Ottaviano G.I.P., Tabuchi T., Thisse J.-F. Agglomeration and trade revisited. Y.Martemyanov HSE CMSSE The First CMSSE Summer School Nizhny Novgorod 2012 Introduction

Even though several modeling strategies are available to study the emergence of economic agglomerations (Fujita and Thisse, 1996), their potential has not been really explored, as recognized by Krugman (1998) himself: To date, the new economic geography has depended heavily on the tricks summarized in Fujita et al. (1999) with the slogan Dixit-tiglitz, icebergs, evolution, and the computer Motivation

Presenting a model of agglomeration and trade that, while displaying the main features of the core-periphery model by Krugman (1991b), differs under several major respects: a) preferences are not CES but the quadratic

utility model and a broader concept of equilibrium than the one in Dixit and Stiglitz (1977); Motivation b) trade costs absorb resources that are

different from the transported good itself. To derive analytically the results obtained by Krugman (1991b). To study forward-looking location decisions and to determine the exact domain in which expectations matter for agglomeration to arise. Motivation

To establish a bridge between the new economic geography and urban economics. When the manufactured goods' trade costs decrease, the economy now displays a scheme given by dispersion, agglomeration, and redispersion (Alonso, 1980). Plan

The model; Short-run price eqilibria (the equilibrium prices and wages that are determined for any given distribution of firms and workers); When do we observe agglomeration? (The process of agglomeration that analyzed by using the standard myopic approach in selecting the stable equilibria);

Optimality versus eqilibrium (comparing the optimum and market outcomes); Plan The impact of workers' expectations on the agglomeration process (introducing forwardlooking behavior and using of the model to

compare history (in the sense of initial endowments) and expectations in the emergence of an agglomeration); The impact of urban costs (associated with the formation of an agglomeration); Conclusion. The model 2 regions H, F.

2 factors A, L. Factor A is evenly distributed across regions and is spatially immobile. Factor L is mobile between the two regions, - the share of this factor located in

region H. The model Some inputs are nontradeable (such as land), some others have a very low spatial mobility (such as low-skilled workers).

2 goods:1st good is homogenous, 2nd one is differentiated product. Factor A: constant returns to scale and perfect competition; freely traded between regions and is chosen as the numeraire. The model

Factor L: increasing returns to scale and imperfect competition. A continuum N of potential firms. There are increasing returns to scale and no scope economies, so each firm produces only one variety. Each firm is negligible and interaction between any two firms is zero. faces a downwardsloping demand.

The model Aggregate market conditions of some kind (here average price across firms) affect any single firm. Trade costs are for each unit transported from one region to the other.

The model Preferences are identical across individuals and described by a quasi-linear utility with a quadratic subutility that is supposed to be symmetric in all varieties, The model

U is maximized at x=N where variety consumption is maximal. The model There is used a quasi-linear utility that abstracts from general equilibrium income effects for analytical convenience. Although this modeling strategy gives the framework a fairly strong partial equilibrium flavor, it does not remove the interaction between product

and labor markets, thus allowing us to develop a fullfledged model of agglomeration formation, independently of the relative size of the manufacturing sector. The model Any individual is endowed with 1 unit of labor (A or L)

and Budget constraint where y is the individual's labor income, p(i) is the price of variety i, and the price of the agricultural good is normalized to one. The initial endowment qo is supposed to be sufficiently large for the equilibrium consumption of the numeraire to be positive for each individual. The model

Solving the budget constraint for the numeraire consumption, and solving the first-order conditions with respect to q(i) yields The model Increasing the degree of product differentiation among

a given set of varieties amounts to decreasing c. However, assuming that all prices are identical and equal to p, we see that the aggregate demand for the differentiated product equals aN - bpN, which is independent of c. It is possible to decrease (increase) c through a decrease (increase) in the while keeping the other structural parameters a and b of the demand system unchanged. The model

The indirect utility corresponding to the demand system is as follows: The model

Technology in agriculture requires 1 unit of A to produce 1 unit of output. In equilibrium Technology in manufacturing requires b units of L to produce any amount of a variety. The marginal cost of production of a variety is set equal to zero. is a measure of the degree of increasing returns in the manufacturing sector. The model and

The equilibrium wages are determined by a bidding process between firms for workers, which ends when no firm can earn a strictly positive profit at the equilibrium market prices. All operating profits are absorbed by the wage bills. The model

Demands faced by a representative firm located in H in region H: Profits Short-run price eqilibria

The process of competition between firms for a given spatial distribution of workers. Each firm i in region H maximizes its profit , assuming that its price choice has no impact on the regional price indices The prices selected by the firms located within the same region are identical and given by 2 linear expressions These prices must be consistent

Short-run price eqilibria The equilibrium prices under monopolistic competition depend on the demand and firm distributions between regions. Short-run price eqilibria There is freight absorption since only a fraction of the trade cost is passed on to the consumers. Short-run price eqilibria

Deducting the unit trade cost from the prices set on the distant markets, that firms' prices net of trade costs are positive regardless of the workers' distribution if and only if There must be increasing returns for trade to occur. Short-run price eqilibria

The equilibrium gross profits earned by a firm established in H on each separated market: i.e.the profits earned in H, while the profits made from selling in F are An aggregate local demand effect due to the increase in the local population that may compensate firms for the adverse price effect as well as for the

individual demand effect generated by a wider array of local varieties. Short-run price eqilibria The individual consumer surplus Short-run price eqilibria When do we observe agglomeration?

When do we observe agglomeration? The driving force in the migration process is workers' current utility differential between H and F: when t is time.

A spatial equilibrium implies If is positive, some workers will move from F to H; if it is negative, some will go in the opposite direction. When do we observe agglomeration?

A spatial equilibrium is stable if, for any marginal deviation from the equilibrium, this equation of motion brings the distribution of workers back to the original one. Therefore, the agglomerated configuration is always stable when it is an equilibrium, while the dispersed configuration is stable if and only if the slope of is nonpositive in a neighborhood of this point. When do we observe agglomeration?

The immobility of the farmers is a centrifugal force, at least as long as there is trade between the two regions. The centripetal force finds its origin in a demand effect generated by the preference for variety. The indirect utility differential When do we observe agglomeration?

There is always an equilibrium Since the indirect utility differential has always the same sign as otherwise it has the opposite sign. In particular, when there are no increasing returns in the manufacturing sector the coefficient of

is always negative since so that dispersion is the only (stable) equilibrium. When do we observe agglomeration? It remains to determine when

This is so if and only if where the second inequality holds because Otherwise, the coefficient of for all is lower than

is always positive When do we observe agglomeration? When do we observe agglomeration? The best way to convey the economic intuition behind Proposition 1 is probably to make use of a graphical

analysis. Figure 1 depicts the aggregate inverse demand in region H for a typical local firm after choosing, for simplicity, the units of L so that b+cN= 1: Figure 1 is a powerful learning device to understand the forces at work in the model. When do we observe agglomeration?

The demand effect dominates the competition effect when goods are bad substitutes (c small), increasing returns are intense ( large), the farmers are unimportant (A small), and trade costs are low ( small). The entry of new firms in one region would raise the operating profits of all firms, hence wages. Higher operating profits and wages would attract more firms and workers, thus generating circular causation

among locational decisions. Agglomeration would thenbe sustainable as a spatial equilibrium. When do we observe agglomeration? Since the impact of firms' relocation on consumer surplus is always positive, agglomeration could still arise even when operating profits, hence wages, decrease with the size of the local market, because the demand effect is dominated by the competition effect. Furthermore, the same argument is likely to

hold for most downward-sloping demand functions. Optimality versus equilibrium We assume that the planner is able (i) to assign any number of workers (or, equivalently, of firms) to a specific region and (ii) to use lump-sum transfers fromall workers to pay for the loss firms may incur while pricing at marginal cost. Observe that no distortion arises in the total number of varieties since N is determined by the factor endowment (L)

and technology ( ) in the manufacturing sector and is, therefore, the same at both the equilibrium and optimum outcomes. Optimality versus equilibrium The setting assumes transferable utility, the planner chooses 2 in order to maximize the sum of individual indirect utilities: in which all prices have been set equal to marginal

cost: Optimality versus equilibrium Operating profits are zero, so that firms do not incur any loss. where Optimality versus equilibrium Optimality versus equilibrium

The impact of workers' expectations on the agglomeration process The parameter domains for which there exists an equilibrium path consistent with belief, that workers will eventually agglomerate in the smaller region, assuming that workers have perfect foresight (selffulfilling prophecy). Consider the case in which initially region F is larger

than H. The impact of workers' expectations on the agglomeration process Therefore, we want to test the consistency of the belief that, starting from t=0, all workers will end up being concentrated in H at some future date t=T; that is, there exists T >0 such that, given are the instantaneous utility levels of

a worker currently in regions H and F, respectively, at time t > 0. is instantaneous utility level in region H at The impact of workers' expectations on the agglomeration process The intertemporal utility of a worker who moves from F to H at time The impact of workers' expectations on the

agglomeration process The impact of workers' expectations on the agglomeration process The impact of workers' expectations on the agglomeration process Since in equilibrium a worker moving at t must be indifferen between migrating at that date or at any other date, until the final expected date T, along an

equilibrium path it must be that u(t) = u(T) for all Terminal conditions are The impact of workers' expectations on the agglomeration process The impact of workers' expectations on the agglomeration process The impact of workers' expectations on the agglomeration process

As long as obstacles to trade take intermediate values and regions are not initially too different, the equilibrium is determined by workers' expectations and not by history. The impact of workers' expectations on the agglomeration process As to the remaining comparative static properties of

the overlap, they are explained by the fact that proximity to = 0 increases the time period over which workers bear losses, a large rate of time preference gives more weight to them, and a slow speed of adjustment extends the time period over which workers' well-being is reduced. The impact of urban costs

Space is continuous and one-dimensional. Each region has a spatial extension and involves a linear city whose center is given but with a variable size. The city center stands for a central business district (CBD).

The two CBDs are two remote points of the location space. Interregional trade flows go from one CBD to the other The impact of urban costs

Each agglomeration has a spatial extension that imposes commuting and land costs on the corresponding workers. Workers consume a fixed lot size normalized to unity, while commuting costs are linear in distance, the commuting cost per unit of distance being given by units of the numeraire. The opportunity cost of land is normalized to zero. The impact of urban costs

The equilibrium land rent at distance the H-CBD from The difference in urban costs between H and F

The actual utility differential The impact of urban costs The existence of positive commuting costs within the regional centers is sufficient to yield dispersion when the trade costs are sufficiently low. The economy moves from agglomeration to dispersion when trade costs fall, thus confirming the

numerical results obtained by Helpman (1998). The impact of urban costs Excessive agglomeration arises for intermediate values of the trade costs. When urban costs are positive, the equilibrium may yield either suboptimal agglomeration or suboptimal dispersion, depending on the parameter values of the

economy. Conclusion

Authors proposed a different framework that is able not only to confirm those insights but also to produce new results that could barely be obtained within the standard one. They have used this framework to deal with the following issues: the welfare properties of the core-periphery model the impact of expectations in shaping the economic space the effects of urban costs on the interregional distribution of activities.

Conclusion The main results in the literature do not depend on the specific modeling choices made. The model used in this article still displays some undesirable features that should be remedied in future research. First, there is a fixed mass of firms regardless of the consumer distribution.

Furthermore, by ignoring income effects, our setting has a strong partial equilibrium flavor. Conclusion

Authors proposed a different framework that is able not only to confirm those insights but also to produce new results that could barely be obtained within the standard one. They have used this framework to deal with the following issues: the welfare properties of the core-periphery model the impact of expectations in shaping the economic space the effects of urban costs on the interregional

distribution of activities.