Ever since the 1983 Nation at Risk report, America has seemingly gone through one educational reform after another. Have these reforms worked? My guest today, Paul Reville, thinks the reforms have correctly focused on the goals of excellence and equity but have not addressed the systemic problems impacting schools.

Paul Reville is the founding director of the Education Redesign Lab at the Harvard. Prior to his time at Harvard, he was the Education Secretary for the Commonwealth of Massachusetts. As Governor Patrick’s top education adviser, Paul brings valuable insights to his work of the real-life political challenges that sometimes slow educational change.

Paul is the Francis Keppel Professor of Practice of Educational Policy and Administration at the Harvard Graduate School of Education.

Citation: Reville, Paul, interview with Will Brehm, FreshEd, 96, podcast audio, November 20, 2017. https://www.freshedpodcast.com/paulreville/

Transcript, translation, resources:

Read more

OverviewTranscript中文翻译Français TranscriptionResources

What’s the relationship between test scores and gross domestic product? Do higher test scores lead to higher GDP?

This question may seem a bit strange because most people think about the value of education on a much smaller, less abstract scale, usually in terms of “my children” or “my education.” Will my children earn a higher wage in the future if they do well on school examinations today? If I major in engineering, will I earn a higher income than if I majored in English?

The answer to these question is usually assumed to be a resounding “yes.” Doing better on examinations or studying subjects that are perceived to be more valuable will result in higher wages at the individual level and higher GDP at the national level. Such a belief shapes educational policies and influences educational decision making by families. It has even resulted in a global private tutoring industry that prepares students for tests in hopes of getting ahead.

But what if this assumption isn’t true? What if the relationship between test scores and GDP isn’t so straightforward?

With me today are Hikaru Komatsu and Jeremy Rappleye. Recently they have been publishing numerous articles (see here, here, and here) challenging the statistical research supporting the conclusion that higher tests scores cause higher GDPs. Instead, they find that test scores don’t determine GDP by all that much.

Hikaru Komatsu and Jeremy Rappleye are based at the Graduate School of Education at Kyoto University. Their most recent op-ed appeared in the Washington Post.

Citation: Komatsu, Hikaro and Rappleye, Jeremy, interview with Will Brehm, FreshEd, 71, podcast audio. May 1, 2017.

Will Brehm:  2:12
Hikaru Komatsu and Jeremy Rappleye, welcome to FreshEd.

Hikaru Komatsu:  2:16
Thanks for having us.

Jeremy Rappleye:  2:18
Thank you, Will. Before we begin, let me just say how much I really enjoy your show. And I learned so much from it. And I really applaud you for creating this space and doing such high quality shows week in and week out. So thank you very much for having us.

Will Brehm:  2:30
Thanks for the kind words. You, two have been doing quite a lot of work lately on really challenging some commonplace assumptions between test scores and GDP – gross domestic product. What is the normal relationship many researchers have between test scores? What students know and gross domestic product? How much a country is growing or how much it’s worth?

Jeremy Rappleye:  2:59
Yeah, thanks, Will. I’ll take the first question here, I would say that the common understanding of the relationship between test scores and gross domestic project is the higher your test scores, the greater your future GDP. This is the claim at its most simple and if we try to be more specific, the understanding is that the higher people test scores in a particular population in fields that are, say relevant for economic growth, the higher GDP will be in the future. So relevant fields here are most likely to be defined as math and science or math and science scores, and also language to a certain extent in particular reading. So the exact fields that international learning assessments such as PISA measure, and in that sense, we can also be more specific about the type of economic model embedded in this common understanding. Specifically, it is one that envisions an economy growing as a result of technological progress. That is, the more technological innovation and the accumulation of knowledge, the higher economic growth rates will be in the future.

Will Brehm:  4:02
And what sort of evidence exists, like do researchers have data, empirical data that shows that this relationship is correct, that higher test scores will equate to higher GDP in the future?

Jeremy Rappleye:  4:19
Yes. So in particular, in the last, let’s say, 10 years, the empirical research base for these claims has become very strong in some circles. Now, try to be specific here, the evidence for this common understanding in its current form, I believe, comes from, primarily from two researchers: one is Eric Hanushek at Stanford University and Ludger Woessmann. I apologize, I probably don’t pronounce the name right, based at Munich University. And I think Eric Hanushek appeared on FreshEd quite recently, if I’m not mistaken. In any case, their work constructs roughly a 40 year history of test scores, and matches that with 40 years of economic growth worldwide. So roughly from the 1960s to the year 2000. And when they say worldwide, they actually mean about 60 countries globally, mostly the high income countries that have participated in international assessments, international achievement tests consistently over that period. So to be more specific, their 40 year history of test scores combines data from two international comparative tests: the IEA-SIMS and TIMSS studies and the OECD’s recent PISA studies. So for GDP, they use a standard Penn World Table data set and listeners who are interested can find kind of full details of this in our paper. But I think the point here is that when Hanushek and Woessmann look at the longitudinal relationship between test scores and GDP growth, they find a very strong correlation. That means that across 60 countries, the higher test score outcomes were, the higher GDP growth was, and I think Hikaro might talk about this in more detail, the difference, particularly between the idea of association and causality.

But the point I want to make here is that it’s been so strong, these empirical claims have been so strong, that’s given a lot of momentum to the idea that there’s this strong empirical basis for the linkage between test scores and GDP. And I think that the work of Hanushek and Woessmann is spelled out in many places, as maybe I’ll discuss later in the interview. But the most comprehensive treatment is found in a book entitled “The Knowledge Capital of Nations – Education and the Economics of Growth”, which was published in the year 2015.

Will Brehm:  6:43
So you said that there’s a strong correlation, but is this the relationship between test scores and GDP causal? And I mean, maybe this is a little getting into some of the more technical statistical language here might be useful, to try and understand this claim.

Jeremy Rappleye:  7:02
Yes. So it’s a very good point. And it’s very important to understand the difference between an association or correlation and causality, I think we probably are betters to wait for Hikaro’s discussion of this. But let me kind of lead into that by giving you two quotes where Hanushek and Woessmann really make the claim that the relationship is causal, not just the correlation. So the first quote and this is this kind of crystallizes or kind of encapsulates the whole findings from their body of work, and they say, quote, “with respect to magnitude, one standard deviation in test scores measured at the OECD student level is associated with an average annual growth rate in GDP per capita, two percentage points higher over the 40 years that we observed.” Now, in that quote, they use the word association, as many of the listeners will have heard. But elsewhere, they talk and talk repeatedly about causality. So here’s the second quote, they say, “our earlier research shows the causal relationship between a nation skills its economic capital, and its long run growth rate, making it possible to estimate how education policies affect each nation’s expected economic performance.” So in simple terms, if you can boost test scores, you will achieve higher GDP growth. And this certainty comes out of the idea that the relationship is indeed causal.

Will Brehm:  8:38
That level of certainty, obviously, must impact education policymakers, right to know that if you increase scores as measured on PISA or TIMSS, you will achieve greater economic growth. I mean, it seems like it makes the lives of policymakers a lot easier.

Jeremy Rappleye:  8:55
Absolutely Will, we believe that the attraction, both the attraction of this claim, and the impact of this claim is growing. And through these types of studies, policymakers who previously had to deal with a very complex equation around education are led to believe that the data shows that an aggressive reform policy that increases test scores will, say, 20 or 30 years in the future, lead to major, quite major economic gains. And if I can give you just another quote, and I apologize for the quotes, but I don’t want you to think I’m misphrasing or summarizing the work of Eric Hanushek and Woessmann, but this quote that of their shows the types of kind of spectacular education gains or economic gains that policies can expect to achieve if they implement this kind of policies directed towards raising test scores. So here, I quote, “for lower middle income countries, future gains would be 13 times current GDP and an average out to a 28% higher GDP over the next 80 years. And for upper middle income countries, it would average out to a 16% higher GDP” unquote. So now, Will, if you are a policymaker, you wouldn’t want to forfeit these games, would you? So this research becomes really a motivation for policymakers putting a much greater emphasis on not just math and science, but on cognitive test scores across the board.

But there’s a fascinating bit here, I want to highlight and maybe we want to unpack it later on in the interview. But you might expect that this kind of narrowing the focus of education around test scores would create a lot of resistance. But actually, the GDP gains of increasing test scores that Hanushek and Woessmann project is actually so great that it is projected to pay for everything in education. So it’s not really a choice between alternatives. But instead of a sure-win policy versus kind of more of the same policy, uncertainty, ambiguity, complexity that we’ve seen in the past.

And sorry, this is a long answer, Will. But I would just really want to emphasize, if we talk about how these academic research claims are finding their way into policy or impacting policy, we have to talk about two organizations that have really latched onto these views, and are advocating them strongly to policymakers worldwide. And the first is the World Bank who hired Hanushek and Woessmann to connect their academic findings to policymaking for low income countries. And this report was published by the World Bank as education quality and economic growth as the title “Education Quality and Economic Growth” was published in 2007. And the second organization that has been really at the forefront here is the OECD, they also hired Hanushek and Woessmann to share their findings and discuss the policy implications. And this report was entitled, “Universal Basic Skills: What Countries Stand to Gain” and that was published in 2015. So maybe towards the end of the interview, after we discuss our study, we can return to discuss how these organizations are really central to making that empirical work into concrete policy recommendations.

Will Brehm:  12:13
So what sort of problems do you find with Hanushek and Woessmann analysis of the relationship between test scores and GDP?

Hikaru Komatsu:  12:23
Okay, the problem we found is temporal mismatch that is Professor Hanushek used an inappropriate period for economic growth. That is Professor Hanushek compare test scores recorded during 1960 to 2000 with economic growth for the same period. But this is a little bit strange. Why, because it takes at least several decades for students to become adults and occupy a major portion of workforce. So from our perspective, test scores for given period should be compared with economic growth in subsequent periods. This is the problem we found.

Will Brehm:  13:17
So for instance, it would be like the test scores from 1960 should be connected to the economic growth rate of say, the 1970s there needs to be some sort of gap between the two, is that correct?

Hikaru Komatsu:  13:31
That’s correct.

Will Brehm:  13:31
Okay. So then in your study, I mean, did you do this and what did you find?

Hikaru Komatsu:  13:36
We did this, our study is very simple. We compare test scores for 1960 to 2000, which is exactly the data used by Professor Hanushek. We compare this data with economic growth in subsequent periods such as 1980 to 2000 or 1990 to 2010 or something like that. And we found that the relationship between test scores and economic growth were much much weaker than that reported by Professor Hanushek. Probably audience would see figure in the web, in the website of FreshEd, and there would be two figures, left one is the original one reported by Professor Hanushek. And there is a strong relationship between economic growth and test scores. While the right one is that we found and the relationship is very unclear. So let me explain how we that relationship is when test score for 1960 to 2000 was compared with economic growth. For 1995 to 2014, only 10% of the variation in economic growth among countries was explained by the variation in test scores. That is, the remaining 90% of the variation in economic growth should be responsible for other factors. This means that it is totally unreasonable to use test scores as the only factor to predict future economic growth. And this is what Professor Hanushek did in his study and policy recommendations.

Will Brehm:  15:40
So in your study, you found that 10% of the variation of GDP can be explained by the variation of test scores. What percentage did Hanushek and Woessmann’s study uncover?

Hikaru Komatsu:  15:53
Probably their percentage was around 70 and we try to replicate Professor Hanushek’s finding, in our case, the percentage was 57 or so and the difference between 57, 70 would be caused by the difference in the version of the data we used. In order to extend that time period, we used an updated version of that data, that data set is exactly the same, but the difference is only the version.

Will Brehm:  16:28
So a difference between 50% and 70% is pretty minimal. But the difference between 10% and 70% is enough to question the basically the conclusions that are drawn from that data.

Hikaru Komatsu:  16:42
Yeah, right. If that relationship originally reported by Professor Hanushek is causal, we should have found that comparably strong relationship between test scores for a given period and economic growth for subsequence periods. But we found very weak relationship so it suggests that that relationship originally reported by Professor Hanushek does not always represent the causal relationship. This is our point.

Will Brehm:  17:17
It seems like this is a very profound point that could be rather earth shattering for many people’s assumptions about education and its value for economic growth.

Jeremy Rappleye:  17:31
Thanks, Will, I’d like to fill in a little bit pick up on what Komatsu sensei was arguing there. One of the important points to understand about this 70% of the variation is explained by test scores, the strength of that correlation leads to very strong policy recommendations. So, I’m going to try to unpack a little bit of what I said earlier, because I think it’s an important point, I’ll try to do it academically first, and then I will try to give a simple version that will be a lot easier for listeners to understand.

But basically, if you have a correlation causality link that strongly then Hanushek and Woessmann are claiming that you can first increase your test scores, and it will produce so much excessive growth in the future, that you can redirect that excessive growth back into all other types of educational goods that you need. So in terms of equity, in terms of inclusiveness, you could even redirect that much extra money into healthcare, to sustainability goals, all of these things. So as we both know, these are two sides of the camp is education for economic growth, or is it for equities, is it for inclusiveness, is it personal development. These are the types of debates that have always been with education as an academic study, but he’s able to transcend, they are able to transcend that debate based on the strong causal claims, though, just to fill that in academically. So the claim is that if 3.5% of GDP is spent for education, this is from the World Bank report in 2007. But if over 20 or 30 years, you could increase your test scores by point five standard deviation, it would lead to 5% higher GDP on average, and quote, “this gross dividend would more than cover all the primary and secondary schools spending.” What that means is if you focus on raising cognitive levels, test scores, you could get enough growth that you would ultimately get more money for education for whatever types of educational goals you want to pursue. In the 2015 OECD report Hanushek and Woessmann write that quote “the economic benefit of cognitive gains, carries tremendous potential as a way to address issues of poverty and limited healthcare and to foster new technologies needed to improve the sustainability and inclusiveness of growth.” So as I mentioned before, instead of a trade off between growth related policies and equity, the Hanushek’s results are so strong that they suggest that first raising test scores will eventually produce enough extra gain to pay for everything. So this really relieves them of the need to engage in the kinds of debates over priorities that have taken place for as long as education policy has been around.

Now, if that sounds academic, I apologize. Let me try to put it in more concrete terms to make it easier to understand. Sure. So if the United States based on the PISA 2000 scores could have a 20 or 30 year reform plan that would eventually lead them to achieve the level of Finland or Korea’s PISA scores in 2006. Hanushek claims that, Hanushek and Woessmann claims that GDP, United States’ GDP will be 5% bigger. Now Hanushek really spotlights that in 1989, the governors of the United States came together with then president George Bush, and they promised by the year 2000 to make America number one in the world in math and science. So at that time, it would have been a 50 point gain. And so Hanushek argues in a different work but produced by Brookings called “Endangering Prosperity”, that if the US would have stayed the course in 1989, and actually achieve the goals rather than getting distracted, its GDP would be 4.5% greater today. And that would allow us to solve all of our distributional, our, I’m American, apologies our distributional or equity issues that have constantly plagued American education. And even in more concrete terms Hanushek is based on that strong causality. He’s saying that the gains would actually equal and I quote, “20% higher paychecks for the average American worker over the entire 21st century.”

Will Brehm:  22:02
So he’s reading the future with that relationship that he assumes he can read the future and basically says that focus on test scores first and worry about everything else later, because we’re going to increase so much GDP that will be able to pay for everything that education needs. That’s kind of that policy gist.

Jeremy Rappleye:  22:25
That’s absolutely right, Will, and he made this calculation for the United States, as you can imagine, he’s based in the United States. But if you look at the OECD report, in 2015, they actually make the same calculations for all countries worldwide. So we spotlight that in our full paper in the introduction, just picking up the case of Ghana, because OECD really picked up the case of Ghana and say, you would have this huge economic gain if you could just stay the course on raising test scores.

Will Brehm:  22:54
So these are future predictions or projections. Has there ever been like a real life example of like a bit working the way Hanushek and Woessmann theorize?

Jeremy Rappleye:  23:07
Well, I believe that they would probably argue that the piece of data itself shows that it works. So of course, you have up and down fluctuations of individual countries, but I think they would have a hard time showing a particular country or giving you a case of a particular country who enacted reforms and then achieved higher GDP growth. It’s all abstracted to the level of correlation and causality rather than brought back into kind of concrete terms of particular countries.

Will Brehm:  23:44
And your analysis obviously shows that those future projections are incorrect, or wouldn’t necessarily work out the way Hanushek and Woessmann claim. You know, how do we begin to theorize that this connection between test scores and GDP, like what sort of implications does your study have on education policymakers?

Hikaru Komatsu:  24:09
In my opinion, or according to the data, it is okay to say that improving test score both lead to higher economic growth on average, but actually test scores are only one factor as we found, as we said, only 10% of the variation in economic growth was explained by the variation in test scores. So in that sense, educational test scores are only one factor affecting economic growth. And as we see there’s a huge variation in fiscal capital, land or enterprise between countries, and those should affect economic growth. So education is only one of those factors. This is, I think, reasonable understanding of the relationship between economic growth education and then affect us.

Will Brehm:  25:14
So in a sense, we recognize that education plays some role in future economic growth. But there are other things that also affect future economic growth. And we shouldn’t lose sight of them either.

Hikaru Komatsu:  25:25
You’re right. So that problem Professor Hanushek had is that he believed or assumed that it is the sole factor and he uses only test scores to predict future economic growth. But the route is not so simple. This is our point.

Will Brehm:  25:45
And it seems like it’s a simple point you’re making. But like I said, I think it’s quite profound, because it really upsets what countries are pursuing in their educational goals. I mean, it challenges the rise of PISA, I mean, are all of these countries that are trying to join PISA? Is this actually what they should be doing, right? I mean, for me, the policy implications become so much more difficult, and the confidence of certain policy prescriptions kind of goes out the door.

Jeremy Rappleye:  26:16
Yeah, Will. We would wholeheartedly agree with that summary of kind of the implications of our study, it is a simple idea. And it’s pretty obvious even to, let’s say, masters level students, that education is not that complex. But one of the problems is all of this big data creates all the potential for kind of this dog fight using data up at the higher stratosphere. And people can’t really touch that. So as long as you are up there fighting it out, then it seems to be pretty solid. And I think that to be very honest, actually, neither Hikaro or I really like doing this work that much. Right. To be very honest, we don’t really enjoy this work. We’d rather be thinking about big ideas and complex ideas and doing those types of things.

But the problem is that it blocks the view, the simple views of education block the complexity or the depth of what we should be seeing in these ecologies of education or these types of things. Now, that was a very kind of big picture response. But if we have time, I’d like to talk more specifically about the specific policy recommendations and what are the implications of our studies. So we believe that our research findings have recommendations but these are not really recommendations in the usual sense of identifying a best practice or a magic bullet or as a magical potion that will improve education worldwide. Instead, the implications of our research is what we might call negative policy recommendations. And by that, I mean it helps policymakers realize what they should not do. Specifically, it tells policymakers that they should not be seduced by promises that focusing on raising test scores, and purely test scores in areas such as science technology, math is a surefire policy that will raise GDP growth rates, and it tells them not to believe kind of advisors who would come in and tell them that raising test scores alone will lead to enough GDP, future growth to quote “and the financial and distributional problems of education,” unquote. Now I want to even be more specific about this, more concrete, two points. As most listeners will know, one of the biggest educational policy trends over the last two decades has been PISA. And there are currently plans to extend PISA to low income countries through the PISA for development exercise. I think that by 2030, the OECD and the World Bank plan to have PISA in every country worldwide, despite a whole range of critiques from academics, from practitioners, from just the normal belief that education is more complex than that.

The central rationale for the expansion of PISA testing is that it will lead to higher GDP growth. In effect, countries are being persuaded to sign up to PISA, because of the types of claims that we reviewed throughout this interview. But our research shows that this will not happen. Some of my favorite research in recent years have come from scholars around Paul Morris, at the Institute of Education in London, working with young scholars, Euan Auld, Yun You or Bob Adamson in Hong Kong, showing how PISA tests are really driven much more by a range of private companies such as Pearson, ETS, and so on. And scholars like Stephen Ball, Bob Lingard, and Sam Seller also write some great stuff along these lines. And another important line of research comes from folks like Radhika Gorur who writes about the dangers of standardization and how it might ultimately destroy the diversity necessary for future adaptation and innovation in education. And so again, we hope that our research removes the belief that research that, academic research somehow proves the PISA and GDP linkage, and that’s let’s policymakers see all of these warnings much more clearly. And if you let me quickly go on to the second dimension of really concrete, what’s happening now is that, as many listeners will also know, is that the world is, the world of education, where the development more generally is talking about the post 2015 goals. Basically, what comes after the Millennium Development Goals that we’re going in the 1990s. And in terms of education, the Sustainable Development Goal Four is the one that deals with education, it sets global targets for improving learning by 2030. And one of the disappointing things we have noticed in these discussions is that it seems the discussion seem to be imitating the OECD and World Bank that is, we see UNESCO and other agencies referring explicitly to the Hanushek and Woessmann studies to argue for why PISA-style assessments are the best way to achieve Sustainable Development Goal number four. And so compared to discussions around EFA, in the early 1990s, the discussion around the SDG number four seems to be taking this knowledge capital claim as truth as academic truth.

And we worried that this will put the whole world on a course for implementing PISA-style tests. And, of course, the change in curriculum that comes in its wake. I don’t want to be, you know, kind of, to overstate this too much. But we worried that there’s really no evidence for that and that these will be very costly exercises that will ultimately do very little to improve education. So, again, we hope our study will give policymakers the kind of academic research basis for resisting the advances made by the OECD and World Bank.

Will Brehm:  31:59
Have you experienced any pushback about some of the findings because I mean, obviously, you’re challenging some of the wisdom that’s taken for granted by the World Bank, by the OECD, by private companies, like you said, Pearson and ETS, the Educational Testing Service, which produces a whole bunch of tests. So I mean, one would imagine that your negative recommendations that come out of your findings may ultimately create a pushback from those who interests are being challenged.

Jeremy Rappleye:  32:30
Yeah, I guess we would love to have a pushback, because pushback implies an explicit engagement. Again, our findings are not new, that there’s no link between educational outcomes and GDP growth. These claims are very, this idea is actually very old. But what happens is that with each kind of wave of data that comes out that kind of dog fight that I was talking about, kind of gets it goes from maybe kids throwing rocks at each other from different trees up to hot air balloons up to airplanes up to jet planes, and it just keeps going up. There’s no real engagement with the ground level realities that would refute all of this. So if we were to get pushback, we would welcome it. We would love to see the evidence because our mind is not made up. It’s quite possible. I mean, there are no certainties, and it would be wonderful to see a more elaborate discussion around these ideas. So our results are conclusive. But in the sense of with that data set, it conclusively disproves a particular hypothesis or claim, but they’re not conclusive in the terms of a terminus of learning. There’s always more that we can understand about the relation the complex relationship between society, economics, culture and these types of things. So we would really welcome that as a way to elaborate.

Will Brehm:  34:00
Well, I really hope that you can kind of open up this door for a much deeper engagement to get to some of those big questions that you obviously have in mind, but Hikaro Komatsu and Jeremy Rappleye, thank you so much for joining Fresh Ed. It was really a pleasure to talk today.

Hikaru Komatsu:  34:15
Thanks for having us. I really enjoyed it.

Jeremy Rappleye:  34:18
Thank you very much, Will. Keep up the great work. We all appreciate the hard work you’re doing on behalf of educational researchers and educational practitioners worldwide.

Will Brehm 2:12
Hikaru Komatsu和Jeremy Rappleye,你们好,欢迎做客FreshEd!

Hikaru Komatsu 2:16

Jeremy Rappleye 2:18

Will Brehm 2:30

Jeremy Rappleye 2:59

Will Brehm 4:02

Jeremy Rappleye 4:19
是的,尤其在过去10年左右的时间里,这一观点在某些圈子里得到了很强的实证研究支持。展开来说,很多关于目前这个共识的证据主要都是源自于两位研究者,分别是斯坦福大学的埃里克·哈努谢克(Eric Hanushek)教授和慕尼黑大学的卢德格尔·沃斯曼因(Ludger Woessmann)教授。如果我没记错的话,FreshEd之前有一期节目就请到过哈努谢克教授。话说回来,他们两人研究了1960至2000年间,差不多40年的考试成绩,并与同时期世界范围内的经济发展情况相对照。这里所谓的世界范围是指全球近60个国家,绝大多数是在上述时期参加过国际评估和国际成果测试的高收入国家。更具体地来说,他们结合了两项国际比较测试的数据,即国际教育成就评价协会(IEA)的第二次国际数学研究(SIMS)和国际数学与科学趋势研究(TIMSS),以及经济合作与发展组织(OECD)的PISA;GDP采用的是佩恩表(Penn World Table)的标准数据库,感兴趣的读者可以在我们的论文里找到所有的详细数据。哈努谢克和沃斯曼因研究了考试成绩和GDP增长之间的纵向关系,发现两者有很强的相关性。也就是说,在这60个国家里,谁的测试得分越高,谁的GDP增长就越高。这一点过会儿Komatsu教授会详细讨论,尤其是关于相关性和因果的概念。

Will Brehm 6:43

Jeremy Rappleye 7:02
你问到点子上了!关联性,或者说相关性,和因果性是不一样的。理解这一点很关键。我想最好是一会儿由Komatsu教授来具体解释。在那之前,我先引用两处哈努谢克和沃斯曼因的原话,他们提到成绩和GDP之间有因果关系,而不仅仅是有相关性。第一处是在概述研究的主要发现时,他们写到:“对于影响的程度,OECD 学生水平测试成绩的一个标准差与GDP的平均年增长率有关,在我们观察到的40年间,这一数字高出了两个百分点。”大家可能也都听到了,他们在这里用的词是“关系到”。但此外的其他地方,他们反复谈论的却是“因果关系”。比如我要引用的第二处,他们写到:“此前我们的研究显示,一个国家的技能(即它的经济资本)与长期增长率之间存在因果关系,因此能够推测教育政策是如何影响国家的预期经济效益。”简单来说就是,只要能提高考试成绩,就能实现GDP增长。也只有因果关系才能得出这么肯定的结论。

Will Brehm 8:38

Jeremy Rappleye 8:55
抱歉说了这么多,我还要强调最后一点,是关于这些学术研究的主张如何进入到政策领域的。有两个组织功不可没,一个是世界银行(World Bank),另一个是OECD。他们牢牢抓住这些研究成果,并向世界各国的政策制定者积极推广。前者聘请了哈努谢克和沃斯曼因,要求他们研究成果运用到低收入国家的政策制定中去。2007年世界银行发表了题为《教育质量在经济增长中的作用》的报告。后者同样也聘请了哈努谢克和沃斯曼因来分享研究成果、讨论政策意义,并于2015年发布了《普及基本技能:国家能获得什么》的报告。在谈完我们的研究之后,如果访谈结束前还有时间的话,可以回过头来再讨论一下这些组织在推动实证研究进入具体政策建议中的重要作用。

Will Brehm 12:13

Hikaru Komatsu 12:23

Will Brehm 13:17

Hikaru Komatsu 13:31

Will Brehm 13:31

Hikaru Komatsu 13:36

Will Brehm 15:40

Hikaru Komatsu 15:53

Will Brehm 16:28

Hikaru Komatsu 16:42

Will Brehm 17:17

Jeremy Rappleye 17:31
这听上去可能比较学术,不好意思。为方便理解,我举个具体例子吧。比如美国,哈努谢克和沃斯曼因声称,基于其在2000年PISA测试中的成绩,如果有一个20年或30年的改革计划使其提高到2006年PISA测试中的芬兰或韩国的水平,那么美国的GDP会有5%的增长。哈努谢克和沃斯曼因他们还指出,早在1989年,时任美国总统的乔治·布什(George Bush)和各州州长曾承诺到2000年要让美国学生的数学和科学成绩成为世界第一,在那时候,也就是提高50分。他在由布鲁金斯学会(Brookings Institution)出版的另一本题为《濒危的繁荣》的书中称,如果美国在1989年坚持这一政策,切实实现目标而没有分心的话,那么美国的GDP应该比现在多4.5%。那么,这样就能解决所有困扰我们教育系统的分配和公平的问题。抱歉,因为我是美国人,所以用“我们”一词。基于那种超强的因果关系,哈努谢克通过具体的数据表示,(成绩提高)带来的效益会使“在整个21世纪,美国工人的平均收入提高20%。”

Will Brehm 22:02

Jeremy Rappleye 22:25

Will Brehm 22:54

Jeremy Rappleye 23:07

Will Brehm 23:44

Hikaru Komatsu 24:09

Will Brehm 25:14

Hikaru Komatsu 25:25

Will Brehm 25:45

Jeremy Rappleye 26:16
然而,我们的研究恰恰表明那一观点所期待的结果未必会发生。近年来也有其他一些我很喜欢的研究,比如伦敦大学学院教育学院的莫礼时(Paul Morris),他和尤安·奥尔德(Euan Auld)、游韵,以及香港的鲍勃·亚当森(Bob Adamson)等中、青年学者的研究认为,PISA其实更多是由培生(Pearson)、美国教育考试服务中心(ETS)等私企所推动的;史蒂芬·鲍尔(Stephen Ball)、鲍勃·林嘉德(Bob Lingard)和山姆·塞勒(Sam Sellar)等学者在诸多论文里也提出了类似的观点。此外,还有一些学者如拉迪卡·哥尔(Radhika Gorur),他们的研究显示出标准化的危害,以及它将如何最终破坏未来教育变革和创新所需的多样性。这些学术研究都表明PISA和GDP之间并无多少联系。希望我们的研究能再次证明这一点,并帮助政策制定者更清楚地看到学界的这些示警。

Will Brehm 31:59

Jeremy Rappleye 32:30

Will Brehm 34:00
我真的希望你们能打开这扇门,探触到你们脑海中已经形成的那些更深刻更复杂的问题。Hikaru Komatsu和Jeremy Rappleye,很高兴你们能来做客FreshEd,再次感谢你们的分享!

Hikaru Komatsu 34:15

Jeremy Rappleye 34:18

Want to help translate this show? Please contact info@freshedpodcast.com 

Will Brehm:  2:12
HikaruKomatsu et Jeremy Rappleye, bienvenue à FreshEd.

Hikaru Komatsu:  2:16
Merci de nous recevoir.

Jeremy Rappleye:  2:18
Merci, Will. Avant de commencer, laissez-moi vous dire à quel point j’apprécie votre spectacle. Et j’ai tellement appris de lui. Et je vous félicite vraiment d’avoir créé cet espace et de faire des spectacles de si grande envergure, semaine après semaine. Merci beaucoup de nous recevoir.

Will Brehm:  2:30
Merci pour ces aimables propos. Vous avez beaucoup travaillé ces derniers temps pour réfuter certaines hypothèses banales entre les résultats des tests et le PIB – produit intérieur brut. Quelle est la relation normale que de nombreux chercheurs entretiennent entre les résultats des tests ? Que savent les étudiants et le produit intérieur brut ? Quelle est la croissance d’un pays ou quelle est sa valeur ?

Jeremy Rappleye:  2:59
Oui, merci, Will. Je répondrai à la première question ici, je dirais que la compréhension commune de la relation entre les résultats des tests et le projet intérieur brut est que plus vos résultats aux tests sont élevés, plus votre PIB futur est important. C’est l’affirmation la plus simple qui soit et si nous essayons d’être plus précis, il est entendu que plus les gens obtiennent des résultats aux tests dans une population particulière dans des domaines qui sont, disons, pertinents pour la croissance économique, plus le PIB sera élevé à l’avenir. Ainsi, les domaines pertinents ici sont plus susceptibles d’être définis comme les mathématiques et les sciences ou les scores en mathématiques et en sciences, ainsi que la langue dans une certaine mesure, en particulier la lecture. Les domaines exacts que les évaluations internationales de l’apprentissage telles que PISA mesurent, et dans ce sens, nous pouvons également être plus précis sur le type de modèle économique qui s’inscrit dans cette compréhension commune. Plus précisément, il s’agit d’un modèle qui envisage une économie en croissance grâce au progrès technologique. En d’autres termes, plus l’innovation technologique et l’accumulation de connaissances seront nombreuses, plus les taux de croissance économique seront élevés à l’avenir.

Will Brehm:  4:02
Et quel type de preuve existe-t-il, comme les chercheurs disposent-ils de données, de données empiriques qui montrent que cette relation est correcte, que des résultats plus élevés aux tests seront synonymes d’un PIB plus élevé à l’avenir ?

Jeremy Rappleye:  4:19
Oui. Donc, au cours des dix dernières années, par exemple, la base de recherche empirique pour ces affirmations est devenue très forte dans certains milieux. Maintenant, tentez d’être précis, les preuves de cette compréhension commune sous sa forme actuelle, je crois, proviennent principalement de deux chercheurs : l’un est Eric Hanushek de l’université de Stanford et l’autre Ludger Woessmann. Je m’excuse, je ne prononce probablement pas correctement le nom, basé à l’université de Munich. Et je pense qu’Eric Hanushek est apparu sur FreshEd tout récemment, si je ne me trompe pas. Quoi qu’il en soit, leur travail retrace environ 40 ans d’histoire des résultats aux tests, ce qui correspond à 40 ans de croissance économique mondiale. Donc à peu près des années 1960 à l’an 2000. Et lorsqu’ils disent mondial, ils font en fait référence à environ 60 pays dans le monde, principalement les pays à revenu élevé qui ont participé à des évaluations internationales, des tests de performance internationaux de façon constante au cours de cette période. Pour être plus précis, l’historique des résultats de ces tests sur 40 ans regroupe des données provenant de deux tests comparatifs internationaux : les études IEA-SIMS et TIMSS et les récentes études PISA de l’OCDE. Ainsi, pour le PIB, ils utilisent un ensemble de données standard de la Penn World Table et les auditeurs intéressés peuvent trouver des détails complets à ce sujet dans notre document. Mais je pense que le point ici est que lorsque Hanushek et Woessmann examinent la relation longitudinale entre les résultats des tests et la croissance du PIB, ils trouvent une très forte corrélation. Cela signifie que dans 60 pays, plus les résultats aux tests sont élevés, plus la croissance du PIB est élevée, et je pense que Hikarumight en parle plus en détail, la différence, en particulier entre l’idée d’association et de causalité.

Mais ce que je veux dire ici, c’est que ces affirmations empiriques sont si fortes, qu’elles ont donné beaucoup d’élan à l’idée qu’il existe cette base empirique solide pour le lien entre les résultats des tests et le PIB. Et je pense que le travail de Hanushek et de Woessmann est bien expliqué à de nombreux endroits, comme je l’expliquerai peut-être plus tard dans l’interview. Mais le traitement le plus exhaustif se trouve dans un livre intitulé “The Knowledge Capital of Nations – Education and the Economics of Growth”, qui a été publié en 2015.

Will Brehm:  6:43
Vous avez donc indiqué qu’il y a une forte corrélation, mais est-ce que la relation entre les résultats des tests et le PIB est causale ? Et je veux dire, peut-être que c’est un peu entrer dans le langage statistique plus technique qui pourrait être utile ici, pour essayer de comprendre cette affirmation.

Jeremy Rappleye:  7:02
Oui, c’est donc un point très pertinent. Et il est essentiel de comprendre la différence entre une association ou une corrélation et une causalité, je trouve que nous sommes probablement mieux placés pour attendre la discussion d’Hikaro à ce sujet. Mais laissez-moi vous guider en vous donnant deux citations où Hanushek et Woessmann affirment que la relation est causale, et pas seulement la corrélation. La première citation, qui est en quelque sorte une synthèse des résultats de leurs travaux, dit, je cite, “en ce qui concerne l’ampleur, un écart-type des résultats aux tests mesurés au niveau des étudiants de l’OCDE est associé à un taux de croissance annuel moyen du PIB par habitant, supérieur de deux points de pourcentage sur les 40 ans que nous avons observés”. Maintenant, dans cette citation, ils utilisent le mot association, comme beaucoup d’auditeurs l’auront entendu. Mais ailleurs, ils parlent et évoquent sans cesse la causalité. Voici donc la deuxième citation, ils disent : “nos recherches antérieures démontrent la relation de causalité entre les compétences d’une nation son capital économique, et son taux de croissance à long terme, ce qui permet d’estimer comment les politiques d’éducation affectent la performance économique attendue de chaque nation”. Donc, pour faire simple, si vous pouvez améliorer les résultats aux tests, vous obtiendrez une croissance du PIB plus élevée. Et cette certitude découle de l’idée que la relation est effectivement causale.

Will Brehm:  8:38
Ce niveau de certitude doit évidemment avoir un impact sur les décideurs politiques en matière d’éducation, qui ont le droit de savoir que si vous augmentez les scores mesurés par PISA ou TIMSS, vous obtiendrez une plus grande croissance économique. Il semble que cela facilite grandement la vie des décideurs politiques.

Jeremy Rappleye:  8:55
Tout à fait Will, nous pensons que l’attrait, à la fois l’attrait de cette revendication, et l’impact de cette revendication sont en croissance. Et grâce à ce type d’études, les décideurs politiques qui devaient auparavant faire face à une équation très complexe autour de l’éducation sont amenés à croire que les données révèlent qu’une politique de réforme dynamique qui accroît les résultats aux tests conduira, disons, dans 20 ou 30 ans, à des gains économiques importants, assez importants. Et si je peux vous donner juste une autre citation, et je m’excuse pour les citations, je ne veux pas que vous pensiez que j’ai mal formulé ou résumé le travail d’Eric Hanushek et de Woessmann, mais cette citation, celle de leur auteur, illustre le type de bénéfices spectaculaires en matière d’éducation ou de gains économiques que les politiques peuvent espérer obtenir si elles mettent en œuvre ce type de politiques visant à augmenter les résultats aux tests. Ainsi, je cite : “Pour les pays à revenu moyen inférieur, les gains futurs seraient de 13 fois le PIB actuel et une moyenne de 28 % de PIB plus élevée au cours des 80 prochaines années. Et pour les pays à revenu moyen supérieur, la moyenne serait de 16 % de PIB en plus”. Alors maintenant, Will, si vous êtes un décideur politique, vous ne voudriez pas renoncer à ces jeux, n’est-ce pas ? Cette recherche devient donc une véritable motivation pour les décideurs politiques, qui mettent davantage l’accent non seulement sur les mathématiques et les sciences, mais aussi sur les résultats des tests cognitifs dans tous les domaines.

Mais il existe une partie fascinante que je veux souligner et que nous voulons peut-être décortiquer plus tard dans l’interview. Mais on peut s’attendre à ce que ce genre de rétrécissement du champ de l’éducation autour des résultats des tests suscite beaucoup de réticences. Mais en réalité, les gains de PIB résultant de l’augmentation des résultats aux tests, que Hanushek et Woessmann projettent, sont si importants qu’ils devraient permettre de financer tout ce qui concerne l’éducation. Il ne s’agit donc pas vraiment de choisir entre plusieurs alternatives. Mais au lieu d’une politique gagnante à coup sûr contre une politique plus ou moins identique, l’incertitude, l’ambiguïté, la complexité que nous avons vues par le passé.

Et je suis désolé, c’est une longue réponse, Will. Mais je tiens à insister sur le fait que si nous évoquons la manière dont ces affirmations de la recherche universitaire se frayent un chemin jusqu’à la politique ou ont un impact sur la politique, nous devons parler de deux organisations qui se sont vraiment appropriées ces points de vue et les défendent avec force auprès des décideurs politiques du monde entier. La première est la Banque mondiale, qui a recruté Hanushek et Woessmann pour faire le lien entre les résultats de leurs recherches universitaires et l’élaboration de politiques pour les pays à faible revenu. Et ce rapport a été publié par la Banque mondiale sous le titre “Qualité de l’éducation et croissance économique” en 2007. La deuxième organisation qui a vraiment été à l’avant-garde dans ce domaine est l’OCDE, qui a également engagé Hanushek et Woessmann pour partager leurs conclusions et discuter des implications politiques. Et ce rapport était intitulé ” Universal Basic Skills : What Countries Stand to Gain”, qui a été publié en 2015. Donc, peut-être que vers la fin de l’entretien, après avoir discuté de notre étude, nous pourrons revenir sur la façon dont ces organisations sont réellement au centre de la transformation de ce travail empirique en recommandations politiques concrètes.

Will Brehm:  12:13
Quels sont donc les types de problèmes que vous trouvez dans l’analyse de Hanushek et Woessmann sur la relation entre les résultats des tests et le PIB ?

Hikaru Komatsu:  12:23
D’accord, le problème rencontré est le déphasage temporel, c’est-à-dire que le professeur Hanushek a utilisé une période inappropriée pour la croissance économique. C’est-à-dire que le professeur Hanushek compare les résultats des tests enregistrés entre 1960 et 2000 avec la croissance économique pour la même période. Mais c’est un peu étrange. Pourquoi, parce qu’il faut au moins plusieurs décennies pour que les étudiants deviennent adultes et occupent une grande partie de la population active. Donc, de notre point de vue, les résultats des tests pour une période donnée devraient être comparés à la croissance économique des périodes suivantes. C’est le problème que nous avons rencontré.

Will Brehm:  13:17
Ainsi, par exemple, les résultats des tests de 1960 devraient être liés au taux de croissance économique, par exemple, dans les années 1970, il doit y avoir une sorte d’écart entre les deux, est-ce exact ?

Hikaru Komatsu:  13:31
C’est tout à fait exact.

Will Brehm:  13:31
D’accord. Alors, dans votre étude, je veux dire, avez-vous fait cela et qu’avez-vous trouvé ?

Hikaru Komatsu:  13:36
Nous l’avons réalisé, notre étude est très simple. Nous comparons les résultats des tests de 1960 à 2000, ce qui correspond exactement aux données utilisées par le professeur Hanushek. Nous mettons ces données en parallèle avec la croissance économique des périodes suivantes, comme 1980 à 2000 ou 1990 à 2010, ou quelque chose comme ça. Et nous avons trouvé que la relation entre les résultats des tests et la croissance économique était beaucoup plus faible que celle rapportée par le professeur Hanushek. Le public verrait probablement un chiffre sur le web, sur le site de FreshEd, et il y aurait deux chiffres, celui de gauche est le chiffre original rapporté par le professeur Hanushek. Et il existe une relation étroite entre la croissance économique et les résultats aux tests. Alors que le chiffre de droite est celui que nous avons trouvé et la relation est très peu claire. Permettez-moi donc d’expliquer comment nous avons établi cette relation lorsque les résultats des tests de 1960 à 2000 ont été comparés à la croissance économique. Pour les années 1995 à 2014, seuls 10 % des variations de la croissance économique entre les pays s’expliquent par la variation des résultats aux tests. Autrement dit, les 90 % restants de la variation de la croissance économique devraient être responsables d’autres facteurs. Cela revient à dire qu’il est totalement déraisonnable d’utiliser les résultats des tests comme seul facteur pour prédire la croissance économique future. Et c’est ce que le professeur Hanushek a fait dans son étude et ses recommandations politiques.

Will Brehm:  15:40
Ainsi, dans votre étude, vous avez constaté que 10 % de la variation du PIB peut s’expliquer par la variation des résultats aux tests. Quel pourcentage l’étude de Hanushek et Woessmann a-t-elle mis en évidence ?

Hikaru Komatsu:  15:53
Leur pourcentage était sans doute d’environ 70 et nous essayons de reproduire la conclusion du professeur Hanushek, dans notre cas, le pourcentage était d’environ 57 et la différence entre 57, 70 serait due à la différence de version des données que nous avons utilisées. Afin de prolonger cette période, nous avons utilisé une version actualisée de ces données, cet ensemble de données est exactement le même, mais la différence n’est que la version.

Will Brehm:  16:28
Une différence entre 50 et 70 % est plutôt minime. Mais la différence entre 10 % et 70 % est suffisante pour remettre en question les conclusions fondamentales qui sont tirées de ces données.

Hikaru Komatsu:  16:42
Oui, en effet. Si cette relation décrite à l’origine par le professeur Hanushek est causale, nous aurions dû trouver cette relation relativement forte entre les résultats des tests pour une période donnée et la croissance économique pour les périodes suivantes. Mais nous avons trouvé une relation très faible, ce qui suggère que la relation rapportée à l’origine par le professeur Hanushek ne représente pas toujours la relation de cause à effet. C’est là où nous voulons en venir.

Will Brehm:  17:17
Il semble que ce soit un point très fondamental qui pourrait être plutôt déconcertant pour les hypothèses de nombreuses personnes sur l’éducation et sa valeur pour la croissance économique.

Jeremy Rappleye:  17:31
Merci, Will, j’aimerais revenir un peu sur ce que Komatsu sensei a évoqué là-bas. Un des points majeurs à comprendre à propos de ces 70 % de variation est expliqué par les résultats des tests, la puissance de cette corrélation conduit à des recommandations politiques très fortes. Donc, je vais essayer de déballer un peu ce que j’ai dit plus tôt, parce que je pense que c’est un point important, je vais essayer de le faire de manière académique d’abord, et ensuite je vais essayer de donner une version simple qui sera beaucoup plus facile à comprendre pour les auditeurs.

Mais fondamentalement, si vous avez un lien de causalité de corrélation aussi fort, alors Hanushek et Woessmann affirment que vous pouvez d’abord accroître vos résultats aux tests, et cela produira une croissance tellement excessive à l’avenir, que vous pouvez réorienter cette croissance excessive vers tous les autres types de biens éducatifs dont vous avez besoin. Ainsi, en termes d’équité, en termes d’inclusion, vous pourriez même réorienter cette somme supplémentaire vers les soins de santé, vers des objectifs de durabilité, toutes ces choses. Comme nous le savons tous les deux, il s’agit là de deux aspects du camp : l’éducation pour la croissance économique, ou bien les actions, l’inclusion, le développement personnel. Ce sont les types de débats qui ont toujours été menés avec l’éducation en tant qu’étude académique, mais il est capable de transcender, ils sont capables de transcender ce débat basé sur les fortes revendications causales, cependant, juste pour remplir cela de manière académique. L’affirmation veut donc que si 3,5 % du PIB est dépensé pour l’éducation, cela provient du rapport de la Banque mondiale de 2007. Mais si, sur 20 ou 30 ans, vous pouviez accroître vos résultats aux tests de cinq points d’écart type, cela entraînerait une hausse de 5 % du PIB en moyenne, et je cite : “ce dividende brut couvrirait largement toutes les dépenses en matière d’écoles primaires et secondaires”. Cela revient à dire que si vous vous concentrez sur l’accroissement des niveaux cognitifs, des résultats aux tests, vous pourriez obtenir une croissance suffisante pour obtenir en fin de compte plus d’argent pour l’éducation, quel que soit le type d’objectifs éducatifs que vous souhaitez poursuivre. Dans le rapport de l’OCDE de 2015, Hanushek et Woessmann écrivent que “les avantages économiques des gains cognitifs offrent un potentiel énorme pour résoudre les problèmes de pauvreté et de soins de santé limités et pour encourager les nouvelles technologies nécessaires pour améliorer la durabilité et l’intégration de la croissance”. Ainsi, comme je l’ai déjà mentionné, au lieu d’un compromis entre les politiques liées à la croissance et l’équité, les résultats du Hanushek sont si probants qu’ils suggèrent qu’une première hausse des résultats aux tests produira finalement suffisamment de gains supplémentaires pour tout payer. Cela les dispense donc vraiment de s’engager dans le genre de débats sur les priorités qui ont lieu depuis que la politique de l’éducation existe.

Maintenant, si cela paraît académique, je m’en excuse. Laissez-moi essayer de le formuler en termes plus concrets pour le rendre plus facile à comprendre. Bien sûr, si les États-Unis se fondaient sur les résultats de PISA 2000, ils pourraient avoir un plan de réforme sur 20 ou 30 ans qui les amènerait à atteindre le niveau des résultats de PISA de la Finlande ou de la Corée en 2006. Hanushek et Woessmann affirment que le PIB, le PIB des États-Unis sera supérieur à 5 %. Hanushek met en avant le fait qu’en 1989, les gouverneurs des États-Unis se sont réunis avec le président George Bush et qu’ils ont promis de faire de l’Amérique le numéro un mondial en mathématiques et en sciences d’ici l’an 2000. À l’époque, cela aurait donc représenté un gain de 50 points. C’est pourquoi Hanushek affirme dans un autre ouvrage, mais produit par Brookings et intitulé “Endangering Prosperity”, que si les États-Unis avaient maintenu le cap en 1989 et avaient réellement atteint les objectifs fixés plutôt que de se laisser distraire, leur PIB serait aujourd’hui supérieur de 4,5 %. Et cela nous permettrait de résoudre tous nos problèmes de distribution, de, je suis américain, d’excuses, de répartition ou d’équité qui ont constamment affligé l’éducation américaine. Et même en termes plus concrets, Hanushek est basé sur cette forte causalité. Il dit que les gains seraient en fait égaux et je cite : “20% de salaires en plus pour le travailleur américain moyen sur l’ensemble du 21ème siècle”.

Will Brehm:  22:02
Ainsi, il lit l’avenir avec cette relation qu’il suppose pouvoir lire l’avenir et dit en gros qu’il faut se focaliser sur les résultats des tests d’abord et se préoccuper de tout le reste ensuite, parce que nous allons tellement augmenter le PIB que nous pourrons payer tout ce dont l’éducation a besoin. C’est en quelque sorte l’essence même de cette politique.

Jeremy Rappleye:  22:25
C’est bien cela, Will, et il a fait ce calcul pour les États-Unis, comme vous pouvez l’imaginer, il est basé aux États-Unis. Mais si vous regardez le rapport de l’OCDE, en 2015, ils font en fait les mêmes calculs pour tous les pays du monde. C’est pourquoi nous mettons l’accent sur ce point dans notre document intégral en introduction, en reprenant le cas du Ghana, parce que l’OCDE a vraiment repris le cas du Ghana et a dit : “Vous auriez cet énorme gain économique si vous pouviez simplement maintenir le cap sur l’amélioration des résultats aux tests.

Will Brehm:  22:54
Il s’agit donc de prévisions ou de projections futures. Y a-t-il jamais eu un exemple concret de la façon dont Hanushek et Woessmann théorisent ?

Jeremy Rappleye:  23:07
Eh bien, je suppose qu’ils pourraient probablement soutenir que l’élément de données lui-même montre que cela fonctionne. Il existe donc bien évidemment des variations à la hausse et à la baisse pour chaque pays, mais je pense qu’il serait difficile de montrer un pays en particulier ou de vous donner le cas d’un pays qui a mis en œuvre des réformes et qui a ensuite enregistré une croissance plus élevée de son PIB. Tout est ramené au niveau de la corrélation et de la causalité plutôt qu’aux termes concrets de pays particuliers.

Will Brehm:  23:44
Votre analyse indique évidemment que ces projections sont incorrectes, ou ne fonctionneraient pas nécessairement comme le prétendent Hanushek et Woessmann. Vous savez, comment pouvons-nous commencer à théoriser que ce lien entre les résultats des tests et le PIB, comme par exemple quel type d’implications votre étude a sur les décideurs politiques de l’éducation ?

Hikaru Komatsu:  24:09
À mon humble avis, ou selon les données, on peut affirmer que l’amélioration des résultats aux tests entraîne une croissance économique plus élevée en moyenne, mais en réalité, les résultats aux tests ne sont qu’un facteur parmi d’autres, car nous avons découvert, comme nous l’avons dit, que seulement 10 % de la variation de la croissance économique s’expliquait par la variation des résultats aux tests. Donc, en ce sens, les résultats des tests éducatifs ne sont qu’un facteur influençant la croissance économique. Et comme nous le voyons, il existe une énorme variation du capital fiscal, des terres ou des entreprises entre les pays, et cela devrait affecter la croissance économique. L’éducation n’est donc qu’un de ces facteurs. C’est, je pense, une compréhension raisonnable de la relation entre la croissance économique, l’éducation et les effets sur nous.

Will Brehm:  25:14
Ainsi, dans un sens, nous reconnaissons que l’éducation joue un rôle dans la croissance économique future. Mais il existe d’autres facteurs qui influent également sur la croissance économique future. Et nous ne devrions pas les perdre de vue non plus.

Hikaru Komatsu:  25:25
Vous avez raison. Le problème du professeur Hanushek est donc qu’il croit ou suppose que c’est le seul facteur et qu’il n’utilise que les résultats des tests pour prédire la croissance économique future. Mais le chemin n’est pas si simple. C’est là où nous voulons en venir.

Will Brehm:  25:45
Et on dirait que c’est un simple point de vue. Mais comme je l’ai dit, je pense que c’est assez sérieux, parce que cela bouleverse vraiment les objectifs éducatifs des pays. Je veux dire que cela remet en question la montée en puissance du PISA, je veux dire, est-ce que tous ces pays qui essaient de rejoindre le PISA ? Est-ce vraiment ce qu’ils devraient faire, n’est-ce pas ? Je veux dire que pour moi, les implications politiques deviennent beaucoup plus difficiles, et la confiance de certaines prescriptions politiques disparaît en quelque sorte.

Jeremy Rappleye:  26:16
En effet, Will. Nous sommes tout à fait d’accord avec ce résumé des implications de notre étude, c’est une idée simple. Et il est assez évident, même pour, disons, des étudiants de niveau master, que l’éducation n’est pas si complexe. Mais l’un des problèmes est que toutes ces grandes données créent un potentiel pour ce genre de combat de chiens en utilisant des données de la haute stratosphère. Et les gens ne peuvent pas vraiment y toucher. Donc, tant que vous êtes là-haut à vous battre, alors cela semble assez solide. Et je pense que pour être très honnête, en fait, ni Hikaruor ni moi n’aimons vraiment faire ce travail à ce point. C’est vrai. Pour être très honnête, nous n’aimons pas vraiment ce travail. Nous préférons penser à de grandes idées et à des idées complexes et faire ce genre de choses.

Mais le problème est que cela limite la vue, les vues simples de l’éducation limitent la complexité ou la profondeur de ce que nous devrions voir dans ces écologies de l’éducation ou ce genre de choses. C’était une sorte de réponse globale. Mais si nous avons le temps, j’aimerais parler plus précisément des recommandations politiques spécifiques et des implications de nos études. Nous pensons donc que les résultats de nos recherches comportent des recommandations, mais il ne s’agit pas vraiment de recommandations au sens habituel du terme, c’est-à-dire l’identification d’une meilleure pratique ou d’une solution miracle ou d’une potion magique qui améliorera l’éducation dans le monde entier. Au contraire, les implications de nos recherches sont ce que l’on pourrait appeler des recommandations politiques négatives. Et par là, je veux dire qu’elles aident les décideurs politiques à réaliser ce qu’ils ne devraient pas faire. Plus précisément, elle leur dit qu’ils ne doivent pas se laisser séduire par des promesses selon lesquelles le fait de se concentrer sur l’augmentation des notes d’examen, et uniquement des notes d’examen dans des domaines tels que la science et la technologie, les mathématiques est une politique infaillible qui augmentera les taux de croissance du PIB, et elle leur dit de ne pas croire le genre de conseillers qui viendraient leur dire que l’augmentation des notes d’examen à elle seule conduira à un PIB suffisant, à une croissance future pour citer “et aux problèmes financiers et de répartition de l’éducation”, pour ne pas citer. Je voudrais maintenant être encore plus précis sur ce point, plus concret, sur deux points. Comme la plupart des auditeurs le savent, l’une des plus grandes tendances en matière de politique de l’éducation au cours des deux dernières décennies a été l’enquête PISA. Et il est actuellement prévu d’étendre le PISA aux pays à faible revenu dans le cadre de l’exercice PISA pour le développement. Je pense que d’ici 2030, l’OCDE et la Banque mondiale prévoient de mettre en place le PISA dans tous les pays du monde, en dépit de toute une série de critiques émanant d’universitaires, de praticiens, de la croyance normale selon laquelle l’éducation est plus complexe que cela.

La principale raison de l’extension du test PISA est qu’elle mènera à une plus forte croissance du PIB. En effet, les pays sont amenés à s’inscrire au PISA, en raison des types d’allégations que nous avons analysées tout au long de cet entretien. Mais nos recherches indiquent que ce ne sera pas le cas. Certaines de mes recherches préférées de ces dernières années ont été menées par des chercheurs autour de Paul Morris, à l’Institut de l’éducation de Londres, en collaboration avec de jeunes universitaires, Euan Auld, Yun You ou Bob Adamson à Hong Kong, ce qui illustre le fait que les tests PISA sont en réalité beaucoup plus pilotés par une série d’entreprises privées telles que Pearson, ETS, etc. Et des universitaires comme Stephen Ball, Bob Lingard et Sam Seller écrivent également de très bons articles dans ce sens. Une autre ligne de recherche importante émane de personnes comme Radhika Gorur qui écrit sur les dangers de la normalisation et sur la manière dont elle pourrait finalement détruire la diversité nécessaire à l’adaptation et à l’innovation futures dans l’éducation. Nous voulons donc, une fois de plus, que nos recherches fassent disparaître la croyance selon laquelle les recherches universitaires prouvent d’une manière ou d’une autre le lien entre le PISA et le PIB, et que les décideurs politiques perçoivent beaucoup plus clairement tous ces avertissements. Et si vous me permettez de passer rapidement à la deuxième dimension du concret, ce qui se passe maintenant, comme beaucoup d’auditeurs le savent aussi, c’est que le monde est, le monde de l’éducation, où le développement parle plus globalement des objectifs de l’après 2015. Fondamentalement, ce qui suit les objectifs du Millénaire pour le développement que nous allons dans les années 1990. Et en matière d’éducation, le quatrième objectif de développement durable est celui qui concerne l’éducation, il fixe des objectifs au niveau mondial pour améliorer l’apprentissage d’ici 2030. Et l’une des choses regrettables que nous avons remarquées dans ces discussions, c’est qu’il semble que la discussion imite l’OCDE et la Banque mondiale ; c’est-à-dire que nous voyons l’UNESCO et d’autres agences se référer explicitement aux études de Hanushek et Woessmann pour défendre les raisons pour lesquelles les évaluations de type PISA sont le meilleur moyen d’atteindre le quatrième objectif de développement durable. Ainsi, par rapport aux discussions autour de l’EPT, au début des années 1990, la discussion autour du quatrième objectif de développement durable semble prendre cette affirmation de capital de connaissances comme une vérité, une vérité académique.

Et nous nous sommes inquiétés de ce que cela mette le monde entier sur la voie de la mise en œuvre de tests de type PISA. Et, bien sûr, le changement de programme qui s’ensuit. Je ne veux pas, vous savez, un peu trop exagérer. Mais nous craignons qu’il n’y ait pas vraiment de preuves à cet égard et qu’il s’agisse d’exercices très coûteux qui, en fin de compte, ne contribueront que très peu à améliorer l’éducation. Nous espérons donc, une fois encore, que notre étude donnera aux décideurs politiques le type de base de recherche universitaire qui leur permettra de résister aux avancées de l’OCDE et de la Banque mondiale.

Will Brehm:  31:59
Avez-vous fait l’expérience d’un recul par rapport à certaines des conclusions parce que, de toute évidence, vous remettez en question certaines des idées reçues par la Banque mondiale, par l’OCDE, par des entreprises privées, comme vous l’avez dit, par Pearson et par ETS, le Service d’évaluation de l’éducation, qui produit toute une série de tests. On pourrait donc supposer que vos recommandations négatives qui découlent de vos conclusions pourraient finalement créer une réaction de rétraction de la part de ceux dont les intérêts sont contestés.

Jeremy Rappleye:  32:30
Oui, je crois que nous aimerions vivement avoir un retour en arrière, parce que le retour en arrière implique un engagement explicite. Encore une fois, nos observations ne sont pas nouvelles, à savoir qu’il n’y a pas de lien entre les résultats scolaires et la croissance du PIB. Ces affirmations sont très, très anciennes, cette idée est en réalité très ancienne. Mais ce qui se passe, c’est qu’à chaque vague de données qui sortent de ce genre de combat de chiens dont je parlais, cela va des enfants qui se lancent des pierres depuis différents arbres jusqu’aux ballons à air chaud en passant par les avions à réaction, et cela ne cesse de croître. Il n’y a pas de véritable engagement avec les réalités du terrain qui réfuterait tout cela. Donc, si nous obtenions un retour en arrière, nous l’accueillerions avec plaisir. Nous aimerions voir les preuves parce que notre décision n’est pas prise. C’est tout à fait possible. Je veux dire qu’il n’y a pas de certitudes, et il serait merveilleux de voir une discussion plus élaborée autour de ces idées. Nos résultats sont donc concluants. Mais dans le sens où avec cet ensemble de données, ils réfutent de manière concluante une hypothèse ou une affirmation particulière, mais ils ne sont pas concluants en termes de fin d’apprentissage. Nous pouvons toujours en comprendre davantage sur la relation complexe entre la société, l’économie, la culture et ce genre de choses. Nous serions donc très heureux de pouvoir élaborer sur ce point.

Will Brehm:  34:00
J’espère vraiment que vous pourrez ouvrir cette porte à un engagement beaucoup plus profond pour aborder certaines de ces grandes questions que vous avez évidemment à l’esprit, mais HikaruKomatsu et Jeremy Rappleye, merci beaucoup d’avoir rejoint Fresh Ed. C’était vraiment un plaisir de vous parler aujourd’hui.

Hikaru Komatsu:  34:15
Merci de nous recevoir. J’ai vraiment apprécié.

Jeremy Rappleye:  34:18
Merci infiniment, Will. Continuez votre excellent travail. Nous apprécions tous le dur labeur que vous accomplissez au nom des chercheurs en éducation et des praticiens de l’éducation du monde entier.

Translation sponsored by NORRAG.

Want to help translate this show? Please contact info@freshedpodcast.com 

Have any useful resources related to this show? Please send them to info@freshedpodcast.com

Rwanda is perhaps most well-known for the genocide it experienced in the 1990s. In its post-conflict development, the country has had to balance colonial legacies, state centralizing tendencies, and the zeitgeist of neoliberalism. This has made for a careful balancing — one that has left the government regulating the society and economy while simultaneously reducing its responsibility to citizens.

In education, this balancing act manifests in the government’s three aims: credentials, controls, and creativity. The education system is based on credentials awarded through examinations, a colonial hangover, and controls students as part of the state’s centralization efforts; yet, somehow, the system promotes creativity so students can pursue a learner-centered education tailored to their own needs, preparing them for the 21st century labor market of precarious work.

My guest today, Catherine Honeyman, has a new book that explores Rwanda’s opportunities, challenges, and paradoxes in post-conflict development through the policy of mandatory entrepreneurship education, which is believed to be the country’s beacon for economic growth. Catherine Honeyman is a visiting scholar at the Duke Center for International Development and Managing Director of Ishya Consulting. Her new book, The Orderly Entrepreneur, takes us inside both policy making circles and classrooms to understand part of Rwanda’s social transformation. The Orderly Entrepreneur received an honorable mention from the Globalization and Education SIG’s 2016 Book Award.

Citation: Honeyman, Catherine, A., interview with Will Brehm, FreshEd, 64, podcast audio, March 13, 2017. https://www.freshedpodcast.com/catherinehoneyman/

Transcript, translation, and resources:

Read more


What is the connection between education and the economy? For many neoclassical economists, the connection is found in Human Capital theory.

My guest today, Professor Steve Klees, thinks human capital theory and rates of return analyses are very problematic.

In our conversation, Steve talks about his new article, “Human Capital and Rates of Return: Brilliant Ideas or ideological dead ends?”, which can be found in the latest issue of the Comparative Education Review. He takes us through human capital theory, its internal logical fallacies, and proposes a set of alternatives.

Steve Klees is a professor of International Education Policy in the College of Education, University of Maryland.

Citation: Klees, Steve, interview with Will Brehm, FreshEd, 54, podcast audio, December 5, 2016. https://www.freshedpodcast.com/steveklees/

Will Brehm  1:45
Steve Klees, welcome to FreshEd.,

Steve Klees  1:52
I’m very glad to be here. Thank you for having me.

Will Brehm  1:55
You are an economist by training. But you have spent most of your career in the comparative and international education field. How do economists typically think about, or look at education?

Steve Klees  2:12
That’s a huge question, but the answer really depends on what kind of economist you’re talking about; they’re different schools of thought. But the the main dominant school of thought is called neoclassical economics. And neoclassical economics is really about competitive capitalist market systems. And within that education is a very important piece of understanding education, economics and development. In particular, neoclassical economists have developed something they call “human capital theory” that is a framework for understanding education’s role in the economy and in society.

Will Brehm  3:10
And how is human capital theory measured? How do economists see human capital?

Steve Klees  3:22
Well, human capital is a latecomer to economics, to neoclassical economics. Neoclassical economics goes back to Adam Smith, and the Wealth of Nations in the 1700s and the term “neoclassical” actually was coined at the end of the 1800s, and it’s about how a competitive market system operates. Human capital theory wasn’t developed till the late 1950s, early 1960s. Prior to human capital theory, economists understood the economy in terms of supply and demand, you’d always see economists talking about supply and demand – of small companies, small firms, small households competing with each other – and trying to understand how that competition worked, what you got out of it, how a market system worked.

Prior to human capital theory, economists had a lot of difficulty understanding labor and work. Labor prior to human capital theory was an anomaly. It wasn’t something you could talk about in terms of supply and demand. The economists in those days, in the 50s, look more like sociologists; it was a whole field of labor economics where they studied real world labor, they studied strikes, they studied unions, they studied how large firms operated. But education didn’t really fit into that structure at all – that way of thinking. And there were odd people out in neoclassical economics because they were more like sociologists, and they weren’t talking about competitive market structures and supply and demand, and human capital changed all that

Will Brehm  5:28
How so?

Steve Klees  5:29
Well, it really made economists able to talk and think about education and labor. Labor, especially, as a commodity like any other commodity that’s bought and sold on a marketplace, that has a price, that’s determined by supply and demand in the marketplace. Human Capital theory developed because it was explaining puzzles. People were trying to understand how economies grew. And they understood that there were more workers, and there was more capital investment, but they didn’t really have any idea about quality of work. And the whole idea of human capital, was it explained better, to neoclassical economists anyway, why some countries grew faster versus slower. They called it a revolution in thought, and the idea behind it was essentially simple – that education wasn’t just a consumer good, it was an investment. It was an investment in individuals, and it was an investment by society and societal development.

Will Brehm  6:48
So in a sense, it would be that if an individual were to receive education, or more education than another, they or he or she would be more productive in an economy, and maybe measured through income? Is this the way the neoclassical economists were seeing this?

Steve Klees  7:12
Yes. They looked at two outcomes of education in particular They looked at earnings, and they weren’t interested in private benefits as much; earnings were a benefit to you, income is a benefit to you. But they were interested in, within their framework earnings as a proxy for people’s productivity, like you said. And so, they were trying to get a handle on education’s connection to individual productivity. And secondly, education’s direct influence on economic growth, its effect on gross national product. So you saw starting in the 60s, lots of studies of the “rate of return”, they called it, the return on investment. So education in terms of earnings as a proxy for productivity, and in terms of economic growth measured by gross national product.

Will Brehm  8:15
So based on the rate of return methodology is is some education better than other education for foreign economy or for foreign individuals’ productivity?

Steve Klees  8:27
Yes. I should explain a little bit about rates of return. Rates of return are a measure of benefits and costs. In neoclassical economics, the private sector is motivated by profit. Profit is a signal that this endeavor is valuable. Adam Smith talked about the invisible hand of supply and demand working in the public interest. That’s the profits supposedly representing where peoples’ benefits exceed their costs; where the outcome of whatever you’re making, tables or software or whatever, the benefits exceed the costs. And so economists were looking for something as analogous to that in the public sector. So the idea was to explicitly study the benefits and costs of public sector activities, whatever field, education, health care, environment, transportation, and rate of return is a summary measure, after you figure out what are all the benefits to an education investment, what are all the costs of that investment, and it’s a summary measure to try and get at  – gosh, you know, you’re making 20% on your investment, the benefits exceed costs by 20%. And so that’s applied to lots of different types of educational activities, and other sector activities, to study the returns to education, of various types of vocational education versus academic education; of higher different levels of education, higher education versus early childhood, or primary education; different programs of education. Anything where you can find reasonable monetary measures of outcomes. Sometimes you can’t do that, you’re just looking at test score differences between different programs. And then economists do a more limited array of what they call “cost effectiveness” analysis. But mostly economists really like to go after cost-benefit analysis, because that gives them a metric that they can compare with returns in the private sector: Is this a better investment to take your money out of the private sector, tax it and put in an education or health care or environmental protection?

Will Brehm  11:13
This sort of thinking of cost-benefit analysis of education to an economy, do you see this is problematic in anyway?

Steve Klees  11:24
Yes, the paper you mentioned that I did, and actually much of my work over the last – I hesitate to say it – 40 years (I’ve been working in this field for a while) has been with the problems of neoclassical economics, generally. And more specifically, with the internal dynamic, the internal problems with that field that gives you measures like benefit-cost studies of rates of return. My work has been recently basically saying that even not taking in a critical outside neoclassical economics look, which we can talk about in a little bit, of political economy perspective, for example. But even if you take the neoclassical economics perspective, there are so many problems within that framework, that for me, the benefit-cost analysis/rate of return type measures just fall apart; that they become almost meaningless.

Will Brehm  12:41
How, so? Let’s dig into it, human capital theory, rates of return analyses. If you’re saying that there are problems of the internal logic of neoclassical economics for human capital theory, and for rates of return analysis, can you can you dig more into that? Like how, so? What are some examples of this?

Steve Klees  13:05
I don’t know whether to start with the details or the broader picture. Let me just start with the broader picture, because I think, then the problems with the details become clearer. And the broader picture really revolves around one central idea of neoclassical economics, and that’s the idea of economic efficiency or societal efficiency. They sometimes called it Pareto efficiency after an Italian economist a century ago called Pareto, and it’s a complex idea that I find completely unsound and unreal. And I’ll try and explain the idea briefly, explain why I think it’s unsound, and then give you how it manifests itself in this cost-benefit/rate of return type studies. So efficiency is something, you know, it’s a common sense concept. So to us, people talk about efficiency of this or that; it’s an engineering concept, it’s a physics concept of, you know, you can do more with less somehow. But in education, you can talk about it sensibly, right? Limited ideas of efficiency, like you can talk about an educational system as inefficient because it has a lot of dropouts, or a lot of people repeating grades, or a lot of people who aren’t learning much. So there’s a common sense idea of efficiency that makes sense to all of us. And I have no objections to that. It’s the economist concept of efficiency that’s problematic. And that’s not about an individual sector or individual project, as much as it’s the overall society is deemed economically efficient if it operates according to the assumptions of a very highly competitive market framework that in abstract neoclassical economics discussion is called “perfect competition”. Perfect competition is a competitive system that is so highly competitive that you’ve got many buyers and many sellers of identical products: that nobody’s big, nobody can influence prices, they’re all taking prices in the market, they’re all small potatoes. Consumers and producers are the two major motivators and movers of the economy. Consumers are just out there maximizing their happiness, and producers are just out there maximizing their profits. And if everything functions according to, and information is perfect, you know everything about everything. If you operate according to these simple assumptions, the whole economy is deemed efficient. And what they mean by that is that somehow, not only is there no waste, you’re you’re doing everything as cheaply as you possibly can. But you’ve got the right balance of everything, the correct balance of everything. So you’re producing the right amount of chairs and tables, and movies and hamburgers, and software. There’s something called “correct balance”. And that’s what’s efficient in this. And it’s completely separable from their other major concept, which is equity or fairness or the distribution of these things. So the distribution is sort of irrelevant to efficiency. You can have an efficient society in which half the people in the world are starving. That’s that can be efficient, because efficiency is just about those people who have effective demand, meaning they have money. And they can wave that money in the marketplace and demand goods and services. And so efficiency is really to economists about forgetting equity, forgetting distribution. Are we producing as much as possible with the inputs of land/labor/capital/technology that we have? In the theory, and in practice, this is just simply absurd.

There’s actually in theory neoclassical economists have something called “second best theory”. And second best theory says that if you don’t live in the first best world of perfect competition with all those tight assumptions – unreal, impossible assumptions – but let’s say have one monopoly in one sector in which everything else is highly competitive, second best says in the second best world with just one imperfection, you have no idea if the economy’s efficient at all, there’s no idea if it’s close to efficient at all. Because this framework is so tight that you only get this overall efficiency of the correct balance of things if prices are the accurate signal sending benefit and costs signals to producers that act in the public interest. With one price off, all of the prices are affected. So in practice, efficiency demands, for example, that you have the correct balance, the correct inputs balance of producing yachts for rich people and rice and beans for poor people. Well, that’s just a distributional issue to me. That’s an equity issue to me. There’s no right balance of yachts and rice and beans; there’s no right balance of computer software, higher education, early childhood education, nutritional programs, roads building. There’s no correct balance of that. And in practice, there’s just no vantage point in the sky. That’s what this efficiency idea is. Where you could separate what we produce from who gets it. They’re all integrally tied in practice. And this is what in practice they’re trying to do with cost-benefit analysis of rate of return: get an idea of whether something in particular is efficient or not.

Will Brehm  20:06
Just hearing that, it just makes me think that the theory of the world in neoclassical economics doesn’t match the reality that I live in. I mean, certainly people do not have perfect information when it comes to to buying anything. But at the same time, I also think that this separation of equity and distribution from efficiency seems to have actually happened. The world I see today, there seems to be a huge gap between the rich and the poor, that those eating rice and beans and those on their yachts.

Steve Klees  20:48
Yes, absolutely. The real world today is very problematic in terms of distribution, as we all know. In terms of equity.

Will Brehm  21:04
How is this embodied in rates of return?

Steve Klees  21:08
The whole efficiency framework is translated into guidelines for the public sector through cost-benefit analysis. And not costs and benefits to private individuals they’re after, they’re after costs and benefits to society as a whole. Because they want to correct the market, to account for all the costs and benefits to society as a whole. And so if you’re thinking about education, you think about the benefits of education. They buy benefits in terms of earnings to an individual. That’s a benefit to society if earnings reflect productivity. That’s problematic because earnings are a price, and prices are distorted in real world economies. So there’s no reason to believe that earnings reflect productivity at all. Earnings are determined by market power, by the vagaries of who’s got skills and who doesn’t have skills, on where firms do their business. So the idea of earnings as a proxy for productivity is a problem.

A second problem is that even if you wanted earnings as a proxy for productivity, that’s just one individual benefit. That is a social benefit, because it measures productivity, but there are what they call externalities. There are benefits to other people who were not included in the market transaction, you decide how much education you’re getting, some supplier gives it to you, a public school or private school, a training program, the benefits to education go way beyond you. And those are not taken into account in the market. So when there are benefits beyond the individual, they’re called externalities, because the market doesn’t take them into account. And therefore, the market is making inefficient decisions, because it’s not counting all the benefits. So in education, you can think of lots of benefits that aren’t just to you, your education benefits other people through your coworker productivity, through your family, through household health decisions, through helping your children, through lowering crime rates, through lowering welfare rates. It’s got lots of these external effects. And the second problem with rates of return is measuring those all are very problematic.

And the third problem with rates of return is that even if we were trusting earnings as a good measure, it’s very hard, I would say impossible, to figure out the effect of education on earnings. This goes to our problems, not neoclassical economics, but our problems with research methods generally. Separating our causes from effect, impact evaluation is extremely difficult to the point where I think it can’t be done quantitatively. If, for example, you took 1000 people and you ask them what’s their income, and then you try to figure out what are the dozens of factors that make those incomes different. And then you’ve actually tried to build a mathematical model that would separate those dozens of factors, so that you could say, “Well, their income went up, because they were a union this much, because they had another year of education this much, because they were in a high demand field this much, because they were healthy that much.” I mean, it boggles the mind. I’ve done another paper on the economist statistical procedure called regression analysis that tries to do that – it tries to take the dozens of factors affecting some outcome and separate them out. And my view is that we just can’t do that. So that even the minimal idea of looking at the impact of education on individual earnings is problematic. Taking them all together, I find rates of return and cost-benefit analysis, generally not a good basis for decision-making.

Will Brehm  25:48
But yet, it has been. These methods and this particular theory have been dominant and have been used to make decisions in education systems, among other sectors. So what’s the scholarly track record of those using rates of return and human capital theory? All of the critiques that you put forward seem very plausible to me, but yet rates of return and human capital theory has had quite the long longevity in academic research.

Steve Klees  26:23
Yes, it’s certainly has. And just two things to say in response to that. One is you asked about their track record, and in terms of track record, there’s no testing this. This isn’t something you can predict and then find out was a true. Because, I say the rate of return to expanding higher education is 12% in your country at this time. Is that a good investment, if you decide 12% is a good return, and you put your money in. But there’s no validation of whether you got 12% or not, so there’s no track record in terms of these predictions. They’re making predictions now, for example, about education and GNP. And I just find those scary and absurd. There’s some very interesting economists, very competent economists doing this. Eric Hanushek and [Ludger] Woessmann, and they tried to do regression analysis to say if your PISA scores go up (the PISA being that international test that people take as a proxy for cognitive achievement), how much will your GNP go up? And they come to these conclusions like a 10% increase in cognitive skills gives you a 2% boost on GDP and if everybody moved a standard deviation on PISA, your GNP would grow seven times in the next 30 years. I mean, this is carrying this framework to an absurdity to me. They can’t separate out the impact of education from the dozens and dozens of other factors that influence GDP or GNP, and then to take that out as its influence now and project that 30 or 40 years in the future is just the height of irrational use of a framework of this kind. And I understand why they do it. These are reasonable people, you want good information for decision-making, and to economists, this idea of efficiency, separable from equity is the touchstone.

But the real question for me is the second question you asked: basically, why has this framework been dominant for so long? And neoclassical economists would say at the very simple answer – it’s because of its explanatory power. It explains differences in investments in health and differences in individual behavior. And it’s true, I mean, in differences in individual behavior, this is useful framework because your decisions are affected by the returns to you, and you make decisions about your educational investment versus your investment in health care, versus your decision to go to a movie or your decision to buy a house, about returns to you. And that’s fine. So there is some use for this framework in terms of understanding people’s motivations. But in terms of societal’s efficiency and investment preferences, this framework is bankrupt and it’s empty. So to me, and to many critics, it’s not the dominant framework because of its explanatory power, it’s the dominant framework because it fits. Human capital theory is embedded in neoclassical economics, and that’s embedded in capital market – it fits with a capitalist market economy. The critics would argue the reason there’s so much attention to efficiency and rates of return and technical views of whether you invest in this thing, or that and how much do you do is because it makes sense in terms of efficiency. If you lose that efficiency framework, you realize that this is just a way to support a market system. This is neoclassical economics is an ideological justification for capitalist market systems to be efficient. To act in everybody’s interests aside from equity. If you question that, then you can see neoclassical economics generally, and human capital theory, as basically an ideological framework and ideological bulwark.

The whole skills discourse today comes from human capital theory and a skills discourse seems like common sense. It says if people only had better skills, they would be better off, and their countries and societies would be better off. That skills discourse based as it is on human capital theory and neoclassical economics is very problematic. People today have talked about the “triple economic challenge” that we face, and they talk about the three things: job creation; poverty elimination; and inequality reduction. Human capital theory and neoclassical economics generally gives one simplistic answer to all three challenges: lack of skills or equivalently, the mismatch between what education is producing and what businesses in the economy need. For the critics, lack of skills is not why people are poor, are not why jobs are scarce, and not why societies are so unequal. The culprit for the critics is that the very structure of the world system in which we are living, capitalism most particularly, but patriarchy, racism and other structures. Those very structures are problematic. While capitalism has increased our ability to produce material goods tremendously, so it looks very productive in that way, in another sense, it’s one of the most inefficient and destructive structures that you can imagine. Why? Because almost half the world – the World Bank says 3 billion people – are relegated to the margins of society. Capitalism has not created jobs for them, livelihoods for them. For the vast majority of our global population, if capitalism was an efficient system, we would be taking advantage of the skills and develop the skills of the 7 billion people on the planet and produced a lot more. Capitalism in its 200, 300 years hasn’t done that, and isn’t doing that. You know, some people talk about we live in a meritocracy. What nonsense! These 3 billion people are relegated to the margins of society because they’re not meritorious? It’s not that at all. It’s poverty, unemployment and inequality, not to mention environmental destruction and other problems, are not failures of capitalism, as they’re sometimes seen, but the logical outcome of its inherent structure. So that that in many ways, contrary to prevailing economic views, human capital has been a very destructive discourse. This is contrary to what the majority of economists think as it’s been brilliant, but it’s been a destructive discourse, because it’s really blamed individuals for their lack of skills, their lack of investment in the right skills, the lack of good choices. And so instead of understanding problematic structures that we need to do something about, we’ve been directing attention towards the supply of individuals and how to fix that. And we’ve been fixing it for decades. And the payoff with poor countries is abysmal. And the payoff even within rich countries is abysmal. The inequalities within the US, the level of hunger in the United States, the level of marginalization, the level of poor dead end jobs, the level of insecurity, the level of environmental destruction, this is not an efficient system.

Will Brehm  36:04
Turning to alternatives. I mean, is it even possible, or can we even have an education system in a capitalist economy without human capital theory? It almost seems like many of these problems that we see in education in terms of equity that you were just mentioning stem from the capitalist economic systems that that are pervasive in most countries. So how can we envision and create education systems in alternative ways that account for equity while still being in capitalist economies?

Steve Klees  36:48
Yes. All you’re asking today’s tough questions.

Will Brehm  36:55
I apologize.

Steve Klees  36:57
That’s okay. Let me preface my response to education with a little bit on how these alternatives are viewed to the very structures in which we live. Because education can only be successful if it’s a part of a challenge to those structures in fundamental ways. And there’s a lot of alternatives. Everything is contested terrain in this world; everything is up for grabs, up for debates with different views. So I view the alternative to neoclassical economics as what I call “political economy”. Political economy is a contested term and people on the right use it as well as people on the left. I’m using it more from the left of center point of view of critics of capitalism, critics of other world system structures. And for me, a political economy perspective today raises questions about the structures of the world system in which we live.

It’s the intersection of feminist perspectives, of post- perspectives, postcolonial, post structural, neo-Marxist perspectives, queer theories about heterosexism in society, disability theories, critical race theories. Not that these theories are identical, not that these approaches are identical, but all of them see marginalization as central, and all of them see marginalization not as failures of the world system -they’re failures for sure – but more is a logical consequence of the structures of patriarchy and racism and capitalism in which we live. And while there’s agreement that reproduction is pervasive, that is this marginalization is not an aberration, systems are out there that reproduce and legitimate marginalization and inequality. And the education system is part of that, as are all of the systems in which we live.

But the critics, the political economists, as I label them agree that while reproduction is pervasive, there are lots of spaces for progressive action. Through exercising individual and collective agency. You have the ability, and especially collectively, we have the ability to challenge these structures. And collective challenge is perhaps the watchword of political economists. Social movements like the women’s movement, like the civil rights movement. These are worldwide now. Like the landless movement in Brazil and now other countries. The poorest people in the world are organized and having an influence on policy. The untouchable movement in India, not anti-globalization, but the alter-globalization movement and in human rights movements and the children rights movement. And so there’s lots of examples of contestation at the systemic level and in education.

There’s lots of examples in every education in every city, in every country, and in every school system of what political economists call more progressive approaches to education. The legacy of people at Paolo Freire, the famous Brazilian educator who founded a field that we call today “critical pedagogy”. Critical pedagogy is a political economy approach to education, arguing that while reproduction is pervasive in schools, there’s lots of ways to challenge that. And so people, individual teachers challenge that all the time. They close their doors to their classroom, they use different learning materials, they teach their students differently about fairness, about equity, about the structures in which they live, they raise questions on that. And it’s not just individual teachers, there are systems of it. In Brazil, the landless movement which I just mentioned, have their own schools that are Freirean, that are participatory, that are so different from the technicist technical approach to education that we have today throughout the world. In Brazil, something called the citizen school movement that, again, is very participatory, that involves the community. We say “community involvement” all the time, but this is serious community involvement. This is serious democracy for students, for teachers, for administrators, participating and directing curriculum, directing grading, making decisions at a local level together, and sometimes very explicitly challenging the types of feeding education into work and into the labor market that dominates so strongly. On the alternative, for most political economists, when you reject the sort of functionalist view of sociologists, of society, of efficiency of markets, and say, “This is not something in which everybody is benefiting, there’s conflict here, there’s different interests, and the only way that’s going to change is through struggle, through individual and collective struggle.”

What that means in terms of alternative system-wide is difficult to say. At a minimum, where we’re not a neoliberal form of capitalism. Capitalism in the 60s, and 70s was a much more liberal capitalism in which government intervention was recognized as necessary to correct the ills that were essentially built into the structure of capitalism. Neoliberalism starting in the 80s with Reagan and Thatcher and Kohl in Germany said: Government is the problem, the market is the solution. We need to get away from that. We need to restore the legitimacy of government action, we’ve got the sustainable development goals of the United Nations on the table. Goals that are very ambitious about improving the world, we’re not going to get there under neoliberal capitalism. We’re not going to get there when we think it’s illegitimate for government to direct action. We’re not going to get there if everything is a public-private partnership and depends on corporate profitability in order to direct that system.

And maybe we have to move beyond capitalism. At the local level there’s lots of alternatives. And  broadly speaking, it’s the subject to a whole another conversation. And I have a paper coming out next year on capitalism in education that talks about alternatives, so maybe we’ll do another podcast there. But the broad answer is, you need to build towards a more participatory democracy and more towards a workplace democracy. The problem with capitalism is that our workplace is authoritarian. We teach democracy in the political sphere; we don’t have a lot of that at a very participatory level either. But we need democracy in the workplace as well.

Will Brehm  45:28
It seems like a lot of what you’re saying is that we have to think beyond the connection of education as being for the development of human capital, and having a different value of education. And there can be many it sounds like, and many different ways of achieving those values or putting those values into action. But it seems like that’s the first step: decoupling, or de-linking the connection between education and human capital development.

Steve Klees  46:05
No, I would agree. And to be fair to human capital theorists, some of them recognize that broad connection. It’s got narrowed in practice so much that all we’re looking at as the connection to education to the workplace. But citizenship can be subsumed in that human capital framework. The problem is its basis in efficiency. You want to talk about the many things education does, the many more things we want it to do. We don’t want to just make education about workplace. We don’t want to make education just about literacy and numeracy. We need education for peacebuilding, for people to not be aggressive, for people to be fair with each other, for people to have resilience, and people to be creative. So there’s lots of purposes of education, and human capital theory and practices just narrow the field too much. And more broadly speaking about this political economy framework versus a more mainstream dominant human capital neoclassical framework, the political economy framework doesn’t offer the technical policy guidance that rates of return give. For neoclassical economists, policies are a dime a dozen. You just do your cost-benefit analysis and this year, vocational education is better than academic education, higher education is better than primary education.

If you reject that framework, what you have is human capital theory, neoclassical economics – our ideologies masking as science. It’s absurd to think that there’s some way to assess technically, the tradeoff between higher education and primary education, between education and health and the environment. All we have is a messy, participatory democratic struggle of individuals and groups with some common interests and some different interests. And for me, what we have to do is find ways to facilitate that struggle and in doing so, economics and the dominant scientific perspective says you need to stay neutral and objective. For me, I’ve learned that you always take sides. That when I write a paper, when I teach a class, when I’m doing research, when I engage in policy, when I engage in my life, I always have to take a side. And if you don’t think you’re taking a side, you are because this is a struggle. This is contestation. And I guess my concluding point is that for me, this is what I said in the paper you cited, neoclassical economics and human capital theory are ideological dead ends. But fortunately for all of us, there are lots of alternatives.

Will Brehm  49:29
Well, Steve, Klees, you gave us a lot to think about in this conversation. Thank you so much for joining FreshEd, and you’re definitely welcome back when that new paper comes out on capitalism in education.

Steve Klees  49:41
Thank you very much for having me.

Want to help translate this show? Please contact info@freshedpodcast.com
Have any useful resources related to this show? Please send them to info@freshedpodcast.com

Today we explore some of the problems with global learning metrics from the perspective of teacher unions. In particular, we look at outcomes-based approaches to international education development.

Such an approach uses global learning metrics to quantify supposed outcomes of education. But as a result, education is reduced and simplified.

My guest today is David Edwards, Deputy General Secretary of Education International in Brussels. Education International is the global federation of teacher unions. He will present some of the ideas discussed today at the CIES Symposium in November. Check out FreshEdpodcast.com for more details about the event.

Today marks the first installment of a seven-part miniseries on Global Learning Metrics. In effort to promote the inaugural Symposium of the Comparative and International Education Society, FreshEd will air interviews with some of the invited speakers.

To kick things off in this episode, I speak with renowned educational economist Eric A. Hanushek about global learning metrics and his use of cross national educational data to understand what is possible in education systems around the globe. He has authored or edited twenty-three books along with over 200 articles.

Dr. Hanushek is perhaps most famous for introducing the idea of measuring teacher quality through the growth in student achievement, which forms the basis for value-added measures for teachers and schools. More recently, his work has focused on the quality of education and its connection to national economic growth.

Eric A. Hanushek is the Paul and Jean Hanna Senior Fellow at the Hoover Institution of Stanford University and will speak at the CIES Symposium this November.

I hope these shows will spark your interest in joining the Symposium. It starts November 10 in Scottsdale, Arizona.

It’s been over two months since the United Kingdom voted to leave the European Union. Right after it happened, I invited Prof. Susan Robertson on FreshEd to talk about the possible consequences the Brexit vote would have on education. During that conversation, I asked if this vote would open the possibility for a new left to emerge within the British Labour Party.

Well, how have things turned out?

To update the situation in the United Kingdom, I recently spoke with Mario Novelli. Mario Novelli is Professor of the Political Economy of Education and Director of the Centre for International Education (CIE) at the University of Sussex. For years, Mario has followed the solidarity work of Jeremy Corbyn, who is now Leader of the labour party and currently in a leadership battle with Owen Smith.

This short episode of FreshEd has been taken from a longer conversation I had with Mario about his research on inequality and education, which will air on September 12.

The Global Partnership for Education is a powerful multi-stakeholder organization in educational development. It funnels millions of dollars to develop education systems in dozens of low-income countries. Yet the board of directors of the organization strategically avoids some of the most important and controversial topics in education today.

My guest today, Francine Menashy, has researched the Global Partnership for Education and the ways in which its board of directors avoids the topic of low-fee private schools, which is a heavily debated idea in both education policy and research.

Francine Menashy is an Assistant Professor in the Department of Leadership in Education at the University of Massachusetts Boston. She researches aid to education and non-state sector engagement, including the policies of international organizations, companies, and philanthropies.

Her research discussed in today’s show was funded through a fellowship with the National Academy of Education and the Spencer Foundation.

Citation:Menashy, Francine, interview with Will Brehm, FreshEd, 33, podcast audio, July 21, 2016. https://www.freshedpodcast.com/francinemenashy/

Transcript, translation, and resources:

Read more


Human capital theory connects education to the labor market. It posits that more education makes workers more productive, which increases earnings. A more educated and productive workforce subsequently increases the gross domestic product of a nation. This theory has been prevalent since the 1950s and continues to play a central role in minds of both policy makers and parents. You go to school because you will get a better job in the future. The government invests in education because it will have a return on investment in larger GDPs.

My guest today says human capital theory is dead.

Hugh Lauder is Professor of Education and Political Economy at the University of Bath. He specialises in the relationship of education to the economy and has for over 10 years worked on national skill strategies and more recently on the global skill strategies of multinational companies.

Citation: Lauder, Hugh, interview with Will Brehm, FreshEd, 29, podcast audio, July 21, 2016. https://www.freshedpodcast.com/hughlauder/

Will Brehm  0:15
Hugh Lauder, welcome to FreshEd.

Hugh Lauder  0:24
Pleased to be here.

Will Brehm  1:08
Human capital theory is such a common place theory in many respects, because when people think about education, they think of it as for human capital development. What is human capital?

Hugh Lauder  1:54
Okay, so I need to take you back a little to the beginning of the theory. The theory was the first sophisticated account of the relationship between education and the economy, and it said that basically people who were better educated would be more productive. And in being more productive, they would then earn a higher income. So that brought education into the picture because what it required was for higher numbers of people to be educated in order they could become more productive so economies could grow, and their income would also accordingly grow. So that’s the basic idea behind it. And it’s an idea, of course, which has permeated through society. It first began really in Chicago in the 1950s at the university there in the Economics department. And then policymakers took it on board and policymakers thought, “Wow, we’ve got a win-win here. Because what’s happening is that if we increase the opportunities for education, so our economies will grow, so people will gain a greater income.” And at the same time, there’s a kind of connection with social justice. So that, for example, as long as people are prepared to work hard and are motivated in terms of education, then they will get their just rewards. And they’ll get their just rewards because employers will always choose the most talented; those that are likely to be most productive. So underlying what seems like an economic theory is actually also a theory of meritocracy. So that’s the economists, that’s the policymakers. And on top of that, of course, now we have parents and students who are going, “Okay, if I want a good job, then I’ve got to get a good education.” So that’s basically the idea behind human capital theory.

Will Brehm  4:01
And it’s led to some interesting notions in education like this rate of return. Can you talk a little bit about what this notion is?

Hugh Lauder  4:11
Yes, sure. So how do economists know, and policymakers know that this claim that education will lead to increased productivity, which will lead to increased income? Well, not how do they know it, but how do they make that assumption? They make that assumption by saying, “Let’s have a look at the rates of return for different kinds of education and skill in the economy.” And in the past – not now, but in the past – what they seem to have found is that the better educated you are, the greater your rate of return in terms of your income.

Will Brehm  4:52
So more schooling means higher income in the future.

Hugh Lauder  4:55
That was the idea, yes.

Will Brehm  4:58
So it’s like it’s predicting the future in many ways; that’s what they’re trying to do.

Hugh Lauder  5:02
For sure it is.  Yes, they really thought that they had a theory which would actually explain and predict the future. And in fact, it has been a theory which has been around, as I said earlier, since the 1950s. And so in terms of social science theories, it’s one of the longest living. But it’s now coming to an end.

Will Brehm  5:27
Before we go into those critiques about the end of human capital theory, can you talk a little bit about what sort of impact it had since the 1950s on education, on education policy, on education development?

Hugh Lauder  5:44
I think the impact has come about in a number of ways. First of all, one of the immediate forms of impact was in development. So the World Bank took on the notion of human capital theory and has argued consistently, since the 80s, that human capital embodied in educated workers would raise the income of countries and of individuals in developing countries. So that was one clear example of the consequence of that particular theory. But, at the same time, it’s also been the case that in developed countries, it’s been seen that if you can increase your higher education system, then you’ll also get a win-win. You get the win-win because people will earn more money as workers and countries will have higher levels of gross domestic product. So these have been the two major consequences of the theory. But it’s also had an extra twist. And that was the notion of the knowledge economy. And the knowledge economy, which sort of started to develop as an idea in the late 80s, also seem to reinforce the idea that we now needed more educated workers. And the more educated workers there would be, so they would become more productive. And this was known as skill bias theory because at the heart of this form of human capital theory was the idea that technology would drive the demand for higher educated workers. So the skill would be biased in favor of the technology and the demand for higher skills.

Will Brehm  7:34
And it would be education that would provide those skills to operate that technology that is driving the economy?

Hugh Lauder  7:44
Precisely that. Yes. Now in more recent times, economists have become a little more sophisticated in one sense and they’ve started to look at particular kinds of skill for which there’s a higher return. But at the same time they’ve been kind of “atomizing” education into particular kinds of skill. So employers have gone in the other direction, and very often look at potential employees holistically. They want to know about their all-around capability in character, rather than also the specific skills.

Will Brehm  8:21
The work that I’ve done in in Cambodia, I’m just amazed by the prevalence of the idea of human capital being the main purpose of education. It is always meant to build and develop human capital because it will increase incomes, and also increase GDPs of the nation. And the conversations that we have are always about this idea of projecting into the future: what sort of economy Cambodia is going to have in 2030, for instance, and what skills are needed? And it just seems like it’s a fool’s errand of trying to predict the skills that are needed in the economy in 2030 for a country like Cambodia that’s rapidly changing; for a global economy that’s rapidly changing.

Hugh Lauder  9:14
Yes, I think this is a very good point. Let me just step back for a moment and say that in developing countries, there are certain sorts of skill that are clearly required for their development. And these forms of skill are to do with the state and state workers. They’re to do with various forms of craft work, so electricians, builders of various sorts, carpenters, that kind of thing. You need those kinds of skill. But the idea that you can predict in 2030 what’s going to happen is more problematic. And it’s more problematic, because just at the time when these developing countries are emerging into the global economy, so many of the techniques which are adopted in the global economy will hit them hard. So, for example, computer algorithms – what Phil Brown, my coauthor, and I  have called “digital Taylorism” – that is moving up the skill chain very quickly, and robots. So, for example, if you look at China right now, there are less people in manufacturing in China now than there were in 2000. So in other words, many of the techniques which have been used in the knowledge economy – and actually it’s not the knowledge economy, its knowledge capitalism, because capitalism is always trying to reduce the cost of labor, including skilled labor – many of those techniques that have been developed in the developed countries are now being applied to developing countries. So that makes it kind of problematic as well.

Will Brehm  10:57
Let’s shift to your specific critiques of human capital theory. What do you find so problematic about the theory itself, maybe not the method that’s employed by the theory?

Hugh Lauder  11:11
Okay, let’s have a look at the theory itself. We start with education, and that’s meant to lead to greater productivity, which is then meant to lead to greater income for the individual and the GDP of the country. Well, when we look at that set of connections, we find that they are all problematic. They’re all problematic for this reason: That first of all, education. There’s now considerable split amongst economists as to what we mean by education. Is it something as I suggested earlier, which is a form of all-round development of an individual? Or is it about particular skills? And that debate has really not taken off yet, but it will. So the education itself in terms of human capital – ‘what is the capital’ is a problem. Then when you look at productivity, what we see overall is that there are more and more educated people in the world, more and more educated people in particular countries like the UK or USA, and yet productivity is either flatlining or is very uneven. So the link between education and productivity is now become wholly problematic. Then when you look at the relationship between productivity and income, it becomes even more problematic because what you see is that instead of workers getting rewarded for their productivity, since around 1978 to 1980 in the United States and the United Kingdom, what you see is that increasingly, the wealthy are creaming off the productivity of other workers. So there are problems with all these different accounts of the relationship between education, productivity and income.

Will Brehm  13:10
So this would be the Piketty’s argument of the rise of the 1%.

Hugh Lauder  13:16
The rise of the 1% certainly has been, in part, because they’ve creamed off the productivity of other workers, but we need to look more closely at the relationship between productivity and income than what Piketty was talking about. Because as far as I can see, and read him, he does assume that most of the rest of the income that people get is a reflection of their productivity. In other words, he becomes quite orthodox once he’s had to look at the 1% in terms of his account of wage determination, and I don’t think that’s right. You only have to look at feminist critiques of human capital theory – and I’m thinking in particular now of the work of Antonia Kupfer in Dresden – and you see that there are a whole range of jobs for which it’s very difficult to determine productivity. It’s not only super managers, as Piketty would say, but it’s care workers. How do we measure their productivity? Why is it that women who can be very skilled at care work get such low wages? There’s a whole range of different questions that can be asked about this relationship between productivity and income. And the idea that productivity simply determines income is taken as a truism in orthodox economics. But I don’t think we can take it as such anymore.

Will Brehm  14:48
So let’s turn to the way in which human capital theory has been studied empirically. What sort of critiques do you see in the way in which it’s been studied?

Hugh Lauder  14:59
Well the way it’s been studied empirically – I’ll give you a clear example since you raised the idea of 30 years as a future timeline for prediction. There’s work by two leading economists, Eric Hanushek and Ludger Woessmann, and Hanushek and Woessmann have published a series of papers for the World Bank, the OECD, where they look at the quality of PISA data (this is international test data for different countries). And on that basis, they then predict that in the future if countries can raise their education standards and their educational achievement so this will increase GDP in 20 or 30 years by X amount. And that has become kind of a standard way of analyzing the returns to education in terms of human capital theory. But I don’t need to tell you this, you will know it and so will all your listeners, that that kind of assumption simply doesn’t take into account the real world. We know, for example – and this is often an example I use – that when you compare Korea in the 1950s and Ireland in the 1950s, what you see as two countries with large numbers of relatively unemployed graduates. Both countries then began to take off, but if you look at the path of Korea where much of the takeoff was state led, and is still highly state influenced, what you see is a totally different kind of success story to the story of Ireland, which of course collapsed in 2008. So different trajectories for countries based on different ways of developing them produce different results. So what Hanushek and Woessmann don’t really do is take into account strategy, institutions, all the things that actually make a difference to whether countries and individuals in them do well or not.

Will Brehm  17:16
One of the critiques you you put forward is that human capital theory or the scholars who are using human capital theory often employ methodological individualism. And we hear this quite a bit also in other education research, and I just would like to ask, what does that actually mean?

Hugh Lauder  17:37
Sure. Basically, the assumption of methodological individualism (which is an ugly term, I know), the basic assumption is that the only thing that exists in society are individuals, and therefore it is to the individuals that we look to explain educational outcomes, to explain income, to explain the key features of social life and economic life.

Will Brehm  18:06
And so it neglects things like history, and perhaps the privilege that one could get from his or her parents rather than just their individual unique ability to learn.

Hugh Lauder  18:22
Yes, absolutely, that’s correct. So it neglects history. It neglects the structures which govern our societies such as class, patriarchy, racism. They don’t enter the story at all. And at the same time it neglects institutions, specifically institutions of education, for example; institutions that steer an economy. All that is simply discounted in this kind of explanation, which focuses on individuals.

Will Brehm  18:58
So if we were to talk about alternatives to human capital theory, how would you describe the link between education, productivity and income?

Hugh Lauder  19:08
Okay, well, first of all, these are now very, very complex connections. They’re not at all simple in the way that the original theory assumed. So we need to think about this very, very differently indeed. Let me just come back to the issue of structures and institutions. When you look at, for example, skill bias theory, it says that we understand that in the 20th century, technology was skill biased – that actually what happened was that as technology developed, so the demand for skills increased. But when you look at the history, it can be read completely differently. And it can be read like this, it can be read: Well, actually, the basis of 20th century industry was Fordism, the idea that people could put a nut on a bolt on a production line and out would roll many cars, many televisions. All the consumer goods that we now take for granted. These people were not up-skilled, they were de-skilled, because originally the people that made the cars were craftspeople. So that’s where you have what they call “skill replacing”, where the technology replaces the skill, doesn’t enhance or demand an increased skill. So then you say, “Well, where did the skill bias, the skill enhancement and demand for it, come from?” And actually, it came from the large numbers of white-collar workers you needed to run a large corporation like it. So these are the people that did the marketing, these are the people that did the accounts, these are the people that did all the other finance work and the planning.

But in order to understand how those corporations grew, you also then have to go to a much wider political economy. You have to go to a political economy which talks about the structures of the labor market – and here we’re looking at trade unions as well as employers. And back in the 50s, for example, and the 60s, trade unions were very strong, and they could increase their wages so that their workers could then buy the cars that were rolling off these production lines. Now, you’ll see for a moment there that the story I’m telling is a very much more complicated story than the one that skill bias theorists assume. Now, they assume that because in the past, we have had skill bias theories, so we will in the future. But the political economy around skill and skill development has now changed dramatically, and we need to understand it in terms of globalization, not in terms of Keynesianism and the idea that you could get some kind of agreement between trade unions, employers and the state, because now trade unions are much weaker, for example. They’ve been weakened through neoliberalism.

So you need to tell a completely different story. And you tell a story now about globalization and the demand for skilled workers can occur anywhere; it doesn’t have to be in any particular country. Multinational companies can simply say, “Okay, these skilled workers we want, they’re cheaper in Shanghai than they are in London. We’ll shift the demand to Shanghai.”  So you can see that we’re living in a very, very different kind of world in which the sorts of prediction that human capital theorists made, or assumed they could make, simply no longer exist in that particular way. So we need a different kind of theory. But – and here’s the big but – the world we’re about to enter is going to be even more radically different from the one I’ve just described.

Will Brehm  23:16
How so?

Hugh Lauder  23:16
Well, robots.  People make a lot of robots. And I used to be very skeptical about this. But I’ve just been talking to very senior infocom officials in multinational companies, and they tell me they’re scared of the consequences. And if they’re telling me that, then I’m really beginning to sit up and look at the other studies which suggests that robots can take many of the jobs that skilled workers used to take. We are moving, I think, into an era in which jobs and income will become increasingly uncertain for many, including many graduates. And that requires us to rethink the entire relationship between education and the labor market, because the labor market is so radically changing.

Will Brehm  24:12
Right. It’s fragmented and global, and you see further changes in the future.

Hugh Lauder  24:17
Absolutely. And they’re going to cause policymakers huge problems, which I think they’re reluctant to really start thinking about and confronting.

Will Brehm  24:28
Before we we turn to, “What then of education?”, you use this term, “the global auction for jobs”. Can you talk a little bit about that?

Hugh Lauder  24:38
Yes. So this is a book that Phil Brown, and I wrote with David Ashton back in 2011, which has kind of taken off a bit, and it’s taken off because up until then, the assumption was that technology would always lead to an increase in demand for skilled workers, and particularly for graduate workers. The research we did was on the skill strategies of multinational companies. And they told us a very, very different story. And I’ll give you an example of that, and it goes like this: The first interview that we did was with a human resources, very senior executive for a German engineering company in Germany. So the interview was in Germany; multinational company, though. And I said to him, because I have still had the human capital thinking cap on, as it were. I said to him, “Do you have a shortage of engineers?” And he said, “No”. And I said, “Do you get them from Germany?” And he goes, “No”. So I said, “Do you get them from England?” And he goes, “No”. And I said, “Do you get them from America?” And he said, “No”. Now you can see how my mindset was. I was thinking, Germany, Britain, America, right?

Will Brehm  25:58
Right, the place where engineers you thought were being produced.

Hugh Lauder  26:01
Exactly. And I said, frustrated, “Okay, where do you get them from?”  He says, “We get them from China, we get them from India, we get them from Russia, especially if they’re computer engineers and mathematicians. And we get them from Bulgaria, because in the Soviet bloc, this was designated as the leading place for computer analysis and development.” And in that moment, our eyes opened to a whole new world that this guy in two sentences had given us. And that meant that we had to then get on airplanes, and go and interview executives of multinational companies from around the world to see what was going on. And two things were going on: First of all, because they are in such an intense competition, they’re always seeking to drive down costs, and brainpower they want to make as cheap as possible. So “cut price brainpower” we call it. Now you get that because you can get engineers, for example, in China and in India, for a fraction of the price you can get them in the West. So what you see then is the offshoring of jobs, or the movement of jobs, from particular countries like the United States and the United Kingdom to East Asia.

But at the same time, we picked up something else that was going on. And that was this notion of digital Taylorism, the idea that you can take skilled work that graduates used to do and you can break it down into discrete tasks, standardize it, routinize it and then put it into algorithms that you can ship across the world so that work can be done anywhere. So these are the two key features of the global auction. Now, there is one exception to this and that is at the same time as we’re producing all these graduates, highly skilled workers from around the world so then in comes a particular ideology which suggests that it’s only the very few of those graduates who are really talented. And so now on every bookshelf of every HR executive office that we went to, was this War for Talent book. And this is about how you recruit the most talented in competition with your other corporates. So this is the one exception: there are a few people who are now designated as talented. Now, there’s a major debate as to what’s really going on there and whether these people really are talented, or whether it’s just executives or corporations wanting to see a kind of great reflection of themselves in the younger new recruits coming into their company. Because, of course, these people designated as talented earn much more money than everyone else. So that’s the global auction in a nutshell. And that began to open up two debates related. The first was, “No, we don’t live in a knowledge economy. No, if you’re a graduate, you’re not going to enter a world where you’ll be highly rewarded necessarily, where you’ll have status, creativity and autonomy. Quite the opposite might happen, that you’ll be entering routinized work.” And alongside that, and following from that, is the idea that actually knowledge work itself is now being stratified. So that you’ll get an elite which is the talented, you might get another group beneath them that do their bidding, and then you’ll get these routinized workers. So that was why the book cause something of a stir, because we were arguing, for the first time I think, that the idea of the knowledge economy and of human capital and skill bias theory really didn’t work in the way that had been assumed.

Will Brehm  30:01
So what then of education? How do we make sense of education in this this world that you are painting for us here?

Hugh Lauder  30:09
Okay, this is, I think, a really important question. Because if you were to just think that we’re talking about today and tomorrow, then there could be a critique which comes in, especially from the right wing, which is: “Oh well, we’re just educating too many people to too high a level.”  And in itself, that is problematic, because what else are graduates going to do when in countries like the United States and Britain, we no longer have the forms of industrialization where people could do high skilled, high paid work that, for example, still obtains in parts of Germany. So that’s one problem, but there’s a much bigger problem on the horizon. And I kind of signaled it when I talked about the robots. Because if so many of the skilled jobs that we have are going to be done by robots, then what’s going to happen to graduates? What’s going to happen to those who are educated? And I think the answer to that is something like this: We are going to have to give people a basic wage, a universal basic wage. Because the insecurities in the labor market will be so great that many will simply not survive unless they get a universal basic wage. Now, that universal basic wage will enable people to do a number of different things. It will enable them to retrain, to re-skill, for which they will need learning accounts so that they can draw on an account to upscale where they see a need. It will enable them to innovate and to develop different ways of interacting with this world. And the universal basic income will expand the labor market from beyond the confines of a market to work which is seen as important and contributing to society. And of course, care workers would be a clear example of that. So, that’s the labor market part of it in a nutshell. Then what about education? Well, if we’re thinking about that world, and you reflect on that for a moment, the uncertainties of that world, then clearly we need people to be as best educated as we possibly can make them. We need people who are reflective, alert, resilient in order to be able to make the best of the opportunities they have. So education becomes more important in these terms than in the past.

Will Brehm  32:41
Well Hugh Lauder, thank you very much for joining FreshEd.

Hugh Lauder  32:44
It was a delight. I hope it was of some value to you.

Want to help translate this show? Please contact
Have any useful resources related to this show? Please send them to info@freshedpodcast.com