The paper examines a model of sequential search and information revision in a housing-market context and relates the outcome of search to the sequence of vacancies observed. The model implies the existence of a critical frontier that separates acceptable from nonacceptable vacancies in the price–quality plane characterizing housing vacancies. The searcher's beliefs concerning the distribution of vacancies change as vacancies are viewed, resulting in shifts in the frontier. In the present case, increases in perceived mean market price or price–quality covariance shift the frontier downwards, whereas increases in perceived mean market quality and in price and quality variances shift the frontier upwards. The observation of a single vacancy causes the frontier to shift downwards if and only if the vacancy is contained in a set of vacancies lying below the frontier. The set is characterized by finite upper and lower limits on the quality of the vacancy for any given price, and contains a line of maximally effective vacancies. The greatest downward shifts in the frontier result from the observation of vacancies with relatively extreme prices. Similar results hold if several vacancies are viewed. The revision of beliefs appears to reduce average search time when vacancies are generated at random. By ordering a given set of vacancies in a nonrandom manner, an agent may significantly affect the search time, the price and quality of the vacancy purchased, and the utility level of the vacancy purchased under conditions of information revision. The showing of relatively extreme vacancies may induce shorter search times as well as higher purchase prices than would otherwise hold for a given buyer, although the particular set of vacancies chosen to be shown by an agent, and the sequencing of the vacancies, depends on the goals of the agent.
