The Rise and Fall of Development Economics provides us with a precursor to the Rodrik reading. While Rodrik saw the importance of looking at cases individually, Krugman emphasizes “silly” models that can help us explain things about the world even if they are oversimplified. I agree with Krugman in seeing the value in both ideas of thinking, and the importance of using both methods simultaneously. Krugman’s third conclusion about the Big-Push model was where the interdependence of models and experimentation really stuck out to me. As Krugman explains, while the Big Push theory might be true, we need to go out and test it in order to see (or not see) it in action. Additionally, the African map example was interesting because while I had never heard of the evolution of maps of Africa, the European’s methodology sounded like something that I would have followed as well. If you don’t know something for sure, don’t write it down. But what happens when you don’t know anything for sure? You get a blank map. From what we have learned about development economics so far, there have been some times of blank maps in its history because it is difficult to simplify such a complex subject in to a few lines on a graph. Even though Krugman wonders if the blank maps were necessary in development theory’s history, I would argue that it is a mute point to look back and ask that question. Instead, we need to be glad that this work has been done, and continue to work to both improve it and gather more data in the field to either credit or discredit the model.
It was very interesting to read another paper that calls into question the traditional focus on models in economics. Rodrik’s paper last week stated that economic models could not be translated into packages. Sometimes, he argued, these models limit policy makers by making them think there are specific recipes or packages for development that can be applied universally. However, he showed that this is not the case, and that traditional economic policies stated in the Washington Consensus are not always the best fit for a country. Similarly, here, the argument is that Hirschman’s high development theory was abandoned because its assumption of economies of scale did not fit traditional models. “Nobody knew how to model economies of scale,” and the theory died out.
It seems as though models can truly be a double-edged sword in economics. During the 1950’s, society demanded more empirical work from development theories. Though Hirschman’s theories previously had recognition, high development theory lost credibility for being too methodological. When models were suggested though, high development theory was further criticized for being impossible to model. It is interesting that I am studying something similar in my Communications Theory class. During this same decade, many mass media theorists were also criticized for lacking any empirical evidence in mass media theories. It seems as though there is an obsession to provide models and empirical evidence, even when they may not be necessary.
While models are important for teaching fundamentals, perhaps models are not the best approach for development economics. Rodrik’s paper showed that no two countries are alike when it comes to development. This author also states, “the relationship between good economic analysis and successful policy is far weaker than we like to imagine.” Perhaps the emphasis on models, then, should not exist for this branch of economics. The need for numbers and equations may limit some theories from expanding or force policy makers to skew their decisions.
That being said, I am not sure I can fully believe that Hirschman’s theory was abandoned simply because it did not fit into a model. I believe that there were other factors this paper ignored. It was a very dramatic article and at times perhaps too subjective. Nevertheless, I think the underlying question about how important models are to development economics is important.
I agree with the two bloggers above, the author does make an enlightening point that many including myself could have missed with regards to the development of economic models. He acknowledges the advantages that models brought to the table but emphasizes the point that the period when these models were adapted suffered loss of theories like HDT that could not be modeled.
The author gives great examples to illustrate his point but a key point I think he makes but does not deeply consider is the emphasis of the models at this time. In the 1950s the models that were developed focused on perfect competition and constant returns to scale, but none looked at economies of scale or oligopolies. I think that since these models were easier to build, it's what the school of economics focused on developing first. And the author mentions the need for simplicity in modelling. With time then, as economists got a good grip on how simple models worked and could be developed they begun to navigate more complex issues like oligopoly market structures and incorporate economies of scale.
Thus, it is arguable that it was not necessarily that HDT was forgotten but focus was changed towards understanding the development and use of models. More so, it can be argued that efforts by economists like Hirschman were put in place to allow HDT to remain a focus while models continued to be developed that could encompass this theory.
I agree with Alexandra’s point that in economics, models can be a double-edged sword, but maybe in some situations it goes beyond that and models are more harmful than good even with consciousness. Models provide an easy understanding of what is going on, but often (practically always) they are oversimplified. So even with a warning from your professor about the oversimplification, it is easier to understand the clear-cut model than it is to accept the complications and blank maps. Krugman says, “The problem is that there is no alternative to models. We all think in simplified models, all the time. The sophisticated thing to do is not to pretend to stop, but to be self-conscious -- to be aware that your models are maps rather than reality.” Self-awareness is obviously the key here, but what I want to know is when is an over-simplified model okay and when is it dangerous? For instance, Kuznet’s curve is an oversimplified model that happens to me “wrong” because there is no real correlation between income and its distribution yet, so is there any situation where that is appropriate, even if we are self-aware? I even think beyond development economics. Should some disciplines within economics (or all social sciences) avoid models all together based on the sheer complexity of the subject? Maybe monetary policy shouldn’t use models, while labor economics can. Are some oversimplifications, simply that, oversimplifications.
I thought that the comparison of European maps of Africa to the growing need for precise, accurate measurements in Economics set the tone and focus of the article well. I had not thought about the fact that old info being replaced by newer info and new standards could actually leave blanks in theories. This makes perfect sense however, and this metaphor example actually indirectly supports another important point made by Krugman about metaphors being models. This point is that whether we want to believe it or not, we explain and understand our world through models. Actually, until professor Casey referenced metaphors as a heuristic modeling technique in class, I would have been less inclined to think that models were as essential to development economics as I now do, especially after reading this article. Maybe, as Krugman suggests, we should not get caught up in the fact that we primarily use models. Instead, using the same quote as Samantha, we should be "self-conscious" in using models and know that they are "maps rather than reality". This way we hopefully will avoid "blanks" but still be able to progress in explaining development.
I thought that this piece had perfect timing for me, as we discussed the importance of models in a different social science, public policy, in POL 232 this morning. We concluded that the best models in public policy:
• order and simply reality
• identify what is significant
• are congruent with reality
• provide communication
• direct inquiry and research
• suggest explanations
In essence, as most previous bloggers have pointed out, is Krugman’s argument for models in economics. Models provide a way to communicate a complex thought in a more clear, straightforward way. Every model in public policy that we have come across so far has a “but” or a “sometimes, in this certain situation” attached to the end of it, like our economic assumptions. I think something that we can take away from the best practice list from public policy is how models should direct inquiry and research. Murphy et al.’s big push model shows, visually, how rational actors might get stuck in a low level trap, and why looking into how to get communities out of these traps is so important. It also shows where specifically the problem might, or then again might not be happening, directing research even further. As Krugman put it, “ Verbal expositions of the Big Push story make it seem like something that must be true. In this model we see that it is something that might be true. A model like this makes one want to go out and start measuring.”
I read this paper for my Econ of Social Issues class with Professor Goldsmith last semester, and the model that really stood out to me was the Fultz’s dishpan. This simple model of a dishpan, a turning table and some light for heat gave great insights into weather and ocean patterns. It was not meant to explain everything or take into account all of the details and variables associated with weather patterns, but it was a start and helped explain and identify some important weather patterns. Models can be great tools, and although they have limitations, I agree with Krugman’s argument that they should not be forgotten, even when they do not take into account all of the nuances of a particular situation. Several bloggers before also mentioned this idea. On the converse though I think it cannot be forgotten that models cannot explain everything.
I wonder at times if Krugman puts too much emphasis on how explaining things and not using models is wrong. I think that a combination of models and explanations/metaphors is the most appropriate. Models are not going to be able to explain everything and that is when I think more metaphors and written explanations might work. It would not devalue models, but it could fill in some gaps that models might create. Either way it is not perfect, because we have not learned how to describe many situations yet through economics, but maybe a combination will facilitate less loss of information before new information is gained. Maybe a combination will make it so that when the "map of Africa" is created and filled in, there will not be any information lost when new mapping techniques are developed.
The other bloggers seem to have come to an agreement that models should be used with caution and that we all must be cognizant of what we take away from them; they are a double-edged sword. The contrast between Rodrik’s emphasis on individual case studies and the importance of being context specific with Krugman’s emphasis on general models is striking and raises many questions, as these two seemingly opposing thought processes are in the same field.
I’m not sure how accurate this is, but it seems that models may be more fitting for certain variables but not others. For instance, when modeling education’s impact on child health, the production function is fairly worldwide: there is a positive relationship but there are diminishing marginal returns. However, when modeling Nepal’s transition to the modern sector, the model may not be so clear. Lucy said in her blog in relation to her public policy class that models should ‘order and simplify reality.’ However, the reality we focus on in public policy (Pol 232) is specific to the U.S. While a model may be useful for the bureaucratic system of America, it’s likely not useful for comparing development strategies between Nepal and Ghana. Nepal got a new government system in 1996, is largely influenced by the caste system, and has much religious strife between Hinduism and Buddhism. Ghana obtained its independence in 1957, has a large divide between the North and South and the many ethnic groups, and constantly faces resource extraction issues. I would argue that no model should be used to compare the development possibilities for these two countries.
So, while the simplification of models does allow us to gain some knowledge, maybe they should be confined to certain sectors of development economics. Human capital models but not labor market ones, etc.
As Jean has already mentioned in her comment, Krugman's paper reminded me a lot of Professor Goldsmith's Economics of Social Issue last year. Professor Goldsmith would often say "think like economists! Build models to explain a social phenomenon". This approach really helped me simplify and thus better understand particular issues in the contemporary society.
Unlike Jean and Kate however, I feel more optimistic about the usefulness of formal models in economics. Yes, Rodrik does argue in his paper that we should not restrict our understanding to a standardized set of ideas. However, he does acknowledges that there are indeed economic principles that are common to every successful nations: property rights, fiscal solvency, market oriented incentive, and sound money. The difference between these nations is that they adopted unique policy measures that implemented and utilized all these principles in a varying degree. Clearly, formal models and other generalization help us more easily observe a common theme.
Formal models are also handy in that it does not aim to conclusively declare an irrevocable truth. As Krugman argues, "verbal expositions...make [it] seem like something that 'must' be true. In...model we see that it is something that 'might' be true". In other words, formal models simplify a phenomenon and enable us to test the hypothesis before actually empirically proving it.
I thought that Krugman's metaphor of the African map was really appropriate to explain what we have been saying in class when we want to see the effect of an economic decision, and we end up in our "it depends,...". Complex models, as well as simple ones, would only give us an insight on the area/case/situation, they will fill in some gaps in our map, as the travelers' reports did to the African map. Simple observations, even the ones that do not have a very deep theoretical explanation can help as much as complex ones (as long as they are true). Perhaps, instead of starting with big generalized models as we do when we first take Econ 101 and 102, it would be better to start by examining smaller models that are more precise in their context (as we studied before in Rodriks article with the unorthodox decisions generating different results depending on where they were applied), and then build these models up to make more generalized ones or decide if we really want big models when we could have these more precise, simpler and smaller models, that could prevent the 'prescription' solutions to create economic growth. Going back to the metaphor, do we want the empty map of Africa, or the one with more details added by simpler observations (even if these do not have an exhaustive scientific proof)?
Like some of the bloggers above, Krugman's map of Africa metaphor really struck a tone with me and made the fall and rise of of development economics more sensible and easier to understand. While economists started to represent development economics with metaphors and not models, it was similar to the earlier maps with the small observations but inabilities to prove observations with scientific technologies. As economists started to use complex and simplemodels with scientific reasoning, metaphors disappeared, just like the interior of Africa as mapmaking improved. In the big picture, what we might have to consider is that not one model or metaphor holds the right answer. As economists we often consider our models with the disclaimer ceteris parabus, but how often do real-world applications hold such assumptions? Smaller models, which often have logical and credible evidence, are limited in application, while larger models, like a microeconomic case of supply and demand, make assumptions that aren't always true. Perhaps a solution to the model vs. metaphor debate is a combination of both models and observations or metaphors, as observations or metaphors could provide details, assuming they are true, while models provide the necessary scientific reasoning that can support or disclaim metaphors. In the end, it is clear from our readings and discussion that neither one metaphor nor a model alone can provide a complete understanding of development economics, just like a couple observations or a reading on a compass can't make the map of Africa.
Like the others who have posted on the forum I see Krugman’s point that models are necessary when looking at economics. I also think that their fear of oversimplification is also a valid point. Krugman believes that we must simplify in order to deal with complex systems because, after all, that is what economists are trying to do when they build these models. I agree with Krugman in that models are very useful tools especially when our resources limit us. However, as he gave his example of the dish-pan I was reminded of other models that had been created to use climate change. Back in the 1970s when computer models were first being used to determine climate change they were very inaccurate. They predicted an ice age in the coming years when in reality the world continued to get warmer. As the models became more advanced the scientists realized how wrong their previous models were. Because there are so many factors that go in to predicting what happens in a complex system it is often very difficult to know what will happen. Although this example isn’t perfect when comparing it to development economic models I believe it demonstrates Krugman’s point that models shouldn’t be taken too literally. While their results weren’t accurate the models still showed that human interaction with the climate could have drastic effects. However, unlike Krugman it also shows how dangerous oversimplification can be. Although he dismisses the inaccurate information found in the fifteenth century maps I believe this misinformation to be almost as dangerous as a blank map. A person who uses this map believes the information presented to him and will therefore make inaccurate conclusions about Africa, conclusions that are still reflected in our stereotypes about Africa today. While simplification may be a useful tool making inaccurate conclusions based on these models, no matter how conscious we are of the simplification we will still be making inaccurate conclusions. I believe it is very important to at least try and understand all of the factors that go in to a complex system, as difficult as this may be and that we should be trying to fill in all the blank spaces of the map.
As other people have stated, I really like two of the examples he gives to illustrate his point about economic models. The African cartography example shows that even when you think you are advancing in skill and knowledge, some other important aspects can be lost or forgotten for a period of time. I had never thought about this and it’s a great point he brings up. The other model, Fultz’s dishpan, explains parts of weather patterns, but it is so simple that it obviously cannot explain everything. I also really like Juan’s point that we, as humans, want answers. We want the exact “physics” answer with no uncertainty as the article states. In economic models, there are so many assumptions and “it depends” that it can be difficult to make conclusions. Models though overall are used to simplify and teach concepts so it is vey important to make them relatively simple and to have assumptions to make them work. On the other hand, it is essential not to oversimplify some complex concepts because then it does not do that specific concept justice and people are not able to learn what it is all about. The other option is that people not take these oversimplified models too literally. Experts need to explain what this simplifies and what would normally happen when taking into account all parts of the concept. One new idea I’ve thought of is: what direct effects, if any, has this “fall” in development economics affected third world countries? I ask this question because I believe this gap must have hurt the progress of development economics and thus individual countries.
Rodrik's piece last week concluded that model set a goof foundation for analyzing economies, but in order to be effective, and ultimately not misleading and damaging, these models need to be tailored to specific economies. Krugman analyzes in depth the effect that economists love of models can have on the development of theory. As Krugman says, physics is admired for its ability to be proven and quantified. Sadly, this has set the bench mark of what "good science" is and how it is conducted. Just as the field of development economics was coming into form in the mid 20th century, so were extremely prominent scientific philosophies, most notably Karl Popper's theory of falsifiability, which set the difference between science and non-science as science having falsifiability whereas non-science can not be proven wrong. This paradigm shift not only encouraged, but made it necessary, for scientists to try and narrow down and quantify their theories. A large, overarching and non quantifiable, or modeled here, theory is not falsifiable. Although this philosophy led to advancements in many aspects of natural sciences, it made it very difficult for young, fragile ideas to make it past that infantile stage.
High Development Theory was one of these casualties. Now knowing that HDT "made perfectly good sense after all", we are able to reflect on the fact that scientific "advancement" can actually lead to the de-evolution of some scientific fields. It is true that economic models can not accurately portray the real world. In every model I've done as an economics major their are always some factors that are being held constant. But this does not mean that these models are of no value, assuming that would be implying that every economic model holds no truth. The problem here is not with modeling, it is with the scientific community. I would hope that this serves as an example to current scientists and economists to not overlook proposed theorem simply due to its inability to be quantified and modeled out.
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
In the fall of last year I read this for a make-up session of Casey's macroeconomics class, and before opening the document to read it again I recalled a few things from last year: that it isn't a paper about Albert Hirschman, the story about the mapping of Africa, and the example of the simple climate model. I'm not sure if these are the points Krugman would want to stick, but in those two stories the pros and cons of modeling are summed up rather nicely. My classmates above have all discussed already how classroom models too often oversimplify the real world, but undergraduate econ models are still capable of illustrating many real world situations, such as why we haven't seen crowding out despite high government spending.
I agree with Andrew's post above that it would be enlightening to see if the "fall" of HDT affected the developing world, or if the the loss of of support among economists was accompanied with a parallel loss of support among policy makers. It seems often that what has widespread support among economists doesn't have much support among policy makers anyway, so it is possible it didn't have any pronounced effects on developing countries.
I think the temptation when reading this article is to take a pro-model or anti-model stance. Much like Kate said above, I believe there is a place for models and a place where they may not fit as well. Clearly this would include the early ideas behind any theory. If all ideas are thrown out for not being well-modelled, we would never be able to develop new theories, since ideas which have not yet been modelled generally stand at the beginning of any advancement.
Unlike Kate, I think modelling has a role at least at some point in every area of development economics. As she said, many of the other posts express a cautious approach to models. This caution is, in my opinion, excessive since the application of the model is much more dangerous than the model itself. To reference our discussion of the Washington Consensus, the problems were not with the models but with the misinterpretation of the models and unilateral belief that the models were capable of more than they had claimed.
Generally speaking, economic models do not claim to be universally true and perfectly applicable in every situation. Instead, they base a conclusion off a set of assumptions. This simple model does not show what will happen in every situation. Rather it identifies a trend. If models are used in this way, then there is no danger in models except the chance of the model proving false. Consequently I think it is important that we focus not on where models apply, but how they apply. To use the metaphor of Africa on European maps, it is important that we fill in as much as we can with models but not to ignore the parts we cannot yet model.
In my very limited exposure to the study of economics I’ve grappled with the sentiments expressed in this article. I’m also a math major, so I feel more comfortable with models, and ‘rules’ that hold in all situations. When one has a model, however ‘vague’ it might be, so much more (sometime unforeseen) information can be garnered. When I compare my econ classes to my math classes, the two contrast starkly. At no point in math can I write a proof as a ‘story;, weaving together some logical based statements and loosely realized results. But in econ one can’t find results that are simply always true, as one can in math. So personally, it took some time for me to adjust to the fact that there aren’t steadfast theorems, lemmas and laws in economics as there are in math.
As the article says, a balance of models and story telling are necessary to the advancement of developmental economics. For the most part the article has been thoroughly dissected by the comments above, so I wanted to share my personal realization of this exact phenomenon. The author took the right approach though, saying that the ignorance must go through stages to reach a place of less ignorance. Likewise, he acknowledged the need to find a balance, and the importance of simplified models.
I really enjoyed reading Kurgman’s article, and like many of my classmates agree that models are needed to understand and teach economics. However, it is important understand the limitation of models, especially very simplistic models that make many, sometimes unrealistic, assumptions. We need to be cognizant that models do not represent a whole phenomenon; that models show what the model-maker believed to be important and had the resources to model. As Kurgman said, in order to create models, “there is a narrowing of vision imposed by the limitations of one’s framework and tools.” He illustrates this idea with the example of the African map. Mapmakers of the 15th century only mapped the parts of Africa that they were absolutely certain about, leaving out many details, mainly of the interior, that were included in previous maps. It was not until the end of the 19th century that these details were added back to the maps. Kurgman also suggests that models are not the end-all-be-all, and that one should not “let important ideas slip by just because they haven’t been formulated your way.”
Similar to many of my classmates, I also found Krugman's Africa metaphor interesting. As a social science, the economics needs to balance numerical examples and models with an understanding of human nature and intuition. While it can be difficult to model human nature, it is important these models are developed to increase our understanding of economics. I personally agree with the post-1970 emphasis on these models as the foundation for our understanding. Through the economics classes I have taken at Washington and Lee I understand economists prefer the simplest models they can generate while still illustrating a point. This is to illustrate ideas as simply as possible, but does lead to some limitations. Despite my support of models in economic theory, I cannot entirely discount that models do not capture every element of human nature. In many ways humans are irrational and cannot be perfectly predicted through a model. As Curtis said, though, models do not claim they can perfectly predict human economic nature. Every economic journal I have ever read has a section that directly states the limitations of that model and suggests areas for further discussion. The numerical aspect of economics gives us a scientific and concrete approach to understanding human economic nature, but these models will always have their limitations. This was one of the reasons I initially became interested in economics; because it requires one to understand models, but still respect the limitations of those models.
Thanks Zach, for your repetitive insight into the article.
I did enjoy how Krugman sets himself up for this essay about the fall and rise of high development. He begins the article claiming not to be qualified to write such a paper on the topic. A nice literary technique as he then articulates an argument for the next seven thousand words. But I did find his narrative compelling.
His Norwegian meteorology folklore, maps of Africa, and dishpan analogies are all creative but still just analogies, literary techniques to drive an argument, and are not the focus of the paper. The focus is that the leader of high development theory, Hirschman, led his followers astray and into the darkness when he alienated his economic thoughts from mainstream economic by not conforming to the traditional model-deriving approach the social science. Here, in the darkness, the ideas fell apart and lost the zeal to inspire others. These ideas were not "teachable," as Krugman writes. According to Krugman, the drive to isolation is all due to one key facet, that economies of scale (crucial to high development theory) were difficult to include in formal, economic models. That simple, according to Krugman. Whether or not that is the only reason, I have no idea.
Krugman does an expert's job at outlining the factors that caused high development economic thought to evolve. It's fall was due to the mainstream economists attempting to derive complex models, which were apparently beyond the scope of high development economists at the time. Krugman goes on to argue that the break-through was the "silly" model constructed by Murphy and that if Hirschman and others had just been willing to construct simple models of their ideas, then the ideas wouldn't have stagnated for a generation. They didn't create simple models because it wasn't in fashion, economically speaking. It was more than out of fashion, according to Krugman, simple models were beaten down from the likes of economic intellectuals. But isn't the crux of economics built on simple models, supply and demand, that are obviously imperfect, but highlight very broad strokes of the economic discipline? From these foundations, the social science can grow into more developed and mature models. Surely Hirschman and others knew this... But I think it's wrong to argue that simplified models were ripped apart for their inadequacies.
The Rise and Fall of Development Economics provides us with a precursor to the Rodrik reading. While Rodrik saw the importance of looking at cases individually, Krugman emphasizes “silly” models that can help us explain things about the world even if they are oversimplified. I agree with Krugman in seeing the value in both ideas of thinking, and the importance of using both methods simultaneously. Krugman’s third conclusion about the Big-Push model was where the interdependence of models and experimentation really stuck out to me. As Krugman explains, while the Big Push theory might be true, we need to go out and test it in order to see (or not see) it in action. Additionally, the African map example was interesting because while I had never heard of the evolution of maps of Africa, the European’s methodology sounded like something that I would have followed as well. If you don’t know something for sure, don’t write it down. But what happens when you don’t know anything for sure? You get a blank map. From what we have learned about development economics so far, there have been some times of blank maps in its history because it is difficult to simplify such a complex subject in to a few lines on a graph. Even though Krugman wonders if the blank maps were necessary in development theory’s history, I would argue that it is a mute point to look back and ask that question. Instead, we need to be glad that this work has been done, and continue to work to both improve it and gather more data in the field to either credit or discredit the model.
Posted by: Madison Smith | 10/07/2014 at 09:28 AM
It was very interesting to read another paper that calls into question the traditional focus on models in economics. Rodrik’s paper last week stated that economic models could not be translated into packages. Sometimes, he argued, these models limit policy makers by making them think there are specific recipes or packages for development that can be applied universally. However, he showed that this is not the case, and that traditional economic policies stated in the Washington Consensus are not always the best fit for a country. Similarly, here, the argument is that Hirschman’s high development theory was abandoned because its assumption of economies of scale did not fit traditional models. “Nobody knew how to model economies of scale,” and the theory died out.
It seems as though models can truly be a double-edged sword in economics. During the 1950’s, society demanded more empirical work from development theories. Though Hirschman’s theories previously had recognition, high development theory lost credibility for being too methodological. When models were suggested though, high development theory was further criticized for being impossible to model. It is interesting that I am studying something similar in my Communications Theory class. During this same decade, many mass media theorists were also criticized for lacking any empirical evidence in mass media theories. It seems as though there is an obsession to provide models and empirical evidence, even when they may not be necessary.
While models are important for teaching fundamentals, perhaps models are not the best approach for development economics. Rodrik’s paper showed that no two countries are alike when it comes to development. This author also states, “the relationship between good economic analysis and successful policy is far weaker than we like to imagine.” Perhaps the emphasis on models, then, should not exist for this branch of economics. The need for numbers and equations may limit some theories from expanding or force policy makers to skew their decisions.
That being said, I am not sure I can fully believe that Hirschman’s theory was abandoned simply because it did not fit into a model. I believe that there were other factors this paper ignored. It was a very dramatic article and at times perhaps too subjective. Nevertheless, I think the underlying question about how important models are to development economics is important.
Posted by: Alexandra Butler | 10/07/2014 at 09:31 AM
I agree with the two bloggers above, the author does make an enlightening point that many including myself could have missed with regards to the development of economic models. He acknowledges the advantages that models brought to the table but emphasizes the point that the period when these models were adapted suffered loss of theories like HDT that could not be modeled.
The author gives great examples to illustrate his point but a key point I think he makes but does not deeply consider is the emphasis of the models at this time. In the 1950s the models that were developed focused on perfect competition and constant returns to scale, but none looked at economies of scale or oligopolies. I think that since these models were easier to build, it's what the school of economics focused on developing first. And the author mentions the need for simplicity in modelling. With time then, as economists got a good grip on how simple models worked and could be developed they begun to navigate more complex issues like oligopoly market structures and incorporate economies of scale.
Thus, it is arguable that it was not necessarily that HDT was forgotten but focus was changed towards understanding the development and use of models. More so, it can be argued that efforts by economists like Hirschman were put in place to allow HDT to remain a focus while models continued to be developed that could encompass this theory.
Posted by: Daphine Mugayo | 10/07/2014 at 12:44 PM
I agree with Alexandra’s point that in economics, models can be a double-edged sword, but maybe in some situations it goes beyond that and models are more harmful than good even with consciousness. Models provide an easy understanding of what is going on, but often (practically always) they are oversimplified. So even with a warning from your professor about the oversimplification, it is easier to understand the clear-cut model than it is to accept the complications and blank maps. Krugman says, “The problem is that there is no alternative to models. We all think in simplified models, all the time. The sophisticated thing to do is not to pretend to stop, but to be self-conscious -- to be aware that your models are maps rather than reality.” Self-awareness is obviously the key here, but what I want to know is when is an over-simplified model okay and when is it dangerous? For instance, Kuznet’s curve is an oversimplified model that happens to me “wrong” because there is no real correlation between income and its distribution yet, so is there any situation where that is appropriate, even if we are self-aware? I even think beyond development economics. Should some disciplines within economics (or all social sciences) avoid models all together based on the sheer complexity of the subject? Maybe monetary policy shouldn’t use models, while labor economics can. Are some oversimplifications, simply that, oversimplifications.
Posted by: Samantha Smith | 10/07/2014 at 03:00 PM
I thought that the comparison of European maps of Africa to the growing need for precise, accurate measurements in Economics set the tone and focus of the article well. I had not thought about the fact that old info being replaced by newer info and new standards could actually leave blanks in theories. This makes perfect sense however, and this metaphor example actually indirectly supports another important point made by Krugman about metaphors being models. This point is that whether we want to believe it or not, we explain and understand our world through models. Actually, until professor Casey referenced metaphors as a heuristic modeling technique in class, I would have been less inclined to think that models were as essential to development economics as I now do, especially after reading this article. Maybe, as Krugman suggests, we should not get caught up in the fact that we primarily use models. Instead, using the same quote as Samantha, we should be "self-conscious" in using models and know that they are "maps rather than reality". This way we hopefully will avoid "blanks" but still be able to progress in explaining development.
Posted by: C Wood | 10/07/2014 at 06:09 PM
I thought that this piece had perfect timing for me, as we discussed the importance of models in a different social science, public policy, in POL 232 this morning. We concluded that the best models in public policy:
• order and simply reality
• identify what is significant
• are congruent with reality
• provide communication
• direct inquiry and research
• suggest explanations
In essence, as most previous bloggers have pointed out, is Krugman’s argument for models in economics. Models provide a way to communicate a complex thought in a more clear, straightforward way. Every model in public policy that we have come across so far has a “but” or a “sometimes, in this certain situation” attached to the end of it, like our economic assumptions. I think something that we can take away from the best practice list from public policy is how models should direct inquiry and research. Murphy et al.’s big push model shows, visually, how rational actors might get stuck in a low level trap, and why looking into how to get communities out of these traps is so important. It also shows where specifically the problem might, or then again might not be happening, directing research even further. As Krugman put it, “ Verbal expositions of the Big Push story make it seem like something that must be true. In this model we see that it is something that might be true. A model like this makes one want to go out and start measuring.”
Posted by: Lucy Ortiz | 10/07/2014 at 09:37 PM
I read this paper for my Econ of Social Issues class with Professor Goldsmith last semester, and the model that really stood out to me was the Fultz’s dishpan. This simple model of a dishpan, a turning table and some light for heat gave great insights into weather and ocean patterns. It was not meant to explain everything or take into account all of the details and variables associated with weather patterns, but it was a start and helped explain and identify some important weather patterns. Models can be great tools, and although they have limitations, I agree with Krugman’s argument that they should not be forgotten, even when they do not take into account all of the nuances of a particular situation. Several bloggers before also mentioned this idea. On the converse though I think it cannot be forgotten that models cannot explain everything.
I wonder at times if Krugman puts too much emphasis on how explaining things and not using models is wrong. I think that a combination of models and explanations/metaphors is the most appropriate. Models are not going to be able to explain everything and that is when I think more metaphors and written explanations might work. It would not devalue models, but it could fill in some gaps that models might create. Either way it is not perfect, because we have not learned how to describe many situations yet through economics, but maybe a combination will facilitate less loss of information before new information is gained. Maybe a combination will make it so that when the "map of Africa" is created and filled in, there will not be any information lost when new mapping techniques are developed.
Posted by: Jean Turlington | 10/07/2014 at 09:52 PM
The other bloggers seem to have come to an agreement that models should be used with caution and that we all must be cognizant of what we take away from them; they are a double-edged sword. The contrast between Rodrik’s emphasis on individual case studies and the importance of being context specific with Krugman’s emphasis on general models is striking and raises many questions, as these two seemingly opposing thought processes are in the same field.
I’m not sure how accurate this is, but it seems that models may be more fitting for certain variables but not others. For instance, when modeling education’s impact on child health, the production function is fairly worldwide: there is a positive relationship but there are diminishing marginal returns. However, when modeling Nepal’s transition to the modern sector, the model may not be so clear. Lucy said in her blog in relation to her public policy class that models should ‘order and simplify reality.’ However, the reality we focus on in public policy (Pol 232) is specific to the U.S. While a model may be useful for the bureaucratic system of America, it’s likely not useful for comparing development strategies between Nepal and Ghana. Nepal got a new government system in 1996, is largely influenced by the caste system, and has much religious strife between Hinduism and Buddhism. Ghana obtained its independence in 1957, has a large divide between the North and South and the many ethnic groups, and constantly faces resource extraction issues. I would argue that no model should be used to compare the development possibilities for these two countries.
So, while the simplification of models does allow us to gain some knowledge, maybe they should be confined to certain sectors of development economics. Human capital models but not labor market ones, etc.
Posted by: Kate LeMasters | 10/07/2014 at 11:14 PM
As Jean has already mentioned in her comment, Krugman's paper reminded me a lot of Professor Goldsmith's Economics of Social Issue last year. Professor Goldsmith would often say "think like economists! Build models to explain a social phenomenon". This approach really helped me simplify and thus better understand particular issues in the contemporary society.
Unlike Jean and Kate however, I feel more optimistic about the usefulness of formal models in economics. Yes, Rodrik does argue in his paper that we should not restrict our understanding to a standardized set of ideas. However, he does acknowledges that there are indeed economic principles that are common to every successful nations: property rights, fiscal solvency, market oriented incentive, and sound money. The difference between these nations is that they adopted unique policy measures that implemented and utilized all these principles in a varying degree. Clearly, formal models and other generalization help us more easily observe a common theme.
Formal models are also handy in that it does not aim to conclusively declare an irrevocable truth. As Krugman argues, "verbal expositions...make [it] seem like something that 'must' be true. In...model we see that it is something that 'might' be true". In other words, formal models simplify a phenomenon and enable us to test the hypothesis before actually empirically proving it.
Posted by: HeeJu Jang | 10/08/2014 at 12:56 AM
I thought that Krugman's metaphor of the African map was really appropriate to explain what we have been saying in class when we want to see the effect of an economic decision, and we end up in our "it depends,...". Complex models, as well as simple ones, would only give us an insight on the area/case/situation, they will fill in some gaps in our map, as the travelers' reports did to the African map. Simple observations, even the ones that do not have a very deep theoretical explanation can help as much as complex ones (as long as they are true). Perhaps, instead of starting with big generalized models as we do when we first take Econ 101 and 102, it would be better to start by examining smaller models that are more precise in their context (as we studied before in Rodriks article with the unorthodox decisions generating different results depending on where they were applied), and then build these models up to make more generalized ones or decide if we really want big models when we could have these more precise, simpler and smaller models, that could prevent the 'prescription' solutions to create economic growth. Going back to the metaphor, do we want the empty map of Africa, or the one with more details added by simpler observations (even if these do not have an exhaustive scientific proof)?
Posted by: Juan Cruz Mayol | 10/08/2014 at 01:12 AM
Like some of the bloggers above, Krugman's map of Africa metaphor really struck a tone with me and made the fall and rise of of development economics more sensible and easier to understand. While economists started to represent development economics with metaphors and not models, it was similar to the earlier maps with the small observations but inabilities to prove observations with scientific technologies. As economists started to use complex and simplemodels with scientific reasoning, metaphors disappeared, just like the interior of Africa as mapmaking improved. In the big picture, what we might have to consider is that not one model or metaphor holds the right answer. As economists we often consider our models with the disclaimer ceteris parabus, but how often do real-world applications hold such assumptions? Smaller models, which often have logical and credible evidence, are limited in application, while larger models, like a microeconomic case of supply and demand, make assumptions that aren't always true. Perhaps a solution to the model vs. metaphor debate is a combination of both models and observations or metaphors, as observations or metaphors could provide details, assuming they are true, while models provide the necessary scientific reasoning that can support or disclaim metaphors. In the end, it is clear from our readings and discussion that neither one metaphor nor a model alone can provide a complete understanding of development economics, just like a couple observations or a reading on a compass can't make the map of Africa.
Posted by: Raymond Monasterski | 10/08/2014 at 12:17 PM
Like the others who have posted on the forum I see Krugman’s point that models are necessary when looking at economics. I also think that their fear of oversimplification is also a valid point. Krugman believes that we must simplify in order to deal with complex systems because, after all, that is what economists are trying to do when they build these models. I agree with Krugman in that models are very useful tools especially when our resources limit us. However, as he gave his example of the dish-pan I was reminded of other models that had been created to use climate change. Back in the 1970s when computer models were first being used to determine climate change they were very inaccurate. They predicted an ice age in the coming years when in reality the world continued to get warmer. As the models became more advanced the scientists realized how wrong their previous models were. Because there are so many factors that go in to predicting what happens in a complex system it is often very difficult to know what will happen. Although this example isn’t perfect when comparing it to development economic models I believe it demonstrates Krugman’s point that models shouldn’t be taken too literally. While their results weren’t accurate the models still showed that human interaction with the climate could have drastic effects. However, unlike Krugman it also shows how dangerous oversimplification can be. Although he dismisses the inaccurate information found in the fifteenth century maps I believe this misinformation to be almost as dangerous as a blank map. A person who uses this map believes the information presented to him and will therefore make inaccurate conclusions about Africa, conclusions that are still reflected in our stereotypes about Africa today. While simplification may be a useful tool making inaccurate conclusions based on these models, no matter how conscious we are of the simplification we will still be making inaccurate conclusions. I believe it is very important to at least try and understand all of the factors that go in to a complex system, as difficult as this may be and that we should be trying to fill in all the blank spaces of the map.
Posted by: Taylor Theodossiou | 10/08/2014 at 01:21 PM
As other people have stated, I really like two of the examples he gives to illustrate his point about economic models. The African cartography example shows that even when you think you are advancing in skill and knowledge, some other important aspects can be lost or forgotten for a period of time. I had never thought about this and it’s a great point he brings up. The other model, Fultz’s dishpan, explains parts of weather patterns, but it is so simple that it obviously cannot explain everything. I also really like Juan’s point that we, as humans, want answers. We want the exact “physics” answer with no uncertainty as the article states. In economic models, there are so many assumptions and “it depends” that it can be difficult to make conclusions. Models though overall are used to simplify and teach concepts so it is vey important to make them relatively simple and to have assumptions to make them work. On the other hand, it is essential not to oversimplify some complex concepts because then it does not do that specific concept justice and people are not able to learn what it is all about. The other option is that people not take these oversimplified models too literally. Experts need to explain what this simplifies and what would normally happen when taking into account all parts of the concept. One new idea I’ve thought of is: what direct effects, if any, has this “fall” in development economics affected third world countries? I ask this question because I believe this gap must have hurt the progress of development economics and thus individual countries.
Posted by: Andrew Riehl | 10/08/2014 at 03:01 PM
Rodrik's piece last week concluded that model set a goof foundation for analyzing economies, but in order to be effective, and ultimately not misleading and damaging, these models need to be tailored to specific economies. Krugman analyzes in depth the effect that economists love of models can have on the development of theory. As Krugman says, physics is admired for its ability to be proven and quantified. Sadly, this has set the bench mark of what "good science" is and how it is conducted. Just as the field of development economics was coming into form in the mid 20th century, so were extremely prominent scientific philosophies, most notably Karl Popper's theory of falsifiability, which set the difference between science and non-science as science having falsifiability whereas non-science can not be proven wrong. This paradigm shift not only encouraged, but made it necessary, for scientists to try and narrow down and quantify their theories. A large, overarching and non quantifiable, or modeled here, theory is not falsifiable. Although this philosophy led to advancements in many aspects of natural sciences, it made it very difficult for young, fragile ideas to make it past that infantile stage.
High Development Theory was one of these casualties. Now knowing that HDT "made perfectly good sense after all", we are able to reflect on the fact that scientific "advancement" can actually lead to the de-evolution of some scientific fields. It is true that economic models can not accurately portray the real world. In every model I've done as an economics major their are always some factors that are being held constant. But this does not mean that these models are of no value, assuming that would be implying that every economic model holds no truth. The problem here is not with modeling, it is with the scientific community. I would hope that this serves as an example to current scientists and economists to not overlook proposed theorem simply due to its inability to be quantified and modeled out.
Posted by: Bennett Henson | 10/08/2014 at 03:08 PM
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
Posted by: Zach Colby | 10/08/2014 at 03:18 PM
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
Posted by: Zach Colby | 10/08/2014 at 03:19 PM
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
Posted by: Zach Colby | 10/08/2014 at 03:19 PM
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
Posted by: Zach Colby | 10/08/2014 at 03:19 PM
I agree with Krugman and many of the bloggers above that models are very important to use in understanding the phenomenons of our world. Models can tell us so much and really help to paint pictures of how theory explains real world outcomes. As many have already stated, every situation is unique and complicated, but most situations can also be explained by a simple model. When thinking about the success that models can have I remembered reading an excerpt of Robert Nozick's "Anarchy, State and Utopia" for Pov 101. Now reading Nozick made me think A LOT and really question how I thought about a lot of things. In debating in favor of Nozick I imagined a completely abstract and theoretical world where conditions of models do hold, and in this world models would be successful and policies would be easy to create. Alas, this is not the case in our very complicated world but it does prove a point that there is some truth behind models. I know there aren't too many true anarcho-capitalists or objectivists out there but models can be pretty darn effective.
Posted by: Zach Colby | 10/08/2014 at 03:19 PM
In the fall of last year I read this for a make-up session of Casey's macroeconomics class, and before opening the document to read it again I recalled a few things from last year: that it isn't a paper about Albert Hirschman, the story about the mapping of Africa, and the example of the simple climate model. I'm not sure if these are the points Krugman would want to stick, but in those two stories the pros and cons of modeling are summed up rather nicely. My classmates above have all discussed already how classroom models too often oversimplify the real world, but undergraduate econ models are still capable of illustrating many real world situations, such as why we haven't seen crowding out despite high government spending.
I agree with Andrew's post above that it would be enlightening to see if the "fall" of HDT affected the developing world, or if the the loss of of support among economists was accompanied with a parallel loss of support among policy makers. It seems often that what has widespread support among economists doesn't have much support among policy makers anyway, so it is possible it didn't have any pronounced effects on developing countries.
Posted by: Jacob Strauss | 10/08/2014 at 03:39 PM
I think the temptation when reading this article is to take a pro-model or anti-model stance. Much like Kate said above, I believe there is a place for models and a place where they may not fit as well. Clearly this would include the early ideas behind any theory. If all ideas are thrown out for not being well-modelled, we would never be able to develop new theories, since ideas which have not yet been modelled generally stand at the beginning of any advancement.
Unlike Kate, I think modelling has a role at least at some point in every area of development economics. As she said, many of the other posts express a cautious approach to models. This caution is, in my opinion, excessive since the application of the model is much more dangerous than the model itself. To reference our discussion of the Washington Consensus, the problems were not with the models but with the misinterpretation of the models and unilateral belief that the models were capable of more than they had claimed.
Generally speaking, economic models do not claim to be universally true and perfectly applicable in every situation. Instead, they base a conclusion off a set of assumptions. This simple model does not show what will happen in every situation. Rather it identifies a trend. If models are used in this way, then there is no danger in models except the chance of the model proving false. Consequently I think it is important that we focus not on where models apply, but how they apply. To use the metaphor of Africa on European maps, it is important that we fill in as much as we can with models but not to ignore the parts we cannot yet model.
Posted by: Curtis Jay Correll | 10/08/2014 at 04:49 PM
In my very limited exposure to the study of economics I’ve grappled with the sentiments expressed in this article. I’m also a math major, so I feel more comfortable with models, and ‘rules’ that hold in all situations. When one has a model, however ‘vague’ it might be, so much more (sometime unforeseen) information can be garnered. When I compare my econ classes to my math classes, the two contrast starkly. At no point in math can I write a proof as a ‘story;, weaving together some logical based statements and loosely realized results. But in econ one can’t find results that are simply always true, as one can in math. So personally, it took some time for me to adjust to the fact that there aren’t steadfast theorems, lemmas and laws in economics as there are in math.
As the article says, a balance of models and story telling are necessary to the advancement of developmental economics. For the most part the article has been thoroughly dissected by the comments above, so I wanted to share my personal realization of this exact phenomenon. The author took the right approach though, saying that the ignorance must go through stages to reach a place of less ignorance. Likewise, he acknowledged the need to find a balance, and the importance of simplified models.
Posted by: Austin Hay | 10/08/2014 at 05:20 PM
I really enjoyed reading Kurgman’s article, and like many of my classmates agree that models are needed to understand and teach economics. However, it is important understand the limitation of models, especially very simplistic models that make many, sometimes unrealistic, assumptions. We need to be cognizant that models do not represent a whole phenomenon; that models show what the model-maker believed to be important and had the resources to model. As Kurgman said, in order to create models, “there is a narrowing of vision imposed by the limitations of one’s framework and tools.” He illustrates this idea with the example of the African map. Mapmakers of the 15th century only mapped the parts of Africa that they were absolutely certain about, leaving out many details, mainly of the interior, that were included in previous maps. It was not until the end of the 19th century that these details were added back to the maps. Kurgman also suggests that models are not the end-all-be-all, and that one should not “let important ideas slip by just because they haven’t been formulated your way.”
Posted by: Bobby DeStefano | 10/08/2014 at 05:32 PM
Similar to many of my classmates, I also found Krugman's Africa metaphor interesting. As a social science, the economics needs to balance numerical examples and models with an understanding of human nature and intuition. While it can be difficult to model human nature, it is important these models are developed to increase our understanding of economics. I personally agree with the post-1970 emphasis on these models as the foundation for our understanding. Through the economics classes I have taken at Washington and Lee I understand economists prefer the simplest models they can generate while still illustrating a point. This is to illustrate ideas as simply as possible, but does lead to some limitations. Despite my support of models in economic theory, I cannot entirely discount that models do not capture every element of human nature. In many ways humans are irrational and cannot be perfectly predicted through a model. As Curtis said, though, models do not claim they can perfectly predict human economic nature. Every economic journal I have ever read has a section that directly states the limitations of that model and suggests areas for further discussion. The numerical aspect of economics gives us a scientific and concrete approach to understanding human economic nature, but these models will always have their limitations. This was one of the reasons I initially became interested in economics; because it requires one to understand models, but still respect the limitations of those models.
Posted by: Stephen Moore | 10/08/2014 at 05:36 PM
Thanks Zach, for your repetitive insight into the article.
I did enjoy how Krugman sets himself up for this essay about the fall and rise of high development. He begins the article claiming not to be qualified to write such a paper on the topic. A nice literary technique as he then articulates an argument for the next seven thousand words. But I did find his narrative compelling.
His Norwegian meteorology folklore, maps of Africa, and dishpan analogies are all creative but still just analogies, literary techniques to drive an argument, and are not the focus of the paper. The focus is that the leader of high development theory, Hirschman, led his followers astray and into the darkness when he alienated his economic thoughts from mainstream economic by not conforming to the traditional model-deriving approach the social science. Here, in the darkness, the ideas fell apart and lost the zeal to inspire others. These ideas were not "teachable," as Krugman writes. According to Krugman, the drive to isolation is all due to one key facet, that economies of scale (crucial to high development theory) were difficult to include in formal, economic models. That simple, according to Krugman. Whether or not that is the only reason, I have no idea.
Krugman does an expert's job at outlining the factors that caused high development economic thought to evolve. It's fall was due to the mainstream economists attempting to derive complex models, which were apparently beyond the scope of high development economists at the time. Krugman goes on to argue that the break-through was the "silly" model constructed by Murphy and that if Hirschman and others had just been willing to construct simple models of their ideas, then the ideas wouldn't have stagnated for a generation. They didn't create simple models because it wasn't in fashion, economically speaking. It was more than out of fashion, according to Krugman, simple models were beaten down from the likes of economic intellectuals. But isn't the crux of economics built on simple models, supply and demand, that are obviously imperfect, but highlight very broad strokes of the economic discipline? From these foundations, the social science can grow into more developed and mature models. Surely Hirschman and others knew this... But I think it's wrong to argue that simplified models were ripped apart for their inadequacies.
Posted by: Wilson Hallett | 10/08/2014 at 06:03 PM