Skip to main content
Search form
  • 00:01


  • 00:17

    IAN SHAPIRO: What's the differencebetween problem-driven theorizing,method-driven theorizing, and theory-driven theorizing?Political scientists are constrained to theorize,so we're all, at some level, interested in theory.The question is where you start.Method-driven theorizing and theory-driven theorizing

  • 00:40

    IAN SHAPIRO [continued]: are differentiated from problem-driven social science,but they're also different from one another.I think that theory-driven peoplestart with a particular theory, and then lookfor ways to vindicate it.The joke that comes to mind is the fellow thatgoes up to a farmer in Donegal and says,

  • 01:02

    IAN SHAPIRO [continued]: how do I get to Dublin?And he says, well, I wouldn't start out from here, fella.And the notion is that the theoryis developed without any reference to the problemthat it might illuminate.And I wrote a book about that with Donald Green, Pathologiesof Rational Choice, where we argued at some length

  • 01:24

    IAN SHAPIRO [continued]: against theory-driven research and in favorof problem-driven research.Method-drivenness is a different kettle of fish,and it more calls to mind the joke about the drunk whocomes out of a bar and has lost his keys,and he's crawling around under a lamppost looking.

  • 01:45

    IAN SHAPIRO [continued]: And somebody comes up to him and says,is this where you lost your keys?And he says, no.I lost them over there behind the shed.And he says, well, why are you looking here?And he says, because this is where the light is.So method-driven work is different from theory-drivenin that it's not trying to vindicate a particular model

  • 02:08

    IAN SHAPIRO [continued]: or theory or outlook.But rather it says, we'll only studythings that are susceptible to a certain method,whether it's a quantitative method, an experimental method,or any of the method.Problem-driven work, of which I am an advocate,is where you start with a problem in the world.

  • 02:29

    IAN SHAPIRO [continued]: You then look at previous attempts to grapple with it,and you think about how you can improve upon them.And everything always comes back to the problemthat you began with.And so I have argued against both theory-driven andmethod-driven work in favor of a problem-driven approach.

  • 02:56

    IAN SHAPIRO [continued]: So how is the debate among method-driven, theory-drivenand problem-driven researches playing out,and what is my view about that?Well, when I came along in the 1970s and '80s,that was the heyday of rational choice theory.And that was very much a theory-driven enterprise.

  • 03:18

    IAN SHAPIRO [continued]: And the book I wrote with Don Green, Pathologies of RationalChoice, was really a critique of the theory-drivenness of that.And what we did was we looked at the main literaturesin that field and argued that very little valueadded had actually come out of that research.

  • 03:40

    IAN SHAPIRO [continued]: Despite all the multiplication of modelsand this theoretical sophistication,we didn't know appreciably more about politics as a result.That played itself out.We published our book in 1994, and there was a decadeof fairly heated debate.

  • 04:02

    IAN SHAPIRO [continued]: And the discipline went in a much more empirical directionthereafter.Certainly it was the case in the 1980sthat you could get an article published in the AmericanPolitical Science Review that had nothing but a model.I'd say that by 1998, that was not possible.You really had to have empirical work as well.

  • 04:24

    IAN SHAPIRO [continued]: How much our book had an impact on that, who knows?But we were certainly part of that conversation.Where the discipline has gone sinceis in a heavily behavioral direction and much morein a method-driven direction.This is partly the influence of the rise

  • 04:44

    IAN SHAPIRO [continued]: of behavioral economics in economics and, of course,in psychology where Kahneman's work has been so influential.And the shift away from neoclassical modelsin economics.And what has gone along with the behaviorism

  • 05:05

    IAN SHAPIRO [continued]: has an acute interest in questions of method.And particularly in the last decade or sothere's been the rise of field experiments, fieldexperimental work, in order to try and do experimentsbut solve the so-called problem of external validity.

  • 05:27

    IAN SHAPIRO [continued]: Psychologists have done experimentsfor decades on college kids in North American universities,but there's been a long going running critiqueof that which is that they're not very representativeof general populations.And so the idea of field experimentsis that you can have randomized treatment and control groups,

  • 05:50

    IAN SHAPIRO [continued]: and not rely on college kids.And therefore, you can solve questions of external validity.In there's been, particularly in the field of American politics,but increasingly in areas of development and study of thingslike corruption in third world countries,a lot of use of field experimental techniques.

  • 06:16

    IAN SHAPIRO [continued]: And actually, my-- interestingly,my co-author from the Pathologies book, Don Green,who's now at Columbia, has been a big pioneer of that.I've been something of a critic of that turnbecause I am of the view that one should be problem-driven.And in my judgment, the cost of field experiments,

  • 06:43

    IAN SHAPIRO [continued]: there's been a tendency to study often very small questionsor questions that might only be of interest, for instance,to people running campaigns.So there's been a lot of field experimental work on turnout,for example, on the question of whether knocking on doors

  • 07:03

    IAN SHAPIRO [continued]: has a better return per dollar spent than sending flyers,let's say.Hugely important if you're running a campaign.It's not clear that it's important to anybody else.It's not clear that it's part of any largertheoretical conversation in political science, and so on.Likewise, if you look in the developing world,

  • 07:29

    IAN SHAPIRO [continued]: a lot of the field experimental workis basically driven by what they can get access to,what data they can get access to from governments, and so on.And so often, it's not motivated by anything otherthan the availability of data on which it's possible to runfield experiments.

  • 07:50

    IAN SHAPIRO [continued]: And so there's been something of a critique of the fieldexperimental work-- and I've beenpart of that-- which says, we're learning,in some of these instances, more and more about less and less,because they're studying things thatare susceptible to the technique,but not obviously things that anyone has other than, say,

  • 08:13

    IAN SHAPIRO [continued]: a consultant or somebody running a campaignor somebody implementing a particular policy-- theydo a lot of program evaluation as well-- has an interest in.And so I was earlier out of sorts,if you like, with the theory-driven.Now I'm out of sorts with the method-driven people.

  • 08:36

    IAN SHAPIRO [continued]: So that's sort of where my research fits into that.Well, I would say that's more critiqueof the method-driven work, that ittends to be stuck in the weeds.The old theory-driven, rational choice scholarship was oftenastonishingly ambitious.They had very grand theories, and often

  • 08:59

    IAN SHAPIRO [continued]: were criticized for that as people said-- accused thema kind of [INAUDIBLE] big think thatwas uninformed by what was actually going on in the world.But I think that the questions they studiedtended to be artifacts of the literaturerather than artifacts of problems in the world.

  • 09:19

    IAN SHAPIRO [continued]: So you would find, for example, a researcherwould discover some people cooperatingwhere rational choice models would tell youthat they-- there's a collective action problem,and they shouldn't be cooperating.And so the research question would

  • 09:40

    IAN SHAPIRO [continued]: become, how did they solve the collective action problem?Well, it may never have occurred to themnot to cooperate because of the way in which they were raised,or whatever it might be.So that tended to be-- the debates in the disciplinetended to be artifacts of the theories that

  • 10:03

    IAN SHAPIRO [continued]: were being trafficked in, not drivenby anything in the real world.But it wasn't necessarily an in the weeds thing.What we have now is very much in the weeds, in that oftenhuge amounts of technical sophisticationand methodological sophisticationare cranked up to study very small questions.

  • 10:25

    IAN SHAPIRO [continued]: So over the last several decades,political science has gone throughhuge methodological changes.When I came along in the 1970s, it was a heavily theory-drivenera.And this meant that people were developingvery ambitious models for studying politics, and usually

  • 10:46

    IAN SHAPIRO [continued]: using the tools of microeconomics to do that.And so it went under the heading of rational choice theorybecause they took rational choice models from economistsand then applied them to politics.As I say, they were often extremely ambitious,and often were developed by people who didn't actually

  • 11:08

    IAN SHAPIRO [continued]: know that much about the nitty gritty of politics.That produced eventually a reactionof which my book, Pathologies of Rational Choice that Iwrote with Don Green, who's now at Columbia, was a part.And since that time, there's been a sea changein political science in the direction of what

  • 11:29

    IAN SHAPIRO [continued]: I call method-driven work, which is verydifferent than the theory-driven work in that it'smostly concerned with behavioral changes in human conductin politics, and not particularly done by people

  • 11:49

    IAN SHAPIRO [continued]: who are interested in theory at all,and sometimes get accused of beingatheoretical or anti-theoretical.Method-driven folks are looking for decisive contributionsto the understanding of political behaviorwhere the big emphasis on what counts is decisive.And so they are looking for ways to study

  • 12:14

    IAN SHAPIRO [continued]: things that allow, for example, the useof experimental controls.They engage very often in field experiments,because what they want to do is isolate the causal factorsat work in behavioral change.And if you can have a control group and a treatment group,

  • 12:37

    IAN SHAPIRO [continued]: you can aspire to do that.They're often concerned, though, with quite small questions,such as whether knocking on doors increasesturnout more than sending flyers, for example,or whether a particular way of implementing a policyis better than a different way of implementing that policy.

  • 12:58

    IAN SHAPIRO [continued]: Often because that's what you can get data about.And that's become the method-driven fashionof the last decade or so.The criticism of problem-driven researchis that it doesn't tell researchers what to study,

  • 13:20

    IAN SHAPIRO [continued]: and this is true.We can go all the way back to Max Weber's essay, "Scienceas a Vocation."And one of the principal claims he makes inthat essay is that science tells you how to study things.It does not tell you what to study.And choice of what to study has Weber said in that essay,is actually an ethical choice.

  • 13:41

    IAN SHAPIRO [continued]: It's an ethical question.It's a normative choice.And if you think science will tell youwhat to study scientifically, you're just muddled.That was basically Weber's position in that essay,and I agree with it.So what happens is that in theory-driven work,

  • 14:03

    IAN SHAPIRO [continued]: the questions are all generated by the theory.In method-driven work, the questionsare all generated by what kind of data is available.Problem-driven work, you've got to, I think,make the case for the importance of the problem.Because if you don't convince peopleof the importance of the problem,they're not going to be interested in what

  • 14:24

    IAN SHAPIRO [continued]: you have to say about it.So my own view is that you want to makethe case for the importance of the problemoutside the walls of academia.And if your sense of why the problem is worth studying is

  • 14:46

    IAN SHAPIRO [continued]: connected to academic literatures--which often it isn't-- it should be-- it's up to you,the researcher, to join the dots and explain why these modelsor methods matter from the standpoint of whythe problem's actually important to people out therein the world.If you don't do that, you'll lose them.

  • 15:06

    IAN SHAPIRO [continued]: You'll lose everybody except the other researchersin your subdiscipline.So I think the burden is actuallyon the researcher to make the case for the importanceof the research.And that means explaining why the problem is worthstudying to begin with.So if we go back to turnout, why should

  • 15:28

    IAN SHAPIRO [continued]: anyone other than people running a campaignhave an interest in knowing whether knocking on doorsmakes a bigger difference than sending flyers?Well, you might have some-- it mightbe part of some larger research agenda in psychologyas to face to face relationships being different

  • 15:49

    IAN SHAPIRO [continued]: than other kinds of relationships.But then the burden is on you to explain that.But it is for sure a normative choice that scientists make,and it's not something that science can itselftell you what to do.

  • 16:14

    IAN SHAPIRO [continued]: So why is problem-driven researchsuperior to method- or theory-driven research?The answer is that it's more likely to illuminatereal questions.If you are so wedded to a particular theoretical outlookthat everything you do is geared to vindicating that outlook,

  • 16:35

    IAN SHAPIRO [continued]: you get the problem that if your only tool is a hammer,everything starts to look like a nail.And you, as I said, you discover solutionsto collective action problems that perhaps never actuallyexisted in the first place.Or you will create a whole research agendaon questions that are of no interest to anyone

  • 16:57

    IAN SHAPIRO [continued]: other than the thimble full of scholarswho are working on that agenda.And you'll start to run into the sorts of questionspolitical science has run into, namelysenators in Washington asking why NSF money and SSOCS moneyshould be available for political science.

  • 17:19

    IAN SHAPIRO [continued]: Because at the end of the day, we'respending society's resources studying these questions.And if they're not questions thatare of any importance to anyone other than the scholars who arestudying them, it's a problem.So that's why I would say it's important to make

  • 17:40

    IAN SHAPIRO [continued]: the case for the value of working on these problemsrather than on some other problems.So how has this problem-driven approachaffected my own research?Well, in a number of ways.

  • 18:00

    IAN SHAPIRO [continued]: I'll give you one.One example is I was working on a book on democracyand distribution, which was principally concernedwith the puzzle that the median voter theorem predictsthat democracies would just-- would redistribute downwards.We have a very unequal distribution

  • 18:22

    IAN SHAPIRO [continued]: of income and wealth in democracies.Once you have universal franchise,the median voter earns less than the average voter.And therefore, there are many theoremsthat predict downward redistribution of wealthin democracies.It doesn't happen.

  • 18:43

    IAN SHAPIRO [continued]: It certainly doesn't happen in any systematic way.There's all kinds of redistribution in democracies,but not systematically downward.And people have puzzled over that for a long time.And I was working on a book about that in the late 1990s.And as I-- while I was working on that book,

  • 19:06

    IAN SHAPIRO [continued]: in 2001, the estate tax was repealed.This is the most progressive tax in the code.It's paid by the top 2%.More than half of it is paid by the top half of 1%.That's people worth more than $20 million.Yet it was repealed as part of George W. Bush's 2001 tax cut

  • 19:33

    IAN SHAPIRO [continued]: with significant bipartisan support.And I couldn't understand how this happened.And I reached for the standard political science theories,and they were not illuminating.And there were people, some people writingabout this saying, well, politicianswere doing this to extract rents from donors, and none of it

  • 19:56

    IAN SHAPIRO [continued]: seemed very plausible to me.And I ended up-- I got a grant from Carnegie,and I spent a year and a half during a kindof-- not something that political theorists typicallydo.I did a soak and poke study of the repeal of the estatetax, which eventually became a book that Iwrote with Michael Graetz called Death by a Thousand Cuts.

  • 20:20

    IAN SHAPIRO [continued]: And we interviewed about 150 peopleover a course of a year and a half.Congressmen, senators, staffers, lobbyists,every interest group that had anythingto do with the 2001 tax repeal.And we found out as this kind of political who done it almost,we found out how this was done.

  • 20:42

    IAN SHAPIRO [continued]: And it turned out to be very illuminating for theoriesof democracy and distribution.It'll take me a minute to explain why,but let me take a minute.If you go back to the standard stories

  • 21:03

    IAN SHAPIRO [continued]: in the political science literatureabout why democracies don't redistribute downwardswhen the median voter is earning less than the average income,the typical answer the tended to be,well, there are other the dimensionsthan left-right income that matter to voters.So if you add, for example, race or religion

  • 21:27

    IAN SHAPIRO [continued]: into the mix as a dimension about which people care,then the median voter might be somewhere elseon the income question.And so the standard way of explaining why democraciesdon't redistribute down was essentiallyto add another dimension.

  • 21:48

    IAN SHAPIRO [continued]: What was interesting about the estate tax right offthe bat is there was no other dimensions.This was just about money, clearly.So if large numbers of people are supportingthe repeal of a tax that they will neverbe in any danger of paying, you can't say,well, it's because they care about race or something.

  • 22:10

    IAN SHAPIRO [continued]: So right there, there is a problem.And so what we discovered out of our almost ethnographic studyof the repeal of the estate tax wasthat a different analytical lens then complicatingthe median voter theorem gave you a lot more leverageon the problem.

  • 22:31

    IAN SHAPIRO [continued]: And that's the majority rule divide a dollar game.The majority ruled divide a dollar game saysthat if three people are dividing upa dollar by majority rule, there's no stable equilibrium.Because if A and B split it 50-50,C will come along and offer a 60/40 split to one of them.And then the person who's left out

  • 22:53

    IAN SHAPIRO [continued]: will come and offer some other split.And there's no equilibrium.Every possible distribution is vulnerable to being overthrownby some potential majority.What does that tell you?It tells you that distributive politicsis about building coalitions, busting coalitions, holdingcoalitions together that other people try to bust.

  • 23:15

    IAN SHAPIRO [continued]: And what you really need to understandis how distributive coalitions are created,how they're torn apart, and how peopleresist tearing them apart.What we had done in our study was essentiallywe had studied the creative buildingof a distributive coalition of very unlikely bedfellows.

  • 23:37

    IAN SHAPIRO [continued]: I'm not going to go through the book now.So what came out of that was a realization that sometimesthe median voter theorem helps you understand distributivepolitics, but sometimes it doesn't.And the sort of next steps forward concern issues-- well,

  • 23:58

    IAN SHAPIRO [continued]: what are the situations under which the median votertheory gets it right?And what are the situations under whichthe divide a dollar game is more likely to get it right?And those themselves turn out to be empirical questions thathave to do with things like the intensity of mass preferencesand all sorts of other things that it wouldn't occur to you

  • 24:19

    IAN SHAPIRO [continued]: to study if you hadn't really got downinto the weeds of this problem and puzzled your waythrough it.And so that empirical study actuallycaused me to throw away a book manuscriptthat I'd decided was completely wrong

  • 24:40

    IAN SHAPIRO [continued]: and actually to rethink my scholarshipon democracy, my theoretical work, in very fundamental ways.So what does it mean when you start with a problem?

  • 25:01

    IAN SHAPIRO [continued]: As I said a minute ago, in the case of the estate taxrepeal, the problem-- the question was,how could this have happened?How could a coalition have been builtto repeal this tax which seemed so out of kilter with people'sexpectations about what goes on in politics?And that was an answer that we came up with.

  • 25:25

    IAN SHAPIRO [continued]: If you had tried to answer that question without beingproblem-driven about it, you would tend to,as I said, to reach for things thatactually had nothing to do with why the estate tax repealed.Some people said this was about politicians getting rents

  • 25:46

    IAN SHAPIRO [continued]: from contributors.It absolutely was not.By the time we had worked our way through the role of moneyin all of this, that didn't get into the top 10 reasonsmoney had anything to do with this.Or people would have reached for some other dimension,as I said.But that doesn't help you when you're just

  • 26:08

    IAN SHAPIRO [continued]: talking about tax policy.Another thing that I would say about itis you get different research agendas if you dig deeplyinto some empirical question.I'm a firm believer in theory and theorizing,

  • 26:30

    IAN SHAPIRO [continued]: so as I indicated before, I'm veryskeptical of the value of this returnto what strikes me as mindless behaviorism of allthe method-driven people.But I think for any theorist to do the kind of study wedid on the estate tax, it's extremely humbling

  • 26:51

    IAN SHAPIRO [continued]: when you realize the levels of complexityof what goes on in politics.And that puts new questions on the research agenda,and that's important.For example, one thing that was put on the research agendaas a result of that, at least to my mind,is the intensity of mass preferences

  • 27:14

    IAN SHAPIRO [continued]: is really important.So one thing that went on in the repeal of the estate taxwas that it turns out that most voters don'thave intense preferences about taxes.Activists have intense preferences, but voters don't.And they particularly don't have intense preferencesabout the estate tax.

  • 27:35

    IAN SHAPIRO [continued]: So if you run a poll which is [INAUDIBLE] approval voting,should we get rid of the estate tax,you can get 70% to say yes.But if you pair it with getting rid of some popular benefitlike Medicare Part D or something like that,the support evaporates.Or if you put it in a list of possible tax cuts,

  • 27:56

    IAN SHAPIRO [continued]: it drops to the bottom.So it's not as crazy as it might first appear.And one of the reasons that we wereable to repeal the estate tax is that mass preferences are notintense.So the activists could go on to the Capitol Hilland say, look, Senator.This is a free vote.You're not going to lose anything by voting for this.

  • 28:20

    IAN SHAPIRO [continued]: Whereas if you look at something like social security,after the Bush-- George W. Bush administration had beenso successful with tax cuts and Bush was reelected in 2004,some say elected for the first time in 2004,

  • 28:41

    IAN SHAPIRO [continued]: he decided-- he said, I have a mandate,and I'm going to use it.And the first thing he did was tookto put privatizing social security on the agenda, whichhe did.And they put the president on the road for 60 days.They used all the same techniquesthey had used to repeal the estate tax,and they ran into a brick wall.

  • 29:02

    IAN SHAPIRO [continued]: Why?Because mass preferences about social securityare intense, and most-- the median voterdoes not want the government messingwith their social security.And so that then tells you one of-- if youwant to understand distributive politics,one of the things you need to get a grip on

  • 29:22

    IAN SHAPIRO [continued]: is, what drives the intensity of mass preferences?And is it alterable?What alters it?So whole new research agendas open upall out of this problem-driven-- allstarting from this problem-driven question.Why were they able to repeal the estate tax

  • 29:43

    IAN SHAPIRO [continued]: and not able to privatize social security?And that then generates theoretical research agendas.And I think that that's the way that political science canbe most fruitful.

  • 30:03

    IAN SHAPIRO [continued]: I've been engaged in some discussions in the philosophyof science as to whether the ideas of Thomas Kuhn and othersbear on the discussions of methodologiesin political science.Political scientists are congenitally about methodology.The standard joke is, how many political scientists does it

  • 30:25

    IAN SHAPIRO [continued]: take to change a lightbulb?21.One to hold the ladder, one to change the bulb,and 19 to argue about the methodology.Well, what Kuhn says is that pre-paradigmatic disciplinesare engaged in endemic debates about method.And I think what he had in mind is important because basically,

  • 30:48

    IAN SHAPIRO [continued]: many defenders of rational choice theoryappeal to Kuhn's idea of a paradigm,to argue that rational choice was the presumptive paradigmfor understanding politics.And what I said about Kuhn was, well,if you were to be a genuine Kuhnian,you'd have to recognize that there'sno presumptive paradigm for understanding politics, which

  • 31:10

    IAN SHAPIRO [continued]: is incidentally the view Kuhn himself took of the matterwhen he said that the social sciences are pre-paradigmatic.I think putting it in words of one syllable,it means there's no agreement on what to studyor how to study it.

  • 31:32

    IAN SHAPIRO [continued]: Well, I think the great theory-driven enterprisein the social sciences since the Second WorldWar has been microeconomics.So to the-- you know, we talked about rational choice theoryin political science.There have been comparable moves in sociology.And the biggest challenge to that paradigm,

  • 31:55

    IAN SHAPIRO [continued]: to the extent it is a paradigm in economics,has come from psychology, not from political science.I think it's basically Kahneman and Tversky's prospect theoryhas taken on neoclassical economic models at their coreand articulated a different view of psychology.

  • 32:18

    IAN SHAPIRO [continued]: And that has then given rise to the resurgenceof behavioral economics, behavioral political science,behavioral psychology, sociology.So that's been the biggest single change.I think that Kuhn, were he still around,would say that all the social sciences are still

  • 32:39

    IAN SHAPIRO [continued]: pre-paradigmatic.Economists certainly present themselves to undergraduatesand to the rest of the world as having a reigning paradigm.But if you read the economics journalsand look at the frontiers of the field,I think there's as much foundational disagreement

  • 32:59

    IAN SHAPIRO [continued]: there as there is in any other social science.Rational choice models I think have a very mixed historyin their influence in political science.Some of their influence has been for the good.

  • 33:21

    IAN SHAPIRO [continued]: I think that for political scientiststo learn to think like economistsis generally good for them.And I think that some of the issues raisedby rational choice are directly applicable to politics,particularly when you think about the kinds of politicsthat involves self-consciously strategic action.

  • 33:44

    IAN SHAPIRO [continued]: So rational choice models, for example,apply much more helpfully to parties and politiciansthan to voters, just as in economics,they apply much more helpfully to the behavior of firmsthan to the behavior of consumers.So when you're dealing with self-consciously strategic

  • 34:05

    IAN SHAPIRO [continued]: actors who have well defined goals,and they're dealing with a other strategic actors,then these models can be extraordinarily helpful.So I've done a lot of work in recent years,for example, on democratic transitions, and particularlynegotiated transitions.

  • 34:27

    IAN SHAPIRO [continued]: I've done a number of papers on transitionin South Africa and the failed transitions in the Middle East.And I have found enormously helpfulthe literature that's grown out of the prisoner's dilemma.Now this is a rational choice model that goes back to Hobbes.

  • 34:49

    IAN SHAPIRO [continued]: And I think in situations like thiswhere you have political leaders whodistrust one another for good reasonand negotiate a settlement that requires them to takeenormous risks, understanding how they escape what clearly

  • 35:13

    IAN SHAPIRO [continued]: was a prisoner's dilemma is very interesting and helpful.The analytical tool provided by the prisoner's dilemmais helpful.But it's precisely because it's the sort of problem that lendsitself to strategic analysis.And so a lot of my work on negotiated transitions

  • 35:37

    IAN SHAPIRO [continued]: is informed by rational choice models.But on the other hand, if you wantto understand why people vote or questions like that--which people spent decades concoctingrational maximization models to try and accountfor why people vote-- and it's just ridiculous.

  • 35:59

    IAN SHAPIRO [continued]: Because they're not interested in strategic maximizationof anything in particular very often when they vote.They might be motivated by all kinds of things.And rational choice models are justgoing to make you focus on the wrong piece of the elephant.

  • 36:22

    IAN SHAPIRO [continued]: So that's, again, you've got to start with the problem and say,given the nature of the problem, whywould one think a rational choice model would be helpful?Well, here we have a situation where people are confrontedwith strategic dilemmas.Well, yes, then probably a good way to go.Whereas if we're looking-- studyingpolitical culture and political behavior,

  • 36:45

    IAN SHAPIRO [continued]: maybe the way people are raised isgoing to tell you more about whether they're likely to voteor not.Or how their parents vote will tell you more--or whether their parents vote, more likely,or what their income is.All kinds of things will tell you more than a rational choicemodel.

  • 37:10

    IAN SHAPIRO [continued]: I think that the biggest failing in political science graduateeducation is the divorcing of methods trainingfrom substantive training.What we've seen happen in political science departmentsin the last decade or so has been the expansionof methods training basically at the expense

  • 37:32

    IAN SHAPIRO [continued]: of substantive training.So typical-- when I went to graduate school in the 1970s,at least half the courses we took on the wayto qualifying exams were substantive courses.Introduction to Latin American politics or courseson democracy or courses on communist systems.

  • 37:56

    IAN SHAPIRO [continued]: Today, you'll find that first and second yearstudents in political science PhD programshave to take so many methods courses,and they're usually compulsory.And even if they're not formally compulsory,the students are made to feel that they can't get jobs

  • 38:16

    IAN SHAPIRO [continued]: if they don't take them.So they may take a two or three semester statistics sequence.They may take a two or three semester game theory sequence.They will take methods courses in research design.They'll take methods courses now increasinglyin field experimental design.

  • 38:37

    IAN SHAPIRO [continued]: And so what you get is people whoare teched up to the eyeballs who knowvery little about politics.Because the substantive courses either are not offered at all,or even if they're offered they nowtend to be cross-listed courses with undergraduates

  • 38:57

    IAN SHAPIRO [continued]: because there aren't enough graduate students to take them.And we're producing a generation of political scientistswho know frighteningly little about politics for alltheir technical sophistication.And my own view is that the best wayto learn methods is in the context of substantive courses.

  • 39:20

    IAN SHAPIRO [continued]: So the biggest single change I would makewould be to embed research training in substantive coursesabout politics rather than have this-- what's essentiallyturned out to be this crowding out effect wheremethods training is seen as the gold standard.

  • 39:45

    IAN SHAPIRO [continued]: And as a result, we have methodologically sophisticatedpeople who know less about politicsthan your average person who's reading The Economist.The effect of all this I think is twofold.One is it's producing a kind of political science

  • 40:06

    IAN SHAPIRO [continued]: that is more and more attracting method-driven peopleand less and less attracting people who are actuallyinterested in politics.Because technically sophisticatedpeople find technically sophisticated job applicantsinteresting.And so to some degree, you get a sort of self-replicating cycle.

  • 40:28

    IAN SHAPIRO [continued]: On the other hand, a reaction is setting in.One thing that's happening in the 19--in the 1990s and early 2000s, political science undergraduatemajors were in perpetual growth.They were the biggest majors in all universities,

  • 40:51

    IAN SHAPIRO [continued]: crowding out many other things.What we've seen in the last five to eight yearsin most major universities is the shrinkingof political science undergraduate majors.And the reason is undergraduates are interested in politics, notpolitical science.They really are problem-driven.Graduate students increasingly are interested

  • 41:13

    IAN SHAPIRO [continued]: in political science rather than in politics,where what they want to learn is the methods and soon of the discipline.But that then has knock-on effects in universities,because one of the reasons political sciencehas been such a thriving disciplineand there have been so many jobs it's because there's

  • 41:33

    IAN SHAPIRO [continued]: been so much demand for political scienceteaching in major universities for undergraduates.And now we're seeing the shrinkage of political scienceundergraduate majors.I was very interested to read-- I've noticed it at Yale,but it's an article about-- it's going on at Stanford where they

  • 41:55

    IAN SHAPIRO [continued]: are completely rehashing their undergraduate curriculum to tryand stem the outflow of students from the political sciencemajor at Stanford.Now I don't think it's all the resultof this change in the political science discipline.I think the lawyer glut has something to do with it.Poli sci's a big law major, and there's

  • 42:18

    IAN SHAPIRO [continued]: less interest in people going to law because there'slots of press about lawyers who can't get jobs.I also think the rise of things like global affairsmajors to some degree has eroded political science majors.But I think a big part of it is that whatthe people who are perceived as the cutting edgeresearchers in political science are interested in doing

  • 42:42

    IAN SHAPIRO [continued]: is of very little interest to undergraduateswho are problem-driven.And I think eventually that's goingto come home to roost as university administratorsnotice the shrinking demand for what political scientists doin their universities.And then outside of the academy as well,I think it spells trouble because I

  • 43:04

    IAN SHAPIRO [continued]: think it's harder and harder to make the case for researchfunding for work that outsiders can't understand and can'tsee the importance of.Well, the theorists who have most inspired me

  • 43:26

    IAN SHAPIRO [continued]: in thinking about the methodologiesof the social sciences, I would say Lakatoshhas influenced me the most in thathis view of science, of the progress of scienceis, I think, convincing.And it's, as you start with a problem,you then turn to the theories and methods that have

  • 43:46

    IAN SHAPIRO [continued]: been used to study it to date.And then you ask, how can I improve upon it?How can I explain what was explained before, and somethingmore?And it's a model of progress in scientific knowledgewhere the idea is to accumulate more knowledgerather than to make knowledge more certain.

  • 44:09

    IAN SHAPIRO [continued]: I think that even though I have stressed important differencesbetween the theory-driven people and the method-driven people,one thing that they both agree onis I think they have a kind of fetish for certainty.So the rational choice crowd in the '70s and '80swere very impressed by the possibility of having theorems.

  • 44:33

    IAN SHAPIRO [continued]: And one-- I can still remember one heated panelwhen our book was being much debated in the 1990s.And a particularly emphatic colleaguegot up at some APSA meeting and said,if you ain't got a theorem, you ain't got shit!It's sort of, having a theorem was the gold standard.

  • 44:58

    IAN SHAPIRO [continued]: With the experiments people, I thinkare also obsessed with this idea of certainty.They want a finding that you can take to the bank.They want something that you can say is definitely right.So this is the idea that, well, maybe theseare small questions, but at least we know the answer.And everything else is kind of windy carrying on.

  • 45:19

    IAN SHAPIRO [continued]: That's the zeitgeist.And it's the one thing that all rational choice peopleshare in common with the new method-driven peopleis this preoccupation with finding something certain.As Descartes said a long time ago,find propositions impossible to doubt.Certainty.Science advances by making knowledge more certain.

  • 45:42

    IAN SHAPIRO [continued]: That's the view.I think the Lakatoshian view, the Humean view, my view,the more-- a plausible view when you look at other sciencesis that all scientific knowledge is fallible and corrigible.As Weber said in that essay I mentioned earlier,even the best scientists expect their work

  • 46:02

    IAN SHAPIRO [continued]: eventually to be superseded.So the idea is to accumulate more knowledge, notto make knowledge more certain.And so you recognize that all knowledge claims are fallible,that you usually get things at least partly wrong.But that nonetheless, progress is possible in this Lakatoshiansense, that you can develop theories

  • 46:24

    IAN SHAPIRO [continued]: that explain what was explained before, plus something more.And it's a continual process of new knowledge displacingold knowledge.And it doesn't have this fetish with certaintythat I think sends people down one blind alley after another.

  • 46:52

    IAN SHAPIRO [continued]: Well, I think that the most interesting questionshave to do with the role of business in politics.I think that the big change of our generationis taking communism off the table.So for instance, going back to my earlier interest

  • 47:13

    IAN SHAPIRO [continued]: in distributive politics, if you readwork that was done in the '30s and '40sand '50s on distributive politics,the growth of modern welfare states and so on,

  • 47:33

    IAN SHAPIRO [continued]: everything that happened to some extenthappened in the shadow of communism.That is to say that if you read people--I read recently a biography of Averall Harriman.And there was very much a sense in the New Dealand in the Great Society years that we should create welfare

  • 47:57

    IAN SHAPIRO [continued]: states for a variety of reasons, but part of itis that you don't want the hearts and minds of workersin the West to be seduced by an alternative possibility.Once you take that off the table,I think the relationship between business and politics changesfundamentally.And so my recent work has been focused principally

  • 48:19

    IAN SHAPIRO [continued]: on the interaction between business and governmentin the post-communist era.And it takes us a long way from some of these debates we'vebeen talking about today.But that's where I'm headed.Well, I think the interesting questions about businessare that-- one way of putting it is if there's

  • 48:41

    IAN SHAPIRO [continued]: no real alternative to capitalism,the critique of capitalism sort of loses its point.And really, you start to become interestedin different types of capitalism.And part of my interest in businessis if they're not going to be part of the solution,they're going to be part of the problem.And so I'm most-- and this, again,

  • 49:02

    IAN SHAPIRO [continued]: comes out of some of my work on distributive coalitions--but I'm most interested in how onecan find incentives for business to support progressive change.A lot of my work is on Africa.And particularly there where you havefailed states and semi-failed stateswith demographic demands that can't possibly

  • 49:25

    IAN SHAPIRO [continued]: be met to educate populations and soon by traditional governments, how is it possible to getbusiness to participate in development in ways that arenot extractive but progressive?And so I think that's a really interesting set of challenges

  • 49:47

    IAN SHAPIRO [continued]: and questions.And so that that's the sort of example I'm thinking about.And so for that reason, that has driven me.The concern with that problem hasdriven me to look for examples of progressive roles playedby business historically.And actually, one of the papers I'm giving at APSA in 2015

  • 50:10

    IAN SHAPIRO [continued]: is about the role played by big business in the South Africantransition negotiations.It turned out they were very important behind the scenes.And so I'm trying to understand what their motives wereand whether there are any lessons from thatabout the sorts of circumstances in which business might

  • 50:33

    IAN SHAPIRO [continued]: actually even internalize some costsone wouldn't expect them to and playprogressive political roles.Again, so it's a research agenda that'scome out of the change in the problemin the world that is a result of the biggestchange of our lifetimes, which is taking communismoff the table.[MUSIC PLAYING]


Professor Ian Shapiro discusses the differences between problem-driven theorizing and method-driven or theory-driven theorizing in political science. Problem-driven theorizing takes a problem and examines previous attempts to solve it, then improves on the previous attempts to find a solution. Shapiro discusses his work with problem-driven theorizing, research he has done using this approach, and research directions he wants to take in the future.

Looks like you do not have access to this content.

Ian Shapiro Discusses Problem-Driven Research in Political Science

Professor Ian Shapiro discusses the differences between problem-driven theorizing and method-driven or theory-driven theorizing in political science. Problem-driven theorizing takes a problem and examines previous attempts to solve it, then improves on the previous attempts to find a solution. Shapiro discusses his work with problem-driven theorizing, research he has done using this approach, and research directions he wants to take in the future.

Copy and paste the following HTML into your website