This paper presents a search equilibrium model of aggregate demand management. The model shows that in a frictionless market, the externality from increased production would be pecuniary and have no efficiency implications. However, in a model with trading frictions, this externality matters. The model has multiple equilibria, giving a second reason for potential welfare improvement from government intervention. The model is based on a highly artificial production and trade process. All individuals are assumed to be alike. Instantaneous utility is given by U = y - c, where y is the consumption of output and c is the cost of production. Lifetime utility is the present discounted value of instantaneous utility. Production opportunities are a Poisson process. With arrival rate a, each individual learns of production opportunities. Each opportunity has y units of output and costs c units to produce. The model shows that the steady state rational expectations equilibrium is inefficient. The inefficiency comes from the fact that a greater level of production improves the trading opportunities of others. The unemployment rate rises from each completed transaction, as a previously employed person becomes eligible to undertake a production opportunity and falls whenever a production opportunity is undertaken. Assuming that all production opportunities with costs below c* are undertaken, we have the equation for the steady state unemployment rate. The level of aggregate demand affects investment decisions since the probability of a sale decreases with the unemployment rate. The paper also examines the effects of long run stimulation policy and short run stabilization policy. It shows that the government can control production decisions and induce private decisions at the optimal steady state. The paper concludes that the model is very special and cannot be used to draw policy conclusions directly. The model is constructed to provide a basis for further generalization and study, and to contrast with models that assume the existence of a frictionless, instantaneous trade coordination mechanism. The existence of this simple model indicates the inappropriateness of basing policy on a model with perfect markets.This paper presents a search equilibrium model of aggregate demand management. The model shows that in a frictionless market, the externality from increased production would be pecuniary and have no efficiency implications. However, in a model with trading frictions, this externality matters. The model has multiple equilibria, giving a second reason for potential welfare improvement from government intervention. The model is based on a highly artificial production and trade process. All individuals are assumed to be alike. Instantaneous utility is given by U = y - c, where y is the consumption of output and c is the cost of production. Lifetime utility is the present discounted value of instantaneous utility. Production opportunities are a Poisson process. With arrival rate a, each individual learns of production opportunities. Each opportunity has y units of output and costs c units to produce. The model shows that the steady state rational expectations equilibrium is inefficient. The inefficiency comes from the fact that a greater level of production improves the trading opportunities of others. The unemployment rate rises from each completed transaction, as a previously employed person becomes eligible to undertake a production opportunity and falls whenever a production opportunity is undertaken. Assuming that all production opportunities with costs below c* are undertaken, we have the equation for the steady state unemployment rate. The level of aggregate demand affects investment decisions since the probability of a sale decreases with the unemployment rate. The paper also examines the effects of long run stimulation policy and short run stabilization policy. It shows that the government can control production decisions and induce private decisions at the optimal steady state. The paper concludes that the model is very special and cannot be used to draw policy conclusions directly. The model is constructed to provide a basis for further generalization and study, and to contrast with models that assume the existence of a frictionless, instantaneous trade coordination mechanism. The existence of this simple model indicates the inappropriateness of basing policy on a model with perfect markets.