Data speak louder than politicians.
countries
trade-surplus GDP trade/surplus/GDP (‰)
Kuwait
48.039 148,024 0.324535211
Saudi Arabia
95.762 369,179 0.259391786
Algeria
30.6 140,577 0.217674299
Singapore
39.157 182,232 0.214874446
Iran
70.797 331,015 0.213878525
Norway
59.983 381,766 0.157119806
Malaysia
29.181 191,601 0.152300875
United Arab Emirates
39.113 261,348 0.149658693
Hong Kong
28.038 215,355 0.13019433
Sweden
38.797 406,072 0.09554217
Republic of China (Taiwan)
32.979 379,000 0.087015831
Netherlands
52.522 792,128 0.066304941
Russia
76.163 1,230,726 0.061884611
People's Republic of China
296.2 4,909,280
0.060334713
Switzerland
28.776 500,260 0.057522089
Germany
109.7 3346702 0.032778538
Japan
131.2 5067526 0.025890346
Thursday, September 16, 2010
some thoughts on the growth theory
I have outlined some specific crtique of the growth theory in a previous blog. In this entry, I cannot give any thoughts on modeling, but merely talk about something that the new models should strive to explain, because it has been generally lacking in the literature.
First, I guess the most obvious absence concerns corporations. The models have not given a uitable role for corporations, given our obsession of national corporations. I think one core issue that needs to be addressed before a good model can be developed is that whether the nationality of the corporation has any influence over the decision of location, employment and managerial stuff. I think the answer is absolutely positive given the cost of information and existence of transaction, cost of monitoring, and other forms of friction imposed by language and culture. What is needed here is to quantify these effects, and I expect them to be heterogeneous among countries and regions. My hunch is that these friction cost will be smaller inamong European countries than among other regions. I feel teh absolute value of this friction will have some important implications for the growth model. This primitive model might explain the
Second, the distribution of wealth.
Third, the role of financial market and things like exchange rate regime
First, I guess the most obvious absence concerns corporations. The models have not given a uitable role for corporations, given our obsession of national corporations. I think one core issue that needs to be addressed before a good model can be developed is that whether the nationality of the corporation has any influence over the decision of location, employment and managerial stuff. I think the answer is absolutely positive given the cost of information and existence of transaction, cost of monitoring, and other forms of friction imposed by language and culture. What is needed here is to quantify these effects, and I expect them to be heterogeneous among countries and regions. My hunch is that these friction cost will be smaller inamong European countries than among other regions. I feel teh absolute value of this friction will have some important implications for the growth model. This primitive model might explain the
Second, the distribution of wealth.
Third, the role of financial market and things like exchange rate regime
Wednesday, September 15, 2010
a rational view of superstition
I. A baseline model of superstition
superstition, that is common,. for example, people in China do not want to write will, because they fear that the fact that they write will may cause death. In other words, superstition is a belief that action A will cause event B (event B is an undesirable thing). This causality need not be absolute. More formally, let's suppose the probability that B will take place is f(x,y), where x is a vector denoting all other relevant aspects of life, while y takes on value about action A. y can take on only two values 0 and 1 (1 means making action A, and 0 not) in this case. and for any x0, f(x0,0)
This is the basic model of superstition, however it is incomplete and naive in that it assumes no dynamics and adaptation or expectation. People's belief does evolve. People is not always 100% sure about the superstition. From now on, I will only look at the continuous case, and the result can be easily generalized to discrete case. There two possibilities--1) superstition is true. 2) superstition is false. There are different probabilities involved. Here I take Baysian view of probability. People adapt this belief with the actual data.
III. A complete model of choice under the influence of superstition
1. action A-->event B (desirable)
to be added (key:cost of A)
2. action A-->event B(undesirable)
to be added
IV.Example and explanation (to be revised)
Suppose that superstition goes that action A leads to event B (undesirable). If that superstition is rampant in that region, then it is likely that people will have a a priori distribution favoring the superstition is true. Because people is risk-averse, people are not likely to make action A. Thus there will be few data and people with this superstition cannot update their belief. thus that superstition may continue. Even in the presence of data, incomplete information (non-dissemination, distortion) and wrong interpretation ( guilt of association, causality vs. association) may fail to dispute the superstition.
(why superstition are prevalent?)
V. Afterword
This is just a small game with behavioral economics. I love economics, maths and so forth not because I like the manipulation of numbers and functions, but rather, they leads to a nice way of looking at the world. It is fun and that is all that matters.
superstition, that is common,. for example, people in China do not want to write will, because they fear that the fact that they write will may cause death. In other words, superstition is a belief that action A will cause event B (event B is an undesirable thing). This causality need not be absolute. More formally, let's suppose the probability that B will take place is f(x,y), where x is a vector denoting all other relevant aspects of life, while y takes on value about action A. y can take on only two values 0 and 1 (1 means making action A, and 0 not) in this case. and for any x0, f(x0,0)
This is the basic model of superstition, however it is incomplete and naive in that it assumes no dynamics and adaptation or expectation. People's belief does evolve. People is not always 100% sure about the superstition. From now on, I will only look at the continuous case, and the result can be easily generalized to discrete case. There two possibilities--1) superstition is true. 2) superstition is false. There are different probabilities involved. Here I take Baysian view of probability. People adapt this belief with the actual data.
III. A complete model of choice under the influence of superstition
1. action A-->event B (desirable)
to be added (key:cost of A)
2. action A-->event B(undesirable)
to be added
IV.Example and explanation (to be revised)
Suppose that superstition goes that action A leads to event B (undesirable). If that superstition is rampant in that region, then it is likely that people will have a a priori distribution favoring the superstition is true. Because people is risk-averse, people are not likely to make action A. Thus there will be few data and people with this superstition cannot update their belief. thus that superstition may continue. Even in the presence of data, incomplete information (non-dissemination, distortion) and wrong interpretation ( guilt of association, causality vs. association) may fail to dispute the superstition.
(why superstition are prevalent?)
V. Afterword
This is just a small game with behavioral economics. I love economics, maths and so forth not because I like the manipulation of numbers and functions, but rather, they leads to a nice way of looking at the world. It is fun and that is all that matters.
critique of Solow Model
One serious drawback of the Solow Model the single good it has in the model has only homogeneous quality. This is an unreasonable assumption, with consequences in conclusion. Because of this set up, technology innovation has to be interpreted as efficiency improvement--that is with the same input of labr and capital, the economy is able to produce much more. However, this assumption is unrealistic. In most areas of the world, what can be witnessed is that techonology improves the quality of the products, leading to so-called "climbing the quality ladder". However, if we model the technological innovation as improvements in quality and thus introducing a continuum of quality of the goods, we can accomplish much more. We then have to introduce multiple goods, since quality only matters when there is substitution. But to do this, we also have to abandon the competitive market assumption, without which the whole model could be ridiculously complicated. What I hope to show is that, when there is a small difference in A(t), the less developed counties could still specialize that technology-intensive industry, but reaping less benefit. Another possible consequence is that with higher growth rate (that is higher dA/dt), less developed countries might find it harder to catch up.
This is only a sketch of the model, there are so many things left out and not thought out. Also, currently I do not have the ability to formalize this hypothesis.
Also, I have the crunch that if monetary stuff are introduced, then we might have a model to explain the exchange rate level's influence on growth rate and catching up.
This is only a sketch of the model, there are so many things left out and not thought out. Also, currently I do not have the ability to formalize this hypothesis.
Also, I have the crunch that if monetary stuff are introduced, then we might have a model to explain the exchange rate level's influence on growth rate and catching up.
Tuesday, September 14, 2010
luck as an explanation in cross-country difference in growth
Luck as an explantion has been explored and dismissed by the economics profession. However, from what I read, there are two pieces missing from the luck explanation.
1) Luck is largely explained in terms of multiple equilibria. Many economists oppose this by saying this explanation might explain difference in the short run, but fail to explain the long run stagnation as the country should have coordinated better. Firstly, time allows for coordination is doubtful. Second, and more pertinent to my point is that these economists neglect the dynamics inherent in the economics matter. One result of equilibrium have influence over latter developments. Because of an innovation as a result of a good equilibrium, the economy might be able to revolutionize its whole economy, making itself much more advanced than other economies. They might gain from this afvantage by enjoying more surplus from trade. Apersonal experience on this dynamics is ....
2) randomness
1) Luck is largely explained in terms of multiple equilibria. Many economists oppose this by saying this explanation might explain difference in the short run, but fail to explain the long run stagnation as the country should have coordinated better. Firstly, time allows for coordination is doubtful. Second, and more pertinent to my point is that these economists neglect the dynamics inherent in the economics matter. One result of equilibrium have influence over latter developments. Because of an innovation as a result of a good equilibrium, the economy might be able to revolutionize its whole economy, making itself much more advanced than other economies. They might gain from this afvantage by enjoying more surplus from trade. Apersonal experience on this dynamics is ....
2) randomness
A New way of calculating GDP and its implications for Theories
As proposed by Nordhaus, there should be more scientific way of national accounting for nonmarket accounts. The most important two things for me is 1)nonreproducible assets and 2) non-market activities. These two have implications for growth theory.
1) Oil-rich countries create noise in our data. They are indeed poor countries, since they produce less than what is being accounted for. Without this noise, it is likely that, when plotting growth rate with log of GDP per capita, the band at the wealthy end will be with almost no outliers (with the exception of Luxembourg). Indeed there is reason to kick them out of growth data all together, since they present a very different story from other countries.
2) Growth rate in poor countries might be over-estimated because part of the increase in GDP might result from transformation of nonmarket activites into market activities, thus weakening the case for convergence.
1) Oil-rich countries create noise in our data. They are indeed poor countries, since they produce less than what is being accounted for. Without this noise, it is likely that, when plotting growth rate with log of GDP per capita, the band at the wealthy end will be with almost no outliers (with the exception of Luxembourg). Indeed there is reason to kick them out of growth data all together, since they present a very different story from other countries.
2) Growth rate in poor countries might be over-estimated because part of the increase in GDP might result from transformation of nonmarket activites into market activities, thus weakening the case for convergence.
convergence or divergence
When looking at data that plot average growth rate of GDP to log GDP per worker/capital, evidence of convergence can never been found. Unless we restrict the countries to OECD countries, introducing once again sampling bias. There are might be three reasons behind this lack of backing from data.
1. true lack of convergence. This may imply the standard growth model that economists have been so happy with are incomplete at least or wrong.
2. The fundamentals (I mean fundamentals in a narrow sense--institutions mainly, not including education, which I would want to be modeled endogenously) are different. It is very reasonable to expect that rich countries tend to have good fundamentals while poor countries might not.
3. Poor countries tend to have reversals in fundamentals due to political instability brought by poverty. this point is similar to point 2, but there is a subtle difference in that, point 2 takes a static view of poor countries, on average, have bad institutions. While this take a more dynamic view. If I take a Markov Matrix to represent the chance of countries staying on or changing into different state of institutions (effective, mediocre, and bad), the function of going from good to bad (relatively) should be a decreasing function in the wealth of the country. This can be partially confirmed by the shape of the distribution. The band of growth rate is often narrow at the wealthy end, and VERY broad at the poor end. ( there is some outlier at the wealthy end and I will talk about these in "A New Way of calculating GDP and Its Implications for Theories"). If we clean out periods of regression resulting from institutions reversals, we might get good evidence for convergence.
1. true lack of convergence. This may imply the standard growth model that economists have been so happy with are incomplete at least or wrong.
2. The fundamentals (I mean fundamentals in a narrow sense--institutions mainly, not including education, which I would want to be modeled endogenously) are different. It is very reasonable to expect that rich countries tend to have good fundamentals while poor countries might not.
3. Poor countries tend to have reversals in fundamentals due to political instability brought by poverty. this point is similar to point 2, but there is a subtle difference in that, point 2 takes a static view of poor countries, on average, have bad institutions. While this take a more dynamic view. If I take a Markov Matrix to represent the chance of countries staying on or changing into different state of institutions (effective, mediocre, and bad), the function of going from good to bad (relatively) should be a decreasing function in the wealth of the country. This can be partially confirmed by the shape of the distribution. The band of growth rate is often narrow at the wealthy end, and VERY broad at the poor end. ( there is some outlier at the wealthy end and I will talk about these in "A New Way of calculating GDP and Its Implications for Theories"). If we clean out periods of regression resulting from institutions reversals, we might get good evidence for convergence.
on technological progress and growth
Prof. Love said I have comparative advantage in trade, analyzing things with heterogeneity. Ok, let me do this.
In growth theory, so many models (Ramsey, Solow, Cass-Koopman) has used a single commodity model. I think this simplification could be modified. I would divide the economy into two industries--technology-intensive and labor-intensive (euphemism for technology-nonintensive). If Country A is a leading country in technological innovation, while country B is only catching up by borrowing the innovation from Country A. However, there is a time lag because of the patent law. Also, Country B also has to incur some cost in using the technology developed by Country A to adapt it to its own country. With this setting, I wish to show that country A will always have a comparative advantage in the technology-intensive industry, and thus specializing in it, while country B, never specializing in industry B. Thus, even with the possibility of technology leak, the technological progress made by country A will not be shared by country B, because country B does not have the industry with the innovation.
There are two extensions to this model.
Extension 1:
Consider countries specializing in many different industries (exogenously determined). Progress made in basic R&D may benefit one industry exclusively (or a weaker version, disproportionately)--leading to rapid technological innovation in that industry. Thus, countries where that sector is bigger will benefit more, leading to higher growth rate. This might be a case for industrial strategic planning as done by Japan.
This line of thought has implication for interpreting data on growth. When we look for evidence of divergence or convergence based on a couple of decades' data, what we see might be distorted--countries may concentrate in different industries depending on their wealth. For example, what we see as evidence for divergence might merely be result of the industry that rich countries (like banking) focused on had a boom in that time period. A boom in an industry can last for some extended period of time, thus 20 year period data should be interpreted carefully.
Extension 2:
This is complicated and I have not yet thought it out. I might want to add heterogeneity in the quality of final products making it more compatible with new trade theory.
In growth theory, so many models (Ramsey, Solow, Cass-Koopman) has used a single commodity model. I think this simplification could be modified. I would divide the economy into two industries--technology-intensive and labor-intensive (euphemism for technology-nonintensive). If Country A is a leading country in technological innovation, while country B is only catching up by borrowing the innovation from Country A. However, there is a time lag because of the patent law. Also, Country B also has to incur some cost in using the technology developed by Country A to adapt it to its own country. With this setting, I wish to show that country A will always have a comparative advantage in the technology-intensive industry, and thus specializing in it, while country B, never specializing in industry B. Thus, even with the possibility of technology leak, the technological progress made by country A will not be shared by country B, because country B does not have the industry with the innovation.
There are two extensions to this model.
Extension 1:
Consider countries specializing in many different industries (exogenously determined). Progress made in basic R&D may benefit one industry exclusively (or a weaker version, disproportionately)--leading to rapid technological innovation in that industry. Thus, countries where that sector is bigger will benefit more, leading to higher growth rate. This might be a case for industrial strategic planning as done by Japan.
This line of thought has implication for interpreting data on growth. When we look for evidence of divergence or convergence based on a couple of decades' data, what we see might be distorted--countries may concentrate in different industries depending on their wealth. For example, what we see as evidence for divergence might merely be result of the industry that rich countries (like banking) focused on had a boom in that time period. A boom in an industry can last for some extended period of time, thus 20 year period data should be interpreted carefully.
Extension 2:
This is complicated and I have not yet thought it out. I might want to add heterogeneity in the quality of final products making it more compatible with new trade theory.
recent things
This summer is horrible. Actually, not the summer perse but the consequences. After getting back from the summer, I feel so out of shape, and when I went tot the pool, today, I realized that I had to redo the whole torturing process of getting used to swimming. This is not the worst, after all, I can, after some time and with some effort, recover that strength. But the worst is in economics. This week, I have not generated any worthwhile ideas after reading the paper. I hve some random ideas from time to time, but the quality is significantly lower than last term's and when I actually present them, I do not feel convinced and lost interest even in presenting them in the first place. What has gone wrong with me?
I hope this is just a transitory thing, and no hysteresis is present. Hmm, maybe after reading some more thought provoking paper, I would be able to get some new ideas. Oh, yes, Love asked me to build a file on my ideas. Should do that when I have time, but it is kind of ironical for me, since I have little of value of keep track of right now.
I hope this is just a transitory thing, and no hysteresis is present. Hmm, maybe after reading some more thought provoking paper, I would be able to get some new ideas. Oh, yes, Love asked me to build a file on my ideas. Should do that when I have time, but it is kind of ironical for me, since I have little of value of keep track of right now.
Subscribe to:
Posts (Atom)