Theoretical Economics Quantitative Economics

 Aims and Scope Editorial Board Journal News Annual Reports Browse the journal Search the journal Forthcoming Papers Awards Supplemental Material Information for Authors Library Subscription Personal Membership Permissions Ordering Back Issues Advertising

 Aims and Scope Annual Reports Editors Monographs in Print Submissions Monographs Online

March 1980 - Volume 48 Issue 2 Page 315 - 332

p.315

On the Disaggregation of Excess Demand Functions

J. D. Geanakoplos
H. M. Polemarchakis

### Abstract

We solve the problem of the restrictions imposed on the Jacobian A at prices $\bar{p}$ of the aggregate excess demand function $x(p)$ of m agents in an exchange economy with $l$ commodities, under the assumption of individual rationality. Given an arbitrary differentiable function $x(p)$ satisfying homogeneity and Walras' law, we attribute rational individual excess demand functions $x^{1} (p), \ldots, x^{m} (p)$ to the $m$ agents such that at any arbitrarily specified vector $\bar{p}$ aggregate excess demand is equal to $x(\bar{p})$ and the following condition is satisfied: There exists a subspace $M$ of dimension $m$ such that the Jacobian at $\bar{p}$ of $x(p)$ and the Jacobian at $\bar{p}$ of the aggregate excess demand function define the same linear function on $M$. If $x(\bar{p}) \not={0}, M$ can be taken to have dimension ($m$+1). As an immediate consequence of our proof for $m$=1 we show that even if $\bar{p}, x(\bar{p}),$ and $Dx(\bar{p})$ are known for the excess demand function of a single agent, the substitution effect and the income effect cannot be unambiguously determined without knowledge of the utility function. We extend the results proved at a point to large open neighborhoods. We show that if $x(p)$ is an arbitrary function which bounded from below and satisfies homogeneity and Walras' law, and if $x(\bar{p}) /not={0}$, then we can find an open neighborhood $G$ of $\bar{p}$ and ($l$-1) individually rational excess demand functions $x^{1}(p), \ldots, x^{l-l} (p)$, such that $\Sigma_{k=1}^{l-1} x^{k} (p) = x(p)$ everywhere on $G$.

Note: to view the fulltext of the article, please login first and then click the "full content" button. If you are based at a subscribing Institution or Library or if you have a separate access to JSTOR/Wiley Online Library please click on the "Institutional access" button.
Prev | All Articles | Next