Summary of "Per Capita Income
Putting Back Into Trade Theory," Markusen (2010).
(Lavoix by Roman Régnacq Charles, Bastianini Bruno)
Markusen (2010) takes Linder assumptions to improve the understanding and estimation of trade. Linder (1961), Swedish economist, stressed the important role of per capita income in the international economy. He believes that countries with similar income levels are more likely to trade together. Thus, differences in preferences among nations represent an obstacle to bilateral trade. In this case, the size of the market impacts the production specialization of a country. A country will export goods for which it holds a large domestic market.
His theory was less analyzed by researchers in International Economic models offer. The framework allows standard neoclassical models simpler and more flexible. The classical model Hecksler-Ohlin he considers the distribution of preferences to be identical and homothetic. In his paper, Markusen takes up the Linder hypothesis, it highlights areas of inconsistency in the literature and construct alternative explanations.
By adding non-homothetic preferences in a traditional model (HO 2 X 2 X 2), Markusen was able to explain several phenomena such as:
-
away increasing wages
-
the mystery of "missing trade"
-
through domestic consumption
-
the role of income distribution within a country (all this by looking only aspect of the application).
Furthermore, adding imperfect competition (Cournot oligopoly) can explain the levels of margins and higher prices in a country with high per capita income. The paper concludes with some suggestions of calibration, estimates and equations of gravity.
The model of the article raised six hypotheses (the X is relatively capital-intensive, well there in labor, the host nation is relatively better endowed with capital, the country f in work, production of X is increasing returns, labor supply is equal to the number of households, the country ha improved productivity and income elasticity of X is greater than 1). These assumptions are not really original except that the good X is considered a higher good. The author draws on literature (Bergstrand 1990, Hunter 1991 and Nishioka et al. 2009) and considering the capital intensive good as well as "luxury."
To model the non-homothetic preferences, uses a function Markusen Stone-Geary. This script allows to formalize a consumption threshold, that threshold before the agent consumes only Y and from this level it starts to consume the property X. Indeed, when income increases the budget share allocated to X increases relative to, Y. To study the impact of income per capita, the author Marshallian aggregate demand.
paper first result: a growing economy may have different implications depending on the economy that growth is the result of a productivity growth or the result of an accumulation of factors. productivity growth by increasing income per capita and consumption of X increases more than proportionately to income (income elasticity assumption of X greater than 1) . While a growth factor accumulation keeps constant income per capita and therefore does not alter the basket per capita.
Second result: the author gives an intuitive explanation of wage differentials. If movements of capital and labor are free, workers move in h f (f and capital moves in h). So in this case, h becomes more abundant in capital (increased production of X) at the same time increasing revenues in h. This increase in revenues implies that consumers are more likely to buy the property X. The differences in relative wages are growing.
Third result: the more a country is specialized, the greater the difference between the observed and estimated trade ("missing trade") is important. From the production side, specialization can not grow indefinitely while specialization in consumption can. Thus, homothetic preferences induce missing trade, taking into account the non-dilatation can correct the discrepancy between the observed and estimated trade .
Fourth result: the role of income distribution. If all consumers in one country have an income sufficient to consume X, the distribution of income does not affect aggregate demand. The problem arises if some have no minimum wage in the country and can buy only good Y By observing countries with similar income per capita, aggregate demand for luxury good may be greater in countries with greater inequality, because the demand for the X is increasing faster than income (the author cites the example of Mercedes in Africa).
In imperfect competition (economies of scale, industry X is Cournot competition, free entry free exit of firms ..), when incomes rise in h increases the demand for X giving greater scope for firms localized pm With the assumption of free entry that margin is offset by a larger number of firms. The margin remains constant while h has the tendency to reduce f. Thus, the relative margin increases in capital-intensive countries.
In its final part, the author identifies several paths to follow on sizing, estimating and gravity equations to take into account the non-homothetic preferences and income per capita rehabilitation as an important source of international trade.
Conventional procedures of general equilibrium analysis assumes homothetic preferences. In this method, used by economists (often) a Cobb Douglas standard to express the utility of consumers. They then observe the distribution of spending in the official statistics to calibrate the parameters of the function. In his article, the author guide us to calibrate the model to take into account the non-homothetic preferences. To calibrate z (a good rival and non excludable), he advocates using the sharing of consumption observed in the statistics with a traditional econometric estimation of β. This calibration allows us to distinguish the movements of the Engel curve (graph attached) between AC and AB. The author shows that standard models tend to neglect these effects and to misunderstand the effects of productivity growth.
To estimate the trade, economists resume a pattern HO Classic summarizing the exports to the difference between production and consumption (E = X - C). By integrating the factor content of trade, [A] = E [A] X - [A] C = V - sV (with V - sV trade "expected" or estimated), the results are the most common overestimation of trade ("missing trade"). The author points out that many efforts have been made adjustment of the production side. The vector "consumption" by cons may qualify for a better fit. Assuming non-homothetic preferences, the author rewrites the consumption vector. Under these assumptions, the consignment in which a country is relatively specialized is also a property where that country is specialized in consumption. In doing so, the high consumption of the property on the market reduces the level of exports (down "missing trade", the same reasoning applies to imports and limit the "missing trade").
Finally, the author notes that the equations of gravity tend to play down the impact of distributions of income per capita. If we consider as homothetic preferences, such an approach is not wrong. But when one considers preferences as non-homothetic, the per capita income explain a significant share of trade between two countries. This is especially true if the traded goods are differentiated and have a high income elasticity.
This text puts into perspective the new framework of research. For what we are interested, its contribution to regional and international economy is very important. It allows for extensions on the number of paper relating to the prediction of trade and face the recurring cases of "missing trade".
Lavoix Roman Régnacq Charles Bruno Bastianini
