Tyler Gannon's point is that we should adopt policies that will encourage people to invest in the "right" form of human capital as quickly as possible, since specializing in the sector in which we enjoy a comparative advantage will lead to the greatest amount of gains from trade. To focus the analysis, let's consider a couple of possible scenarios about human capital "investment":
Scenario 1: human capital investment is not costly (but also not immediate).
Here, compensating the losers makes no difference on the incentive to switch sectors. Even the most modest wage differential between the two sectors will leave an incentive for workers to gravitate towards higher-wage professions.
Here, compensating the losers makes no difference on the incentive to switch sectors. Even the most modest wage differential between the two sectors will leave an incentive for workers to gravitate towards higher-wage professions.
Scenario 2: human capital investment is costly (and also not immediate).
Here, not only does it take time to learn about higher-paying jobs in other sectors and find them, but it also costs a certain amount of your income to learn the new trade. Thus, the incentive to switch sectors will be motivated by: (1) the wage differential (net of taxes and wage subsidies), (Wc - t) - (Ww + s); and (2) the cost of re-investing in the new skills needed to become proficient in that sector (C), where:
Wc = Wage in the cloth (export) sector;
Ww = Wage in the wine (import) sector;
t = Tax on each cloth worker (= 0 for a shift to free trade with no transfer);
s = Subsidy to each wine worker (ditto).
Assume that t < Wc - Ww, which would be the case if we raised wine workers' incomes just to the level needed to attain autarky utility. This would leave all of the net gains to be enjoyed by the cloth workers. (Also note that in a static model we also require a government budget-balance condition which implies that t*Lc = s*Lw.)
In this case, wine makers stuck in a cruel transition mechanism (say, because Tyler is their "benevolent dictator") will invest in retraining as long as the benefit is at least as much as the investment cost, i.e. if Wc - Ww ≥ C; wine makers living it up in a redistributionist nannystate will only retrain if (Wc - t) - (Ww + s) ≥ C. Note that the second condition requires the cost of reinvesting to be lower than would be the case in the transfer-less transition to free trade, which means current workers are less likely to reinvest, and the transition to full specialization will be slower. Future generations, however will begin their "careers" by investing in the "right" types of human capital (and surely will also hope and pray that the economy never ever changes again!). As the old fogies who foolishly misinvested in wine-making at the start of their lives die off, we reach the free trade equilibrium with full gains from trade.
But, of course there are problems with even this analysis. What if people can't afford the cost of retraining? Even if the benefits justify the cost, workers might not be able to reinvest because the lower wage now received in the wine sector is barely enough to cover a subsistence level of consumption (call this minimum subsistence expenditure "E"), and will not leave sufficient additional funds to cover the cost. One might argue that credit markets could solve this by allowing you to take out a loan against your future earnings, but credit markets often fail when information is imperfect and lenders may not be willing to give loans. In this case, without the ability to take out a loan, investment can only occur in the cruel transferless world if Wc - Ww ≥ C (incentive must be sufficient), and Ww - E ≥ C (the cost must be feasible). One can imagine a set of parameters where (Wc - t) - (Ww + s) ≥ C and Ww + s - E ≥ C, but Ww - E < C. In other words, there will be sufficient incentive to reinvestment under BOTH policy regimes (transfer and no transfer), but the cost of re-investment is only affordable with the transfer. Thus it is possible that the transfer might actually speed up the rate of convergence to the desired free-trade level of specialization and the hands-off approach will not only slow down this convergence, but will even worsen and perpetuate inequality. Whether this is indeed the case is an empirical question, so make sure you enroll in my econometrics course in the spring!
Also, what if the cost of acquiring cloth-making skills increases (which we might expect since it will be the cloth makers - who earn higher wages using their time to produce cloth - who will be needed to train all of these new cloth makers)? If this is the case then one might imagine how, over time, this might lead to a widening of the gap between "winners" and "losers" as the children of wine workers (who themselves could not afford to invest in wine retraining) now also cannot afford the (rising) cost of training their children to achieve the goal of attaining a better career in the cloth sector.
Here, not only does it take time to learn about higher-paying jobs in other sectors and find them, but it also costs a certain amount of your income to learn the new trade. Thus, the incentive to switch sectors will be motivated by: (1) the wage differential (net of taxes and wage subsidies), (Wc - t) - (Ww + s); and (2) the cost of re-investing in the new skills needed to become proficient in that sector (C), where:
Wc = Wage in the cloth (export) sector;
Ww = Wage in the wine (import) sector;
t = Tax on each cloth worker (= 0 for a shift to free trade with no transfer);
s = Subsidy to each wine worker (ditto).
Assume that t < Wc - Ww, which would be the case if we raised wine workers' incomes just to the level needed to attain autarky utility. This would leave all of the net gains to be enjoyed by the cloth workers. (Also note that in a static model we also require a government budget-balance condition which implies that t*Lc = s*Lw.)
In this case, wine makers stuck in a cruel transition mechanism (say, because Tyler is their "benevolent dictator") will invest in retraining as long as the benefit is at least as much as the investment cost, i.e. if Wc - Ww ≥ C; wine makers living it up in a redistributionist nannystate will only retrain if (Wc - t) - (Ww + s) ≥ C. Note that the second condition requires the cost of reinvesting to be lower than would be the case in the transfer-less transition to free trade, which means current workers are less likely to reinvest, and the transition to full specialization will be slower. Future generations, however will begin their "careers" by investing in the "right" types of human capital (and surely will also hope and pray that the economy never ever changes again!). As the old fogies who foolishly misinvested in wine-making at the start of their lives die off, we reach the free trade equilibrium with full gains from trade.
But, of course there are problems with even this analysis. What if people can't afford the cost of retraining? Even if the benefits justify the cost, workers might not be able to reinvest because the lower wage now received in the wine sector is barely enough to cover a subsistence level of consumption (call this minimum subsistence expenditure "E"), and will not leave sufficient additional funds to cover the cost. One might argue that credit markets could solve this by allowing you to take out a loan against your future earnings, but credit markets often fail when information is imperfect and lenders may not be willing to give loans. In this case, without the ability to take out a loan, investment can only occur in the cruel transferless world if Wc - Ww ≥ C (incentive must be sufficient), and Ww - E ≥ C (the cost must be feasible). One can imagine a set of parameters where (Wc - t) - (Ww + s) ≥ C and Ww + s - E ≥ C, but Ww - E < C. In other words, there will be sufficient incentive to reinvestment under BOTH policy regimes (transfer and no transfer), but the cost of re-investment is only affordable with the transfer. Thus it is possible that the transfer might actually speed up the rate of convergence to the desired free-trade level of specialization and the hands-off approach will not only slow down this convergence, but will even worsen and perpetuate inequality. Whether this is indeed the case is an empirical question, so make sure you enroll in my econometrics course in the spring!
Also, what if the cost of acquiring cloth-making skills increases (which we might expect since it will be the cloth makers - who earn higher wages using their time to produce cloth - who will be needed to train all of these new cloth makers)? If this is the case then one might imagine how, over time, this might lead to a widening of the gap between "winners" and "losers" as the children of wine workers (who themselves could not afford to invest in wine retraining) now also cannot afford the (rising) cost of training their children to achieve the goal of attaining a better career in the cloth sector.
In response to this post i'd like to first note that I will not be attempting to disprove your math with math of my own, we all know how that would turn out.
ReplyDeleteWith that being said I understand your point that under Gannon's scenario it would take longer to reach the desired free trade level of specialization, and under transfer conditions Wine makers will be able to more quickly specialize. So while i don't disagree with your theory as to which scenario will lead to specialization the quickest, i do raise the question of whether the quickest method is the best method for the long run. I raise this question because i look at what subsidizing training and education in our country is doing. I'm assuming the subsidy to wine makers in your scenario is comparable to our(U.S) subsidization of education/training.
Why do i bring this up? It is my thought that a big reason why education is so expensive today comes from the fact that the government subsidizes a large amount of education. What happens from these subsidies, first of all there is higher demand because people can afford to go to school (higher demand higher prices). But more importantly, what happens where there are government dollars, prices go up. I use subsidies on corn for ethanol as an example, corn prices are higher due to these subsidies. Subsidizing the education may have brought about quicker specialization, but in the long run, like now for example, it would be impossible for a student to earn a degree at a university and be able to work to pay for it as you go along. Accumulating so much federal student debt from inflated education prices essentially taxes us young students for years to come before we even get in the work force. Whereas in Tyler's scenario (i'm not advocating a "benevolent dictator") the absence of subsidies allows future attempts to specialize be less pricey. Although in Tyler's scenario the existing wine makers would be screwed, but those in the future would have less education costs when they decide to specialize. Also in Tyler's scenario the wine maker may never be able to afford to specialize because their wage minus their minimum subsistence expenditure doesn't allow for it, but compare that to someone who has accumulated so much student debt they will be paying most of their life. Economically speaking isn't the person in Tyler's scenario technically better off having a small amount of positive dollars than the person with student debt.
Let me clarify what my post claims. First, to review, the only way to make free trade "efficient" in the Pareto sense is to have the transfer. Second, the post merely claims that it is "possible" for a transfer to speed up the transition - which policy does in fact lead to a faster transition is an empirical question.
ReplyDeleteThat said, your point about whether a fast transition is desirable is excellent. In my view, this may be more relevant to the question of how trade liberalization should be sequenced: Should we quit tariffs "cold turkey" or phase them out gradually? In a former life, I dabbled a bit in the literature on privatization and the transition of Soviet economies to Capitalism. There is an interesting article by Mathias Dewatripont and Gerard Roland (1992), "The Virtues of Gradualism and Legitimacy in the Transition to a Market Economy" which you can look up on JStor if you like. Perhaps a motivated student might pick the issue of the pace of liberalization up as a future blog topic.
In the case of the subsidy, I have only considered a cash subsidy. Think of it like trade adjustment assistance or the earned income tax credit if you like. Adversely-affected households may choose to spend the subsidy as they please. Some households' preferences (and especially their discount rates) might lead them to spend it all on gas station hot dogs and cheap beer; others may spend it on education. But supposing the transfer were "tied" to being used on education, the sticker price of education may rise as you say. However, the increase in sticker price would be by less than the amount of the subsidy, leaving households paying less out-of-pocket.
As an aside, looking at sticker prices for education to make the claim that costs to students have gone up by the same margin is that the discount rates and rates of financial aid have risen proportionally more, leaving the cost net of aid growing at a somewhat more modest pace. Also, the fact that students are amassing debt may actually be a sign that markets are working quite well! The gap between the wages of secondary-educated workers and those of college-educated workers has grown at an even faster pace than tuition, making the return (even the NET return) on college higher than ever. All the better reason to take out the loan and pay it back at $300/month for 20 years rather than need to save the cost of the education for 20 years and only be able to afford the education when you are nearing retirement (at which time the benefit will be mostly eroded)!
I understand what you are saying about switching sectors and that it might be costly to re-train and switch sectors, and I would like to answer your question along with making another point. Future generations will get a chance to switch sectors and train. Whether it is immediately or two or three generations down line, families will eventually accumulate enough wealth and/or capital (through inheritances) to sell and then receive education if the opportunity cost of clothe making is that much higher than making wine. Second I really believe that a benevolent dictatorship can be extremely beneficial for an economy if the dictator puts his own interest aside. There are plenty of real world examples, but one that really stands out to me is Lee Kuan Yew of Singapore. I believe that the evidence shows that a dictator can act in almost pure altruism and have a flourishing economy while still providing social justice to his/her citizens. Under Yew there was a 6.5 increase in GDP per capita. The rents required where not significantly different from any democratic state, and you didn’t have to deal with all the corruption and logrolling. Although I do not agree with everything Yew did while in power, I think that a lot can be said about the success of his reign, and show a benevolent dictator, like myself, might make a country better off.
ReplyDelete