part!

        I will conced for the sake of argument that it’s impossible to stop capitalists from making political donations, whereas it’s possible to stop unions from doing so. Are both kinds of donation forced? Are they both unforced? Is one forced and the other not?

        Why is a worker who has to pay union dues that go to a political cause as a condition of employment coerced, but a worker who has to accept wage cuts that go to a cause as a condition of employment not? Why for that matter, is a shareholder who has to accept political donations as a condition of ownership of stock in the company not forced?

        In other words, I’m not interested in whether the norm of political neutrality can be consistently enforced across the board. I’m interested in whether the cases I’ve been talking about are violations of that norm to begin with, or whether the norm is even coherent.

        1. The reduction is larger if you look at the anchored supplemental poverty measure, but even there, the entirety of the poverty reduction took place in the 90s, and the poverty rate flatlined in 2000s (Figure 2, p. 12).

          Nevermind, this is irrelevant. As Bruenig points out in a note at the end of the post, only the official poverty rate is relevant here. The supplemental poverty definition differs from the official one only in that taxes and various kinds of transfers are included. But the causal mechanism by which education supposedly reduces poverty is through increasing market income.

          So, while I conceded that you see a larger drop when looking at the anchored supplemental poverty rate than when looking at the official poverty rate, I now revoke that concession.

        2. The question isn’t whether it helps at all (I didn’t say it didn’t, just that it had “little to do” with poverty). The question is whether it helps enough to justify focusing efforts on “educational reform” (assuming that the “reforms” in question actually will improve education) is worth the opportunity cost of doing other stuff, like increasing transfers. If your educational”reform” are going to undermine unions and thereby make transfers less politically feasible, then there is going to be a trade-off between transfers and educational “reform”. Even assuming that neoliberal education policies improve education (they don’t) and that education matters substantially to poverty (it doesn’t), it could still be the case that those educational policies create more poverty relative to the alternatives, if the alternatives, like transfers, are better at cutting poverty (they are), and neoliberal educational policies are make those alternatives more politically difficult to implement (they do).

          That’s a lot of ifs. As we say in France, if my aunt had a pair of balls, she’d be my uncle. But she doesn’t and she’s not. It’s obviously true that, if the direct effect of educational reform was to reduce poverty by x, if the indirect effect through undermining unions was to increase poverty by y, if x is less than y and if nothing else changes, then educational reform in question would increase poverty. But stating that kind of platitudes does nothing to support the claim that we couldn’t significantly reduce poverty by pursuing educational reform. In fact, as I argue below, your argument is totally unconvincing, because it rests on wildly implausible assumptions.

          It certainly would be nice to know the counter actual of how much poverty would have fallen, if at all, during this period, in the absence of the recession. But we can have the next best thing by looking at the poverty trends before the recession, from 1991-2007, during which the educational composition of the poor also monotonically improved.

          As you can see, poverty went from 14.2% in 1991 to 12.5% in 2007, a fall of just 1.7 percentage points or 12 percent. But even that is doubtfully due to improved education, since the education of the poor improved continuously through these years, yet almost all of the reduction took place in the 90s, which was an economically prosperous time for the country. There was a very modest recession (it may not have even technically been a recession) in the early 2000s that caused a slight uptick in poverty, but even during the good years of 2003-2007 poverty essentially flatlined (2003: 12.5; 2007: 12.5).

          The reduction is larger if you look at the anchored supplemental poverty measure, but even there, the entirety of the poverty reduction took place in the 90s, and the poverty rate flatlined in 2000s (Figure 2, p. 12).

          I don’t want to get into a debate about whether investment in education did or did not reduce poverty since 1990. I just pointed out that Bruenig’s post did not even begin to show that educational reform can’t make a meaningful contribution to the reduction of poverty, which is absolutely indisputable. I just cited one reason why it doesn’t show that and doesn’t even make it plausible in any meaningful way, but that’s only because I didn’t think it was necessary to insist on that point, since it’s so obviously true. Similarly, what you say above does little to improve on Bruenig’s post, as you can surely realize on your own. Causal inference is more complicated than what Bruenig does in his post or what you’re doing above. I honestly don’t even understand why we’re having this conversation, since the claim I’m making is so weak claim that it should be totally uncontroversial. I’m not even saying that you’re wrong about what happened in the US since 1990, I honestly have no idea, I’m just pointing out that you haven’t even come close to showing that you’re right.

          You’re here guilty of the same “statistical incompetence” you accused Bruenig of, except it’s worse here, since Bruenig includes many years before the recession, whereas you don’t. In addition to and related to the statistical confusion, there’s also a conceptual confusion about what it means to “increase social expenditures”. The document you link to compares social expenditures in 2007 (before the recession) to social expenditures in 2014 (after the recession). Of course spending on transfers increased after the recession: more people needed them and were eligible for them! That doesn’t mean that transfers per person who received them have been getting more generous, or that eligibility has been expanding. If anything, they’ve been getting less generous, and eligibility has been restricted in OECD countries.

          The document itself makes this point on p. 2.

          In an economic downturn, social spending-to-GDP ratios usually increase as public spending goes up to address greater need for social support, while simultaneously economic growth falters (GDP as in the denominator). At the onset of the Great Recession both these features contributed to a rapid increase in public social spending-to-GDP ratios on average across the OECD

          But of course that doesn’t mean that the welfare programs have gotten more generous (which is what I was talking about), it just means that more people qualify for the previously existing programs. Those previously existing programs were won through the organized political power of the working class, and as long as that power continues to decline, those programs won’t get more generous or otherwise improve, though more people might receive them if the economy slows down. But of course, a mere expansion of the welfare roles, without an increase in welfare generosity per recipient or an expansion of eligibility, won’t do much besides hold down poverty somewhat in the face of a severe recession.

          If you doubt that labor is necessary for transfer generosit, take a look at the graph on p. 2 of the very document you linked to. It shows that the social spending to GDP ratio in the OECD and the US stayed essentially flat from the end of the early 90s recession to the beginning of the late 2000s recession. This was during a period of severe setbacks to organized labor, during which the political parties traditionally linked to labor retreated far from aggressive redistribution. If you think these phenomena aren’t linked, you’re kidding yourself.

          No, I’m not guilty of statistical incompetence, you just misunderstood my argument. First, I wasn’t referring to the figure on p. 1 but to that on p. 2, as the fact that I used the expression « in the past few decades » should have made clear. (What sense would it make for me to use that expression if I wanted you to look at a graph that only look at a trend between 2007 and 2014.) This figure shows very clearly that public social expenditures in developed countries have massively increased since the 1960’s. Second, unlike Bruenig, I wasn’t making a controversial causal claim, but a very weak one that should be totally uncontroversial. Namely, I was denying your claim, which is extremely strong and clearly false, that high union density was a necessary condition for poverty reduction through social transfers. The fact that public social expenditure has exploded in many countries where, during the same period, union density has been significantly reduced is quite sufficient to make your extremely strong claim wildly implausible. In fact, anyone who has some familiarity with European politics knows that it’s false.

          For instance, take the case of France, which I know intimately. In France, union density has massively diminished since the 1960’s (although it was never as high as in Nordic countries), yet during the same period public social expenditure has increased even more dramatically. Moreover, I can guarantee you that it’s not just or even primarily because of aging and impoverishment, it’s because the welfare state has expanded. Today, France has one of the most generous welfare states in the world, yet one of the lowest union density in the world. It’s true that, at the same time, it has one of the highest rates of union coverage (i. e. the proportion of workers who are covered by a collective bargaining agreement) in the world because the law essentially dictates that pretty much every worker is covered by a collective bargaining agreement at the level of industry branches. (But note that, unlike in the US, union dues are voluntary. As you can see, it doesn’t prevent France from having a very strong welfare state, far from it.) But this has absolutely nothing to do with how developed the welfare state is because, mostly for historical reasons (the trade union tradition in France is heavily influenced by anarcho-syndicalism and, except for the CGT with the Communist Party after 1945, never got involved with political parties, unlike in the UK for instance), unions don’t play a significant political role. (Well, this would need to be qualified since unions have a lot of power in other ways, but the important point is that they had little to do with the expansion of the welfare state since the 1960’s because of the way in which the political system works in France.)

          Now, before you accuse me of engaging in the same kind of unsophisticated causal inference I was accusing Bruenig of doing, I’m just going to repeat that, unlike you and Bruenig, insofar as I’m making a causal claim, it’s a very weak one. In particular, I’m not saying that union density has nothing to do with how generous the welfare state is, I’m just denying your very strong claim that high union density was politically necessary for poverty reduction through a strong welfare state, which is indisputably false. It’s also false that, unless you allow unions to collect dues from people even if they don’t want to, you can’t have a strong welfare state. France, I should add, is hardly the only counter-example, since it’s also true of several other European countries with large welfare states. Now, you can say that union density has a tendency to make a strong welfare state more likely, which is probably true, but that’s a much weaker claim than the one you were making and one that makes it significantly harder for you to say the kind of things you want to say. In fact, a look at the evidence and a back-of-the-envelope calculation makes your argument that educational reforms would increase poverty by decreasing union membership wildly implausible, as I will now argue.

          According to the Bureau of Labor, 14.6 million workers belonged to a union in 2016, which corresponds to a union density of 10.6%. According to the Union Membership and Coverage Database, which is based on data from the CPS, 2,683,436 teachers belonged to a union that same year. (This figure is obtained by adding the number of union members for categories 2200, 2300, 2310, 2320, 2330 and 2340 in the table. It excludes teachers who aren’t members of a union but are nevertheless covered by a collective bargaining agreement, some of whom probably pay dues, but again there shouldn’t be too many of them and, even if we assume that the entire difference should have been included, I could show that it doesn’t affect my calculation in any meaningful way. I have to use the number of teachers who are members of unions and not also include those who are covered by a collective bargaining agreement because the study I’m going to use is about the relationship between public social expenditure and union density, not union coverage.)

          I looked for a something about the relationship between union density and public social expenditure and found this study from 2016. It distinguishes between what it calls coordinated market economies such as Sweden and liberal market economies such as the US and only finds a relationship between union density and public social expenditure in the former. Thus, the data analyzed in this study suggests that, in the US, union density has no statistically significant effect on public social expenditure. But let’s assume that it does, and use the strongest effect of union density on public social expenditure found in any of the models used in that study, which is approximately 0.12. In other words, if we make this gratuitous assumption, public social expenditure as a percentage of GDP increases by 0.12 point for every percentage point increase in union density.

          Now suppose that educational reforms would cut union membership in half among teachers. In fact, none of the reforms that are regularly proposed would even come close to doing that, but I want to grant you as much as I can again. Since teachers constitute only 18% of union members in the US, this would bring the union density to approximately 9.6%, i. e. 1 percentage point less than the current rate. Thus, according to the study I discussed above and even if we make a gratuitous assumption to help your case, the educational reforms in question would only decrease public social expenditure by 0.12 point. Even if we assume that, for every point percentage increase in public social expenditure, poverty is also reduced by 1 percentage point (which is absurd), it follows that the kind of educational reforms conservatives like would only increase the poverty rate by 0.12 point. Now, even if you think that people overestimate the effect educational reforms could have on poverty (I agree with that), if you think that educational reforms couldn’t possibly reduce poverty by more than 0.12 point, then you are completely delusional.

          Moreover, let’s not forget that in order to arrive at this figure, I made totally implausible assumptions just to help your case as much as possible. In fact, the evidence I examined suggests that the decrease in union density resulting from educational reforms would have no effect whatsoever on poverty, which is what my back-of-the-envelope calculation would show if we used more realistic/better supported by the evidence assumptions. Now, don’t get me wrong, this back-of-the-envelope calculation is sloppy in more ways than I care to explain, but I think it’s quite sufficient to show how implausible the assumptions you have to make about the effect of teachers union on poverty in order to argue in the way you do that any educational reforms that reduce the membership of teachers unions would be harmful to the poor.

          It’s part of my argument that the lobbying efforts of unions in general are on the whole good, not necessarily with regard to education but certainly with regard to poverty and inequality. With regard to education, teachers unions, their political efforts, as far as educational policy goes, are devoted primarily to blocking vouchers and charter schools. I’ve never seen much evidence that either of these policies have much of an impact on educational outcomes for either good or ill. Therefore, they should be opposed because they are a waste of time, and also because they hurt an important partner in a progressive political coalition: teachers unions.

          I have already showed that your argument about the effect of teachers unions on poverty was totally implausible. You don’t even need the back-of-the-envelope calculation I did above to see that. Although teachers unions are some of the groups that spend the largest amounts of money in lobbying, they still spend only a very small fraction of the money that is spent on lobbying each year. To be sure, they have a decisive influence on education policy, but that’s for the same reason that AIPAC has a lot of influence on the US foreign policy in the Middle East, namely because they are single-issue lobbying groups. (Since I mention their influence on education policy, I should note that one reason we don’t have as much evidence as we’d like on the effect of educational reforms is that teachers unions systematically block even the experimentation of that kind of reforms, because they are first and foremost concerned with the protection of their rent at the expense of students and taxpayers.) The notion that teachers unions specifically play a significant role on the promotion of social-democratic policies that reduce poverty is ludicrous and can’t be defended without attributing magical powers to them.

          I think the evidence clearly supports the hypothesis that this kind of reforms would have a modestly positive effect on educational outcomes, which becomes quite large if you take into account the issue of costs, as indeed you should. (I think both conservatives and liberals, although in opposite directions, overestimate how substantial the effect would be. I think reducing socio-economic segregation has a much larger effect and it’s actually one of the few issues on which massive state intervention is justified because it would be very effective, although it wouldn’t take the form of cash transfers. But the effect on costs would be substantial and, as I noted before, liberals systematically ignore the issue of costs when they discuss the evidence about the effects of that kind of educational reforms.) It’s also worth noting, since we have been talking about the effect of education on poverty, that the evidence suggests that the positive effect of that kind of educational reforms would be strongest on poor students. But I don’t want to get into that debate, because it’s complicated and deserves a whole separate discussion, so I’m just going to remind that I have already dug up evidence to support the corresponding claim about the direct effect of teachers unions on educational outcomes, which I discuss more below.

          First, the literature suggests the direct educational effects are slightly negative or insignificant.

          This doesn’t really matter. Even assuming the direct educational effects are zero, the direct effect on costs is very large, so it’s still the case that teachers unions have a large negative effect on the effectiveness of expenditure in public education. Moreover, as the review I cited points out, the literature also suggests that the direct educational effect is strongest at both tails of the performance distribution, hence presumably on poor students. Again, since we’ve been talking about the effect of education on poverty, this is worth noting.

          Second, part of my argument is that teachers unions are cost effective. Not necessarily with regard to education specifically, but the benefits unionized teachers get are a necessary part of building and holding together a broad social democratic coalition that will include enough of the educated middle classes to survive. Middle class professionals like teachers won’t support the high middle class tax rates necessary to sustain social democracy unless they get something out of it. Support for unions, even and especially those representing relatively well off workers, is part of the social democratic bargain. It’s the same reason why universal programs are to be preferred to means-tested ones, and why things like tuition-free college are necessary even though the benefits mainly go to the top and the middle.

          But, just like your argument that teachers unions have a significant effect on poverty, this argument that by reducing the number of unionized teachers educational reforms would significantly damaged the coalition that supports the American welfare state is completely implausible. According to the Union Membership and Coverage Database, which I cited above, there are only 2,944,383 teachers in the US who are covered by a collective bargaining agreement, which represent about 2% of the workforce. Even if educational reforms cut the number of unionized teachers in half and the teachers who were no longer covered were paid 20% less, which is preposterous, this would not have any significant political or fiscal effect. Indeed, even if we just do a static analysis, it would just mean that 1% of the workforce would be paid 20% less. This would barely affect the tax base used to finance the welfare state, especially given how progressive the US tax system is. The notion that it would also have a significant effect on the political behavior of teachers is similarly overwhelmingly implausible. In practice, for the vast majority of the teachers concerned, it would only be a loss of a few thousands dollars a year and that’s under the crazy assumption that being covered by a collective bargaining agreement increases the wages of teachers by 20%. (To be clear, I’m not saying that loss of income wouldn’t be significant for the people concerned, but only that there is no reason it would have the kind of massive effect on their political behavior your argument requires.) So the notion that reforms that improve educational outcomes while reducing the union coverage among teachers would not be cost effective all things considered because it would have the kind of dramatic political effect you’re assuming is essentially voodoo magic. Like my back-of-the-envelope calculation above, many of the things I just said are very sloppy, but they’re not sloppy in a way that has any chance of reversing my conclusion.

          I have doubts about the coherence of rule-consequentialism, but leave those aside. It seems at the very least that there is going to be a conflict of rules. Suppose one of your rules is “reduce poverty”, and another is “don’t force people to support political groups they disagree with”, where the latter is interpreted to prohibit unions from “forcibly” collecting dues from their members. Suppose also, as I’ve been arguing, that there is a pragmatic and political conflict between the first and the so-interpreted second rule. Then you’re going to have to choose. For what it’s worth, I’ll take fewer impoverished kids over some weird procedural obsession with “political neutrality” any day.

          One of my arguments in favor of the norm of political neutrality is that, given how complex socio-economic problems are, one should recognize that one’s judgments about them are highly fallible and, in the long-run, having such a norm maximizes the probability that optimal policies will be adopted because it will limit the extent to which the debate is structurally biased. On the other hand, I don’t want to say that it can’t be overruled by other considerations, since for instance I argued that it’s permissible to force people to pay taxes because we need a government that can do that. So you may be right about this, but I would have to think more about it before I make up my mind, and right now I can’t because my brain is fried since I have only slept 4 hours last night and it’s late. But even if I grant you this point, I have shown above that your case that educational reforms which reduce union density/coverage would have a negative effect on poverty is utterly implausible, so it wouldn’t matter all that much.

          (By the way, note that in arguing against your claim that teachers unions have overall a positive effect on poverty, I have conceded for the sake of the argument that except for teachers unions social democratic policies are the best way to solve poverty. Of course, I don’t actually believe that, it’s just that I can grant you that and still show that your argument against educational reforms that weakens teachers unions is unconvincing. I also don’t want to argue against social democracy here, since I think it’s best to wait until I reply to your latest posts about Sanadaji’s book, which I plan to do in a post as soon as I have enough time.)

          But I’m not sure that making people pay union dues as a condition of their employment is even a violation of the political neutrality norm, as you’ve articulated it. That norm, as you describe it, prohibits forcing people to support causes they don’t like. But why is paying union dues as a condition of employment forced any more than accepting a wage cut as a condition of employment forced? You attempt to answer this question later on, but you fail to do so, because you go off on an irrelevant tangent about the pragmatic difference between preventing political donations by unions and preventing political donations by firms. But I wasn’t interested in the “donation” part of the scenario, I was interested in the “force” part!

          I will conced for the sake of argument that it’s impossible to stop capitalists from making political donations, whereas it’s possible to stop unions from doing so. Are both kinds of donation forced? Are they both unforced? Is one forced and the other not?

          Why is a worker who has to pay union dues that go to a political cause as a condition of employment coerced, but a worker who has to accept wage cuts that go to a cause as a condition of employment not? Why for that matter, is a shareholder who has to accept political donations as a condition of ownership of stock in the company not forced?

          In other words, I’m not interested in whether the norm of political neutrality can be consistently enforced across the board. I’m interested in whether the cases I’ve been talking about are violations of that norm to begin with, or whether the norm is even coherent.

          In my opinion, there is a clear sense in which the workers who are covered by a collective bargaining agreement but don’t like the union are forced to pay dues to it. Similarly, if they don’t like the political activity of the company they work for, there is a sense in which they’re also forced to contribute to it anyway. In both cases, their labor is part of what allows a corporation/union to engage in political activity, even though they may disapprove of it. They may get more income because of that, but it doesn’t mean that there isn’t an important sense in which they’re forced to contribute to political activities they don’t like, because in practice it may be very difficult for them to avoid that by quitting their job. This, by the way, is exactly the kind of thing leftists are prone to say as long as the topic of unions doesn’t come up and, if you ask me, they are completely right. (They just tend to draw the wrong conclusions from that observation, but that’s another problem.)

          In the same way, we can say that, in many cases, shareholders are forced to contribute with their capital to the political activity of the company they own if they disapprove of that activity, because often shareholders have little to no influence on the management of the company they own and it may be difficult in practice for them to sell their shares. In a similar way, someone who owns a company and, as a result, has to pay a lot of taxes even though he doesn’t like how they are used is clearly forced by the state to pay taxes.

          Of course, in all of those cases, there is also a sense in which people in the kind of situations I described are not forced to contribute to political activities they disapprove, since they could always quit, give up their shares, dismantle their company, etc., but I agree with leftists — although I suspect that, like conservatives, most of them would not consistently apply the concept of force I think is relevant — that it’s not the important sense. It’s the sense in which someone who puts a gun to my face and asks me for money doesn’t force me to give him money, since I could just have chosen to die.

          In my view, the fact that workers in a unionized firm are forced to pay dues to the union even if they disapprove of it is a defeasible reason against the legal framework that allow it, just like the fact that they are forced to contribute to the political activities in which the company they work force engage is a defeasible reason against the legal framework that allows it. It’s just that I think that, in the former case but not in the latter, that reason is defeated by other considerations in the consequentialist framework I’m assuming. Similarly, as I explain in my post, I think the fact that the existence of a government that forces people to pay taxes which it uses in ways that some of them disapprove is a defeasible reason against the existence of government, but I also think it’s defeated by other considerations. However, in view of the fact that I conceded above that the consequentialist framework I’m assuming may not work in the way I want, what I have just said may not matter in the end.

      2. “First, despite a widespread belief to the contrary among liberals, as far as political contributions are concerned, corporations contribute to both major parties more or less evenly. ”

        Note that this appears to be false:

        https://www.aol.com/article/2015/10/26/one-chart-shows-exactly-how-much-america-s-biggest-companies-giv/21253905/

        Based on the publicly available data (which is limited, because a lot of corporate political spending is funneled through dodgy and secretive non-profits), corporations consistently donate more to Republicans than to Democrats. They’re happy to curry favor with anyone in power, so they end up giving a lot to both sides, but overall there’s still a marked preference for Republicans.

        1. The article you quote says that “Republicans and conservatives still have a slight edge overall”, which I think is consistent with what I said. I think I once read that, based on the data from the Center for Responsive Politics (the data used by the article you shared), the partisan split of contributions by corporations was something like 55/45 in favor of Republicans, but I don’t have time to look for the article where I read that again. Anyway, the important point is that, even if Republicans have a slight edge with respect to contributions from corporations, it’s still nothing like the totally one-sided nature of contributions from unions. I should also say that I’m strongly in favor of limiting the possibilities for corporations to influence politics through lobbying and contributions, which results in the kind of crony capitalism that both libertarians and progressives constantly blast, but I think the only practical way of doing that without subverting the basic principles underlying the liberal and capitalist order is to reduce the role of government in the economy. Obviously, progressives disagree with that, but that’s a debate for another time.

  4. 1. “The problem is that, if the graduate students vote in favor of unionization, even people like me who strongly disapprove of the AFT and NYSUT will be forced to pay dues that will go to these organizations and will be used to support policies they consider deeply immoral.”

    It’s not clear to me that this is true. Arguably, it is Cornell that is really forced (or “forced”) to pay the unions, not the dissenting graduate students. If collective bargaining adds (say) $200 to each student’s stipend, but the union charges a $100 fee for its services, it looks to me like the fee is only nominally assessed against the student and is really going straight from the university to the union. There is not, after all, any possible state of affairs where you actually get your hands on that extra $100.

    By way of analogy, suppose that, having appointed myself your agent without your consent, I approach your neighbor and threaten that you’ll leave a ghastly Christmas display up on your lawn all winter long if he does not immediately fork over $200. He gives me the money; I take a 50% cut and leave the rest in your mailbox. Maybe this is some kind of extortion, but if so, it’s your neighbor’s money I’m extorting, not yours. If I were legitimately acting as your agent, things might be different, but because I’m acting without your approval (just like the union is, on your view, acting without the approval of the dissenting grad students), the transaction doesn’t genuinely involve you at all.

    2. “I imagine that some people will reply that, in a democracy, even people who didn’t vote for the person who won the election have to pay taxes that are used in a way they sometimes find morally repugnant.”

    Perhaps more to the point, in a capitalist society, lots of people work for firms which put the revenue generated by the individual’s labor to uses he or she finds morally repugnant. AT&T offers you an employment contract; you have the choice of accepting, in which case they will donate some of the proceeds of your labor to the RNC, or refusing, in which case you will be asked to find work elsewhere. This seems to me pretty similar to what the union would do at Cornell– henceforth, when a graduate student is offered a contract, they have the choice of accepting, which means the union takes a cut, or refusing, in which case they will be asked to pursue their studies elsewhere. What do you think is the difference, here? Does it rely on a contentious notion of “force” that is supposed to apply in the union case but not in the corporate case?

    1. I’m sorry I didn’t reply to you, but I’m really swamped with work. Matt made some very similar points above, which we have been discussing. Maybe some of it will be of interest to you.

  5. I wanted to comment that the two anecdotes you gave on the evil of AFT/NYSUT were utterly unconvincing. I would’ve hoped that if you were going to pick two anecdotes from an entire book of how bad they were, they’d at least have been shocking anecdotes. But taking two minutes to think about how AFT/NYSUT would respond to these accusations, I can already mount decent defenses.

    1. It’s reasonable for a worker, or a union of workers, to not want their contract unilaterally changed by their employers. If your landlord decided, “Oh hey you’re not using this parking space, let me give it to someone else”, would you be in the wrong for saying ,”Hey I wasn’t using it, but it was *mine* according to the terms of our contract. How about you reduce my rent a bit if you’re going to give the parking spot to someone else?”

    2. It appears that the ALST actually *was* discriminatory. It is possible to interpret NYSUT’s comment about it being redundant as saying that obviously a “literacy” test would be redundant, since what bachelor’s degree holder who’s applying to be a teacher wouldn’t be literate?? So your pointing out that not all applicants passed the ALST is only consistent with their claim that the ALST was discriminatory with respect to race/ethnicity (and associated cultural/linguistic differences). The article you linked to even points out the immense disparity in pass rates between the ethnicities. (In case you’re not aware, there are dialectal differences in English used among different ethnic groups. Eg https://en.wikipedia.org/wiki/African_American_Vernacular_English . In fact a lot of harm has been done to African American children based on the premise that their English was “bad” English: https://www.theatlantic.com/past/docs/issues/95sep/ets/labo.htm .)

    1. 1. I think it’s entirely reasonable to insist that your employer cannot unilaterally change your terms of employment, but if this had really been the reason why the union fought against attempts by the district to get rid of the rider unilaterally and if it had really been opposed to the rider as it claimed, it could have simply given its agreement to end the rider… Indeed, as I explain in my post, at one point the district asks the union to agree to suspend the rider for a year so that the district would not have to lay people off and it refused even though it wouldn’t have involved any other change and was just a temporary measure. Perhaps you don’t think that’s shocking, but I do and I bet the people who had to be let go because the school district had a $10 million deficit also do…

      2. The fact that minorities had a significantly lower rate of success on the literacy test is at best extremely weak evidence that it was discriminatory. In general, a disparity is poor evidence of discrimination, because there are many between-group differences that have nothing to do with discrimination. For instance, women attend university at a significantly higher rate than men in the US, but presumably this has nothing to do with discrimination. The fact that minority applicants — not just African-Americans but also Hispanics and Asians — had a significantly lower success rate than white applicants is probably explained by the fact that they are more likely to be poor and to have parents who didn’t speak English at home. What they need is good teachers at school to compensate for that, but the decision to allow people who can’t even pass basic literacy test to teach is clearly not going to help them. Indeed, the same federal judge who had declared 2 previous versions of the test discriminatory ruled that the version that was just dropped was not discriminatory, as you can see in some of the articles I linked to in my post. Indeed, if you have a look at sample questions (https://www.nystce.nesinc.com/TestView.aspx?f=HTML_FRAG/NY202_PrepMaterials.html), you will see that it just tests basic literacy skills. This also shows that, just because the applicants have a Bachelor, it doesn’t follow that the test is redundant. The fact that many applicants with a Bachelor nevertheless fail the test just shows that many people with a Bachelor lack basic literacy skills. Indeed, I don’t think there is anyone who has actually had a look at the test and has kids who could honestly say that they wouldn’t mind if their kids were going to have teachers that can’t even pass it, but I guess they don’t mind as long as it’s the kids of poor people.

      1. I get the impression you’re speaking from authority that you don’t have.

        1. How confident are you that 100 teachers were actually laid off? Do you know NYSUT’s explanation for why they didn’t agree to the temporary deal? Do you know if the teachers represented by NYSUT generally approved of the move? If you don’t know these things, how are you so sure that NYSUT was in the wrong here?

        2. Honestly neither of us actually know whether or not the thing unfairly discriminates based on cultural or linguistic variations between ethnic groups. I’d contend that the evidence that the test is actually useful for filtering out bad teachers is just as weak as the evidence that it’s discriminatory. In fact I just read the court ruling, and I can say there’s no evidence provided that the test is actually a good determiner of teaching skills. Are you really sure it’s more of a test of “literacy skills” rather than, say, a test of reading/writing speed (in Standard English), or a test of how much preparation/drilling someone did for the test?

  6. 1. I’m relying on the article in The Atlantic. It’s possible that what they say is false, but I have not seen any reason to think so. (In general, lots of things are possible, but it doesn’t mean they’re true.) What is certain is that, for several years, the school district of Buffalo spent millions of dollars in cosmetic surgery for teachers even though it had millions of dollars in deficit, which for years the union has refused to end so they could use it as a bargaining chip to obtain other advantages. I can’t believe that we’re actually having a debate about this. As for your last two questions, as far as I know, neither the AFT nor NYSUT ever spoke out against the BFT on this, even though the situation lasted for several years. Perhaps they did, which I don’t believe for a second, but again I have no reason to think so. If one of their affiliates was adopting racist policies, and the AFT/NYSUT didn’t criticize it, they would be harshly criticized and that would be perfectly justified. I don’t see how the case we’re discussing is different in any way that’s relevant.

    2. I have read the study guide for the literacy test, which is available online for everyone to see. Applicants have 3 hours and a half to answer 40 multiple-choice questions, write two 100-200 words focused responses and one 400-600 words extended response. The rubric also makes it pretty clear that the requirements to pass the exam are not extraordinary. In fact, I have no doubt that I would have passed this test when I was 16, even though I’m not even a native speaker of English. I don’t see how someone who is a competent speaker of English can read that study guide and think that someone who couldn’t pass this test should be able to teach. Of course, one can always be a skeptic if one absolutely wants to be, but I know that none of the people who defended that decision would ever want their kids to have a teacher who couldn’t pass this test.

    But look it doesn’t even matter. Those are just the first 2 anecdotes that crossed my mind when I wrote this post. If you don’t like them, it would only a few minutes on Google to find dozens more. I have also cited a review of the literature, which isn’t based on anecdotes and shows that the evidence suggests teachers union have a substantial negative effect on education in terms of cost-effectiveness, and that’s only when you consider their direct effect.

    1. 1. From what I can discern, the article you cited is an editorial column with a blatant anti-union slant, which interviewed the superintendent of the school system but didn’t bother asking the union for their side of the story. I basically don’t trust the author to have done due diligence on what might be technically true, but misleadingly presented facts. You can easily imagine how this might be a move deliberately crafted by the school system to make the union “look bad”.

      I acknowledge that the situation with the cosmetic rider was dysfunctional. It does seem like the BFT was not negotiating in good faith, and instead was taking advantage of a status quo preference put into place by the Triborough Amendment and the Taylor Law. It’s worth noting that the status quo preference does not discriminate between the two sides: if the union were the side upset with the status quo, would the school system have voluntarily made concessions to the union?

      But you’re laying the responsibility for the deficit at the feet of the teachers, and saying that they should not have negotiated in collective self-interest. You’re demanding that they take a cut to their contractual guarantees in a selfless act for the greater good of a balanced budget. Is there any reason why it should have been the teachers who take that cut, rather than any other adjustment to the budget? When is it okay for a group of people to act in its members’ collective self-interest, and when is it not?

      2. I am agnostic on the merits of the ALST. I’m not quick to jump to conclusions on why a test had the results it did. I know that there has been a lot of other situations where testing was found not to be indicative of actual job performance.

      > I don’t see how someone who is a competent speaker of English can read that study guide and think that someone who couldn’t pass this test should be able to teach.

      This in particular is an argument by personal incredulity. The AFT/NYSUT is not acting immorally if your disagreement with them is over factual matters (Is the ALST a good indicator of teaching effectiveness? Do children receive more benefit from having more ethnically/culturally diverse teachers over teachers who score better on the ALST?) that neither side appears to have that much evidence for either way.

      3. The review of the literature you cited shows that teachers’ unions increase the costs of education, I think primarily via an increase in teachers’ compensation. Isn’t that exactly the purpose of a union to begin with?

      4. I’m a bit jaded with this style of “argument by 1000 papercuts”. When I was discussing the issue of the Democratic Party (or Hillary Clinton)’s alleged corruption, the same argument was made. “Look, there are so many examples, just look for them, here are like 20 examples right here.” So I looked at them, and the vast majority of the examples they came up with were blatantly misinterpreted, had no regard for context, jumped to conclusions, applied double-standards, or were just outright false. There were like 2 or 3 allegations of wrongdoing out of dozens that ended up being either ambiguous or confirmed in my mind.

      So no, I’m not going to hunt down the good examples for your stance.

      Even acknowledging that unions do perverse things sometimes, is that a good enough reason to believe that unions themselves are bad? Couldn’t I make the exact same arguments you made, but instead to claim that the United States government is a “morally bankrupt” organization and therefore nobody should support it?

      1. From what I can discern, the article you cited is an editorial column with a blatant anti-union slant, which interviewed the superintendent of the school system but didn’t bother asking the union for their side of the story. I basically don’t trust the author to have done due diligence on what might be technically true, but misleadingly presented facts. You can easily imagine how this might be a move deliberately crafted by the school system to make the union “look bad”.

        I admit that it’s a possibility, but that something is a possibility doesn’t mean that it’s particularly likely and, in this case, I don’t think it is. If this claim, which has been repeated in several other places (including in the Huffington Post, which isn’t exactly known for union busting), were false, it’s so damaging to the union that I imagine they would have publicly denied it and/or that pro-union news outlets would have debunked it. But when I did my research, I didn’t find any denial or debunking, so I think it’s probably true. Perhaps the union denied it and I just missed it, in which case I would be happy to revert to agnosticism until we have more evidence, but so far I have found no reason to think it’s the case.

        It’s worth noting that the status quo preference does not discriminate between the two sides: if the union were the side upset with the status quo, would the school system have voluntarily made concessions to the union?

        If the union had been the side upset with the status quo and the school system had refused to voluntarily make concessions to it, I would have blamed the school system, but it’s not what happened. You can say a lot of things with ifs.

        But you’re laying the responsibility for the deficit at the feet of the teachers, and saying that they should not have negotiated in collective self-interest. You’re demanding that they take a cut to their contractual guarantees in a selfless act for the greater good of a balanced budget. Is there any reason why it should have been the teachers who take that cut, rather than any other adjustment to the budget? When is it okay for a group of people to act in its members’ collective self-interest, and when is it not?

        I do believe that, when you benefit from a totally unjust advantage that is responsible for a substantial part of the large deficit your organization has to deal with, it’s morally obligatory for you to give up that advantage. You may disagree with that, but I suspect most people don’t.

        This in particular is an argument by personal incredulity. The AFT/NYSUT is not acting immorally if your disagreement with them is over factual matters (Is the ALST a good indicator of teaching effectiveness? Do children receive more benefit from having more ethnically/culturally diverse teachers over teachers who score better on the ALST?) that neither side appears to have that much evidence for either way.

        I think just having a look at the structure of the test in question, sample questions and the rubrics is more than enough to know that anyone who can’t pass that test should not be allowed to teach. I frankly don’t believe that anyone who has done that and nevertheless pretends to be agnostic is arguing in good faith, so I think we’re going to have to agree to disagree on this point. Again, I’m sure there are other people who would take your side on this, but I’m also sure that, if they had looked at the test and you forced them to put their money (or, in this case, their kids) where their mouth is by agreeing to send their kids to schools where teachers couldn’t pass that test, none of them or almost none of them would. There is a point where asking for more evidence is not a sign of rationality but a sign of irrational skepticism. At the very least, I think it would be irrational not to have a strong presumption in favor of the claim that someone who can’t pass that test shouldn’t be allowed to teach and that the burden of proof is on those who deny that to prove it’s not true.

        The review of the literature you cited shows that teachers’ unions increase the costs of education, I think primarily via an increase in teachers’ compensation. Isn’t that exactly the purpose of a union to begin with?

        First, it’s important to keep in mind that it’s not only driven by the increase in wages. But perhaps more importantly nobody denies that teachers unions benefit teachers. The question is whether the advantage teachers gain from them is worse than the harm they do to the far more numerous kids they teach and taxpayers who pay their wages. I personally don’t think so.

        I’m a bit jaded with this style of “argument by 1000 papercuts”. When I was discussing the issue of the Democratic Party (or Hillary Clinton)’s alleged corruption, the same argument was made. “Look, there are so many examples, just look for them, here are like 20 examples right here.” So I looked at them, and the vast majority of the examples they came up with were blatantly misinterpreted, had no regard for context, jumped to conclusions, applied double-standards, or were just outright false. There were like 2 or 3 allegations of wrongdoing out of dozens that ended up being either ambiguous or confirmed in my mind.

        So no, I’m not going to hunt down the good examples for your stance.

        Even acknowledging that unions do perverse things sometimes, is that a good enough reason to believe that unions themselves are bad? Couldn’t I make the exact same arguments you made, but instead to claim that the United States government is a “morally bankrupt” organization and therefore nobody should support it?

        Well, for reasons having to do mostly with foreign policy, I think it’s totally obvious that the US government is a morally bankrupt organization, but let’s not start this debate 🙂 I agree with the general point you make about “arguments by 10,000 papercuts”, but I think the examples I gave were quite good, so I’m not going to waste my time looking for more of them since I know that when one doesn’t want to believe something one can always apply hyperbolic doubt to it. Again it looks as though we’re going to have to agree to disagree.

        1. I’m getting close to agreeing to disagree, but I’m not quite there yet.

          1. Have you considered that not everyone agrees that “it’s so damaging to the union”? The offer itself did not seem to get all that much press coverage compared to the existence of the rider in the first place.

          Or that the reasons for the union to decline the offer are justified, but in a way that’s subtle/technical enough that instead of carefully and visibly/publicly rebutting it they would rather just avoid furthering the controversy? Unfortunately, PR strategy doesn’t occur in a world where everyone is as Enlightenment-rational as you are. They have to tailor their messaging effort to the overwhelming majority of the public that won’t be spending longer than 2 minutes thinking about the issue.

          Do you think that it is unlikely for a powerful player in a negotiation to deliberately attempt to create a situation that looks bad for the other party, in order to gain a bargaining advantage?

          > I do believe that, when you benefit from a totally unjust advantage that is responsible for a substantial part of the large deficit your organization has to deal with, it’s morally obligatory for you to give up that advantage.

          I think the morality of groups of people is complicated. I will agree that on an individual level, it would have been morally good for a member of the union to fight to fix what was evidently a dysfunctional situation that was harmful to the public good.

          It’s not that I think unions are saintly. I do believe that they are effective in opposing other non-saintly institutions though. So given a choice between a flawed union or unilateral control by a flawed school board….

          Perhaps NYSUT should have intervened, but it’s not an easy choice for them, deciding where to draw the line for the autonomy of member unions. You might say that unions *are* working to fix the situation by opposing the Taylor Law and all its consequences.

          I agree that the U.S. government makes very bad decisions sometimes, or frequently. That doesn’t mean I’d prefer anarchism though. I’d say it is morally good for individual citizens to work to prevent those bad decisions.

          2. Sorry, I was a bit unclear. When I said “I am agnostic on the merits of the ALST”, I meant I *personally* don’t know enough. On the other hand, I do believe that the parties involved in making the decisions about the ALST were better equipped to make those judgments than I was. I did not mean to advocate for more evidence to be collected about the ALST. I will however advocate for you to actually look at the evidence that had already been collected.

          You call my attitude irrational skepticism. I call it an awareness that there are entire fields of study that investigate the fairness and merits of different forms of testing, and that what they find doesn’t always align well with “common sense”.

          Here’s some evidence you might find more convincing, on the reasonableness of skepticism here. Looking through the top few results for journalism covering the abolishment of the ALST, it appears that the director of education policy at the Manhattan Institute was initially in favor of the ALST, but after actually trying a 20-minute practice test for it, decided to come out against the use of the ALST.
          Initially in favor: http://nypost.com/2017/03/06/state-considering-major-changes-to-teacher-certification-exams/
          Later against: http://www.nydailynews.com/new-york/test-meant-screen-n-y-teachers-weeded-minorities-article-1.2995796
          https://www.mhpbooks.com/nyc-cuts-burdensome-teacher-literacy-test/

          The Board of Regents convened a task force of 35 teachers who came to the conclusion that the ALST was too expensive and unnecessary in light of the other tests and requirements.
          http://www.nysed.gov/news/2017/board-regents-act-amend-states-teacher-certification-requirements-based-recommendations

          There is for instance, an edTPA test that teachers did and still have to pass. Among other things, the edTPA has applicants actually demonstrate their skills in real classrooms, whereupon footage of the trial would be reviewed by a panel. So the dropping of the ALST only permits people who would’ve failed the ALST but passed the edTPA. And you have to wonder, if other teachers are looking at footage of someone at work and saying “yeah this is good” while the ALST says they’re illiterate… was the ALST really a good measure of anything useful?

          3. So that entire section of your original article was pretty useless, wasn’t it? You were citing a heavyweight piece of literature to make a conclusion that nobody would’ve denied anyway. I suppose you could say it was a pre-emptive rebuttal to the argument that “unions improve teachers’ performance and are therefore good for students”, though I’m not sure if I’ve ever seen that argument made.

          1. Do you think that it is unlikely for a powerful player in a negotiation to deliberately attempt to create a situation that looks bad for the other party, in order to gain a bargaining advantage?

            Of course I don’t think it’s unlikely, but I think it’s even more unlikely that the union would not try to correct the record, yet I have found no evidence of that. I find your explanation for why that might be the case totally unconvincing. We know for a fact that, for several years, the union has refused to give up a benefit that any reasonable person would agree was completely unjustified, even though it costs several million dollars a year and was responsible for a substantial part of the deficit. In light of this fact, together with the many other cases of abuses by teachers unions I’m aware of (have you ever heard of reassignment centers for teachers that can’t be fire in NYC?), I find the story reported by The Atlantic totally plausible, especially since there doesn’t seem to be any denial/debunking out there. However, I don’t want to continue arguing this point, as I think it’s unlikely to be productive.

            It’s not that I think unions are saintly. I do believe that they are effective in opposing other non-saintly institutions though. So given a choice between a flawed union or unilateral control by a flawed school board….

            But I think the evidence suggests that isn’t true. At least, the evidence suggests that teachers unions have a positive effect on education in terms cost-effectiveness, even though it’s probably true that teachers benefit from them. As I noted before, it’s not that I don’t count the advantages they confer to teachers for something, it’s that I don’t believe it compensates the harm done to students and taxpayers.

            Sorry, I was a bit unclear. When I said “I am agnostic on the merits of the ALST”, I meant I *personally* don’t know enough. On the other hand, I do believe that the parties involved in making the decisions about the ALST were better equipped to make those judgments than I was. I did not mean to advocate for more evidence to be collected about the ALST. I will however advocate for you to actually look at the evidence that had already been collected.

            You call my attitude irrational skepticism. I call it an awareness that there are entire fields of study that investigate the fairness and merits of different forms of testing, and that what they find doesn’t always align well with “common sense”.

            The Board of Regents convened a task force of 35 teachers who came to the conclusion that the ALST was too expensive and unnecessary in light of the other tests and requirements.
            http://www.nysed.gov/news/2017/board-regents-act-amend-states-teacher-certification-requirements-based-recommendations

            I actually read that statement, as well as the task force’s recommendations, when I did the research to write this post. Neither this statement nor the report about the recommendation give me any reason to believe that anything like a systematic, evidence-based assessment of the merits of the ALST was conducted. There is only the claim that the test is costly, although no figure is given, as well as the assertion that it was redundant, which as I pointed out in my post is, at least prima facie, false. (I come back to this point below, when I address what you say about the edTPA.) Again, I have read the study guide for the ALST and my conclusion is that it tests very basic literacy skills, which any teacher should possess. But I’m taking back the stuff about irrational skepticism because of the evidence you uncovered about Sahm’s experience with the test, which I briefly discuss below.

            Here’s some evidence you might find more convincing, on the reasonableness of skepticism here. Looking through the top few results for journalism covering the abolishment of the ALST, it appears that the director of education policy at the Manhattan Institute was initially in favor of the ALST, but after actually trying a 20-minute practice test for it, decided to come out against the use of the ALST.
            Initially in favor: http://nypost.com/2017/03/06/state-considering-major-changes-to-teacher-certification-exams/
            Later against: http://www.nydailynews.com/new-york/test-meant-screen-n-y-teachers-weeded-minorities-article-1.2995796
            https://www.mhpbooks.com/nyc-cuts-burdensome-teacher-literacy-test/

            Actually, Sahm doesn’t say that he supports dropping the requirement to pass the ALST in order to be certified, he just says that after taking the test he agrees that « it’s not a great test », which is not exactly the same thing. He didn’t say that he was in favor of dropping the test without replacing it with anything and I doubt that he would say that. But I agree that the fact that he only got 21 out of 40 questions right in the comprehension section of the test suggests that it’s a bad test, so you have convinced me that there is a prima facie case that it should be replaced. But as I argue below, there is also a prima facie case that it needs to be replaced by something and, unfortunately, the Board of Regents didn’t offer any replacement. Instead, it just said that “the Department will work with the state’s testing vendor to review the existing Educating All Students (EAS) exam to determine if the EAS should also assess candidates’ literacy skills”, which is administrative speak for “we are not going to do anything”. Moreover, although Sahm’s testimony is suggestive, I would still like to see a systematic review of the ALST, because again when I look at the study guide, sample questions and rubrics it strikes me as testing basic literacy skills.

            There is for instance, an edTPA test that teachers did and still have to pass. Among other things, the edTPA has applicants actually demonstrate their skills in real classrooms, whereupon footage of the trial would be reviewed by a panel. So the dropping of the ALST only permits people who would’ve failed the ALST but passed the edTPA. And you have to wonder, if other teachers are looking at footage of someone at work and saying “yeah this is good” while the ALST says they’re illiterate… was the ALST really a good measure of anything useful?

            I have also read the description of the edTPA and, for various reasons, I think it’s probably a very bad test, but I don’t even want to get into that. According to this article, which I also cited in my post to make the same point, the success rate for the edTPA is significantly higher than for the ALST. In the case of white applicants, it’s approximately 14.5% higher, while for black applicants it’s approximately 60.5% higher. (Before you say that such a discrepancy is a proof of unfair racial bias, I will point out that it’s exactly what you would expect if the ALST tested basic literacy skills and those skills were normally distributed among both white and black applicants, although the mean was lower for the latter. This does not show that the ALST is not unfairly biased against minority applicants, but it means that before people make that accusation they need to provide more evidence, which as far as I can tell the task force put together by the Board of Regents did not do.) This shows that, since the requirement to pass the ALST was dropped, many applicants who previously would not have been certified will be allowed to teach. It could be that they should be allowed to teach, because how one scores on the ALST is not correlated to one’s ability to teach, but that is hardly obvious. You seem to think that it’s wildly implausible that people who have a Bachelor and apply to be certified to teach lack basic literacy skills, but unfortunately I’m far more pessimistic than you about that. It’s a well-known problem that teachers in the US are academically very weak. (See for instance table 2 in this report for the distribution of their college entrance examination scores. It’s also a well-established fact that people who major in education score consistently at the bottom of standardized tests.) I went to a shitty school in France and I know plenty of people who have a Bachelor yet lack basic literacy skills. Of course, it’s not that they are literally illiterate (pun totally intended), but it would be irresponsible to put them in a classroom to teach kids.

            So that entire section of your original article was pretty useless, wasn’t it? You were citing a heavyweight piece of literature to make a conclusion that nobody would’ve denied anyway. I suppose you could say it was a pre-emptive rebuttal to the argument that “unions improve teachers’ performance and are therefore good for students”, though I’m not sure if I’ve ever seen that argument made.

            Actually, plenty of people claim that teachers unions have a positive effect on education (although they’re mostly unionized teachers and people who are in one way or another affiliated to teachers unions), so I don’t think it was useless at all. What I said nobody denies is that teachers unions benefit teachers, not that teachers unions have a negative effect on cost-effectiveness in education, which is not the same thing. It could have been that they also have a large enough positive effect on educational outcomes to more than compensate the rise in expenditure they cause, but according to the literature, that is not the case.

          2. Fair, I can acknowledge that it would be good to improve teachers’ qualifications, and my model of a “bachelor’s degree holder” is maybe not well calibrated. On that note, I’m curious if you have any particular stance on what the biggest single obstacle to getting better teachers is. My first instinct here is “pay them more”, but I haven’t thought about it much.

          3. Don’t think I didn’t see what you did just now 😉 Anyway, more seriously, I think it’s a good and very complicated question. I’m rather pessimistic about what can be done, because I think the problem is largely a result of deep social changes that have taken place over the past few decades, which I think makes it harder to address effectively.

            For instance, it used to be the case that teaching was one of the few careers available to women, because there were unfair barriers to entry for them in most other careers. As a result, a lot of bright women used to be teachers and they were probably very good, but now they understandably prefer other careers since they have more options. Obviously, undoing the liberation of women is neither possible nor desirable, so if I’m right that it’s a factor we can’t just go back to the way things used to be.

            I agree that paying teachers more would probably help to some extent, but it needs to be done correctly. Teachers unions don’t just drive wages up, they also compress the distribution, because they insist that teachers can only be payed differently based on seniority. This probably contributes to make the profession less appealing to bright people who are considering a career in teaching, since it’s not very attractive to know that, even if you’re a great teacher who works a lot for his students, you will be payed the same as someone who is terrible at his job and doesn’t care about his students.

            There are many other ways in which teachers unions contribute to make the job less attractive to bright people, but to be fair, I doubt it’s the main factor, although it’s a factor. So I think reducing the influence of teachers unions would help with that problem, but also that the notion that it will completely solve it is a conservative fantasy. (In general, while I agree with conservatives about many of the changes they support in education, I think they tend to be irrationally optimistic about the positive effect they would have.)

            So, at the risk of letting you down, I’m really not sure how to fundamentally address the problem. I think it’s a very difficult problem and I really have no magical recipe to solve it. I fear there may be no good way to really address it. (People often seem to assume that, for every problem, there is a solution, but I’m more pessimistic than that.)

          4. Haha, that was actually completely unintentional. I wasn’t thinking of unions to increase teacher pay, but instead some way of paying more for valuable skills. So eg, someone wouldn’t have to take a big pay cut to go from being an engineer to being a teacher. I wasn’t aware that teachers’ unions insist that pay can only be differentiated by seniority.

            But I appreciate your thoughts on it.

  7. Thanks for sharing your views, I have a couple of comments.

    “In my opinion, a union should only be able to bargain on behalf of the workers who agreed to be represented by it, but not anyone else.”

    Are you familiar with labor history at all? There’s a good reason for this type of provision: it’s how you solve the collective action problem in this space. Frankly, the world you want (only people who voted for the union are represented) basically can’t exist. You have two choices: 1) A vote determines whether there will be a union for all; 2) no unions for anyone.

    As a sociologist, it’s unsatisfying that these are basically the two options—but when we talk about collective action problems and social dilemmas, not all solutions are stable given the preferences and incentives involved.

    A note: many of your other arguments have this quality of asking for things that are very difficult or impossible given reality. I would love for everything to be perfect, but usually we have to make choices with tradeoffs involved. It’s important to weigh the pro’s and con’s with an eye towards the actual choice we face, rather than comparing to an unrealizable ideal (which is your main comparison in this piece).

    Additionally, CGSU has quite a lot of autonomy from AFT, so if you want to see specific provisions negotiated in the contract, I’d encourage you to get involved if the vote passes.

    1. Thanks for the comment. Let me reply quickly.

      Are you familiar with labor history at all? There’s a good reason for this type of provision: it’s how you solve the collective action problem in this space. Frankly, the world you want (only people who voted for the union are represented) basically can’t exist. You have two choices: 1) A vote determines whether there will be a union for all; 2) no unions for anyone.

      You may have missed it, but I actually address the “collective action problem” argument in my post. The only reason why there is a collective action problem in the first place is because of the National Labor Relations Act, which allows a union to bargain on behalf of everyone including people who don’t want to as soon as more than 50% of people in the bargaining unit vote in favor of unionization. Were this not the case, there would be no collective action problem, because people who aren’t members of the union wouldn’t benefit from any advantages the union obtains for its members.

      The notion that, without a system in which people can be forced to pay dues even if they don’t want to, no unions could exist is patently false. In many European countries, such as France and Germany, people can’t be forced to pay dues, yet they sometimes have a much higher union density than the US. It’s true that, in both France and Germany, collective bargaining is often done at the industry level, which results in a much higher union coverage than in the US, but this doesn’t show that you can’t solve the collective action problem without collective bargaining. The collective action problem is largely a product of the legal/institutional framework.

      A note: many of your other arguments have this quality of asking for things that are very difficult or impossible given reality. I would love for everything to be perfect, but usually we have to make choices with tradeoffs involved. It’s important to weigh the pro’s and con’s with an eye towards the actual choice we face, rather than comparing to an unrealizable ideal (which is your main comparison in this piece).

      In my post, I show that CGSU argues in a misleading way, by cherry-picking evidence and neglecting to mention crucial information. This is absolutely indisputable and I have yet to see anyone show me that anything I say about that is false. I don’t think it’s asking too much that a union doesn’t try to misrepresent the facts to mislead people, but maybe that’s just me.

      Additionally, CGSU has quite a lot of autonomy from AFT, so if you want to see specific provisions negotiated in the contract, I’d encourage you to get involved if the vote passes.

      This may be true, although I don’t know exactly what this means, since I was told that we can’t see what the terms of the contract between CGSU and the AFT are without being a member of the union and I don’t want to be a member. (I haven’t checked this claim myself, so it may be false, in which case I’m happy to stand corrected.) But even if that were true, it would have no bearing on my actual complaint, which is that at least $400/year will be taken from our wages to pay the AFT that will be able to use that money as it sees fit. Since I strongly disapprove of the activities of the AFT, I don’t want to support them by giving them money. How much autonomy CGSU has relative to the AFT is irrelevant to the worry I raise in my post.

      1. > The only reason why there is a collective action problem in the first place is because of the National Labor Relations Act, which allows a union to bargain on behalf of everyone including people who don’t want to as soon as more than 50% of people in the bargaining unit vote in favor of unionization. Were this not the case, there would be no collective action problem, because people who aren’t members of the union wouldn’t benefit from any advantages the union obtains for its members.

        If this weren’t the case, there would be a different collective action problem. Now employers can hire non-union workers at lower rates than union workers. Assuming transaction costs for hiring/firing are low and that the free market operates, that means that the employer only hires non-union workers, and again there is no union.

      2. See the other commenter for the collective action problem issue. You don’t eliminate a collective action problem by changing the law so that half the employees can be unionized. It’s simply untrue to state otherwise. The lower-paid non-union workers would drive out union work because employers prefer to pay less (think about it from an employer’s perspective—you can hire an equal candidate for $1000 or $900, which do you pick?).

        Additionally, when it’s really hard for (skilled) workers to switch jobs (like switching PhD programs), you can make an argument that a union prevents an employer from having the vast majority of the negotiating power when setting working conditions. I don’t disagree that our wages at Cornell are OK, but the threat of unionization has already lead to lots of material benefits—why weren’t these provided before if Cornell had the means and thought we deserved them? Recall that our current grad health insurance came out of the first unionization push in 2002.

        The system you are referring to in France and Germany, of unionizing by industry rather than profession, is preferable to our system from what I understand. At least in Germany, There is also a fundamentally different corporate structure of co-determination where union leaders sit on major company boards. This practice makes it a requirement that labor have a say in corporate governance, which then in turn makes it less important for unions to include everyone (since labor has a seat in governance regardless of the formal unionization rate).

        This is a philosophy and legal structure of corporate governance that doesn’t obtain in the US, and we are not debating that. Given our law, you need everyone in the union or it doesn’t work for basic collective action problem reasons. We also don’t face a choice of remaking labor practices in the US to go to industry-based (rather than trade-based) unions.

        We face a choice between unionizing under current labor law, or not unionizing at all.

      3. If this weren’t the case, there would be a different collective action problem. Now employers can hire non-union workers at lower rates than union workers. Assuming transaction costs for hiring/firing are low and that the free market operates, that means that the employer only hires non-union workers, and again there is no union.

        But if the union manages to gather the support of enough workers, the transaction cost for firing unionized workers will not be low, because if the union is efficient the workers that are part of it but aren’t fired would go on strike and it’s not easy to replace a large part of your workforce, especially for skilled work which is increasingly the norm in modern economies. Now, don’t get me wrong, I’m not denying that, if you do that, wages will be lower on average in the economy (but note that, as I explain in my post, it doesn’t mean that it would be true for graduate students), but this means that unemployment and prices will be lower. Unions and closed shop agreements benefit people who already have a job at the expense of consumers and people who don’t have a job. There is always a trade-off and it’s a complicated debate what’s the optimal solution.

        Additionally, when it’s really hard for (skilled) workers to switch jobs (like switching PhD programs), you can make an argument that a union prevents an employer from having the vast majority of the negotiating power when setting working conditions. I don’t disagree that our wages at Cornell are OK, but the threat of unionization has already lead to lots of material benefits—why weren’t these provided before if Cornell had the means and thought we deserved them? Recall that our current grad health insurance came out of the first unionization push in 2002.

        You can’t make inferences about what advantages would be gained from unionization from one case, even if I grant that graduate students at Cornell wouldn’t have obtained health insurance had it not been for the unionization push in 2002, which I shouldn’t because we don’t know that. (Since you’re a sociologist, I’m sure you understand that causal inference is more complicated than that.) As I explain at length in my post, we just don’t know that graduate students would benefit from unionization. Perhaps they would and perhaps they wouldn’t. It’s also possible that some would and others wouldn’t. The evidence from the case of graduate students unions at public universities suggest that, on the whole, it would not make any difference, except that we’d be paying dues to the AFT and NYSUT. Perhaps it would be different at Cornell, which is a private university, but we don’t have enough data to say that. Anyone who says otherwise is just dishonest and, indeed, CGSU had to cherry-pick the evidence and neglect to mention crucial information to make the case that we’d benefit materially from unionization. I note that nobody has shown that anything I say in my post on that count was wrong.

        The system you are referring to in France and Germany, of unionizing by industry rather than profession, is preferable to our system from what I understand. At least in Germany, There is also a fundamentally different corporate structure of co-determination where union leaders sit on major company boards. This practice makes it a requirement that labor have a say in corporate governance, which then in turn makes it less important for unions to include everyone (since labor has a seat in governance regardless of the formal unionization rate).

        This is a philosophy and legal structure of corporate governance that doesn’t obtain in the US, and we are not debating that. Given our law, you need everyone in the union or it doesn’t work for basic collective action problem reasons. We also don’t face a choice of remaking labor practices in the US to go to industry-based (rather than trade-based) unions.

        We face a choice between unionizing under current labor law, or not unionizing at all.

        You may not be arguing about that, but that’s exactly what I am arguing about in my post, at least in the last part. I’m precisely arguing that the National Labor Relations Act is unjust. So you’re right that we face a choice between unionization under current law or not unionizing and what I’m arguing is that we should not unionize. As I argue in my post, the evidence doesn’t support the claim by CGSU that we’d benefit, but we know for a fact that we’d be paying dues to the AFT and NYSUT, which I and many other people want nothing to do with. I don’t see why I should agree to financially supporting organizations I find despicable for a zero expected gain in net compensation and work conditions.

        1. We literally got better healthcare in the middle of the union election, which is transparently an attempt to stop the unionization effort. A similar thing happened in 2002. I find it extremely difficult to believe that this is a coincidence given the timing and that the administration is against unionization (in causal inference, to argue against a factor being causal you need a plausible alternative factor, particularly when you have a strong prior that a factor is causal). And it also shows that organizing works: we have already received a material benefit in response to our unionization effort.

          If your position is that we need massive & systemic change before any progress can be made on labor issues, we will never make progress on labor issues. To be honest, this is a position that sounds like it comes from a place of ignorance about how political & social changes happen (which is why your argument frustrates me so much as a sociologist—you have an unreasonable ask before any progress can be made). Check out Theda Skocpol’s work on how new social policies tend to resemble already-existing social policies for some academic work on the subject. There is a path-dependence to national policies and working within that existing framework (while bending the framework by taking a step forward) is usually the way to get stuff done.

          1. We literally got better healthcare in the middle of the union election, which is transparently an attempt to stop the unionization effort. A similar thing happened in 2002. I find it extremely difficult to believe that this is a coincidence given the timing and that the administration is against unionization (in causal inference, to argue against a factor being causal you need a plausible alternative factor, particularly when you have a strong prior that a factor is causal). And it also shows that organizing works: we have already received a material benefit in response to our unionization effort.

            I’m not denying that it happened because the administration wanted to stop the unionization effort, what I said is that you don’t know that it wouldn’t have happened otherwise, which indeed you don’t. It’s entirely possible that graduate students could have obtained the same advantages by other forms of organization than trying to form a union. But let me grant you, for the sake of the argument, that we could only have obtained better health care benefits through efforts to form a union. It would only follow that the threat of allowing a union to represent us through collective bargaining is effective, not that actually voting to let a union represent us through collective bargaining would be effective. Indeed, as I explain in my post, the evidence that is currently available doesn’t support this claim. Which is why, in order to argue that we’d definitely benefit from voting in favor of unionization, CGSU cherry-picks the evidence and neglects to provide crucial information. As I already noted, although people have argued with several of the points I made, nobody has found anything to say against this particular claim which I defend in the first part of my post.

            If your position is that we need massive & systemic change before any progress can be made on labor issues, we will never make progress on labor issues. To be honest, this is a position that sounds like it comes from a place of ignorance about how political & social changes happen (which is why your argument frustrates me so much as a sociologist—you have an unreasonable ask before any progress can be made). Check out Theda Skocpol’s work on how new social policies tend to resemble already-existing social policies for some academic work on the subject. There is a path-dependence to national policies and working within that existing framework (while bending the framework by taking a step forward) is usually the way to get stuff done.

            You’re completely misrepresenting my argument and begging the question in the process. I’m not saying that, although unionization would have a positive effect for graduate students, we should nevertheless vote against it because it’s not perfect or something like that. I’m denying that we’re a position to know that it would have a positive effect or at least that we’re in a position to rationally believe that it’s more likely than not to have a positive effect, but I’m saying that, on the other hand, we know that if we voted in favor of unionization we’d financially support organizations I personally strongly disapprove. You take for granted that voting for unionization would bring about progress, but as I pointed out above, you simply don’t know that or even have very good reasons to think so, because the evidence is clearly insufficient.

            I imagine that you disagree with me on the role of teachers unions, which is fair enough, but don’t assume that unionization would bring about positive changes for graduate students at Cornell unless you can show that, which you can’t. Also, I would appreciate it if you didn’t attribute ignorance to me, unless you can also show that I’m ignorant. The fact that you’re a sociologist doesn’t mean that you have read more than me about this and, to be perfectly honest with you, I doubt that you have. Unlike CGSU, which systematically censors dissenting voice on its Facebook page, I allow people to criticize me in the comments of this blog (which to be clear I think is perfectly normal), so given that I’m assuming you’re a member of CGSU, I think the least you could do is avoid that kind of attacks.

          2. I think we reached reply depth here (haha), but I wanted to reply specifically about teacher’s unions specifically:

            > I imagine that you disagree with me on the role of teachers unions

            You also make the argument above in your main post that teachers unions are rent-seeking and don’t improve student outcomes.

            I actually don’t disagree that teacher’s unions can be rent-seeking organizations, they probably are in some cases (as you cite). However, I think we as grad students are in a unique position of being high-skilled workers where it is *extremely* costly to switch to another equivalent job. We have imperfect information when we sign up in year 1, and must ride out that commitment or lose years of our lives trying to change PhD programs (or abandon a PhD altogether).

            This difficulty of switching jobs means that getting employment conditions correct is super important AND that our main mechanism for bettering conditions is working with our employer. This follows because we can’t use the normal labor market mechanism of switching jobs to better our condition. A union seems to make a whole lot of sense (to me) in this kind of circumstance: it takes a not-really-functional labor market and turns it into a negotiation. I get your skepticism about what we can gain and about the particulars of THIS union, but that’s my argument for A union.

            Also, I stand by the ignorance comment. I am not saying you’re not an intelligent guy (you clearly are). However, just as I haven’t taken the time to carefully consider things in your discipline, your comments about the NLRA and letting part of the grad students unionize indicate a lack of understanding about the incentive structure we face. This isn’t about knowledge of the specific CGSU process (which you might know more about than I do). The NLRA is the way it is for a good reason: because you create at least two big collective action problems if you don’t require everyone in a place to unionize/pay negotiation fees (free rider and incentive to undercut the union). I get your antipathy to the AFT, but there’s really no other way I’m aware of to get the incentives right for unions without having everyone on the boat.

            Thanks for letting me post a dissenting view and have a discussion with you here. I get your views better now, even though I still disagree on the big points.

          3. I actually don’t disagree that teacher’s unions can be rent-seeking organizations, they probably are in some cases (as you cite). However, I think we as grad students are in a unique position of being high-skilled workers where it is *extremely* costly to switch to another equivalent job. We have imperfect information when we sign up in year 1, and must ride out that commitment or lose years of our lives trying to change PhD programs (or abandon a PhD altogether).

            This difficulty of switching jobs means that getting employment conditions correct is super important AND that our main mechanism for bettering conditions is working with our employer. This follows because we can’t use the normal labor market mechanism of switching jobs to better our condition. A union seems to make a whole lot of sense (to me) in this kind of circumstance: it takes a not-really-functional labor market and turns it into a negotiation. I get your skepticism about what we can gain and about the particulars of THIS union, but that’s my argument for A union.

            I agree that, if it’s harder for a category of workers to switch jobs, then it’s a reason in favor of a union. But it doesn’t follow that we should have a union even if we’re in that situation because there may be other reasons not to have one, as indeed I think there are in our case. Perhaps more importantly, I wasn’t arguing that teachers unions are rent-seeking organizations that have a negative effect on education to make the comparison with CGSU, but rather because I think it means that we shouldn’t financially support that kind of organizations. If you agree that teachers unions have a negative effect on education, especially for poor students, then I find it baffling that you have no problem paying them something like $1 million dollar per year.

            Also, I stand by the ignorance comment. I am not saying you’re not an intelligent guy (you clearly are). However, just as I haven’t taken the time to carefully consider things in your discipline, your comments about the NLRA and letting part of the grad students unionize indicate a lack of understanding about the incentive structure we face. This isn’t about knowledge of the specific CGSU process (which you might know more about than I do). The NLRA is the way it is for a good reason: because you create at least two big collective action problems if you don’t require everyone in a place to unionize/pay negotiation fees (free rider and incentive to undercut the union). I get your antipathy to the AFT, but there’s really no other way I’m aware of to get the incentives right for unions without having everyone on the boat.

            As I pointed out above, the choice between different legal/institutional frameworks always involve trade-offs. You’re assuming that I don’t know what the trade-offs are, when in fact I just disagree with you on how to evaluate them. If you think that reasonable people can’t disagree about whether the NLRA is the right solution, you clearly haven’t given this enough thought. The question of how to rank different legal/institutional arrangements not only depends on complicated empirical questions, but also on possibly even more complicated philosophical questions about issues such as distributional justice. For instance, if unions tend to improve the situation of relatively skilled workers at the expense of consumers and people who are relatively unskilled, it’s hardly obvious whether they are a good thing. I think the NLRA is a bad law that should be repealed. As I explain in my post, as long as you don’t allow non-members to benefit from the advantages negotiated by the union for members, it’s simply not true that repealing the NLRA would allow non-members to free ride. It is true that, by getting rid of the kind of the kind of collective bargaining allowed under the NLRA, you create some incentive to undercut the union, but how strong that incentive is depends on the specifics of each situation and, on the whole, I think it would have a positive effect for society. I may be wrong about that, but telling me that I’m ignoring basic facts about the trade-offs involved does not do anything to show this, because I’m not.

            Thanks for letting me post a dissenting view and have a discussion with you here. I get your views better now, even though I still disagree on the big points.

            To be clear, I don’t think there is anything particularly praiseworthy about allowing dissenting views, I just think there is something extremely blameworthy about not doing so, which is another thing that turns me off about CGSU. Having said that, I appreciate the pushback I got from you and Jonathan, even though I think you’re wrong. I wish there could have been a real debate about the union, but unfortunately there wasn’t, largely because CGSU didn’t want it.

  8. unions ruin work !! one time i had to sweep floors because robbie fell asleep and the union made me do his work ! wouldn’t wake him up fear of reprimand if u hear union run the other way and hide trust me !!

    thanks
    karl

Comments are closed.