Determining the equilibrium effects of minimum wage changes on labor market outcomes is a challenging modeling and estimation problem; arriving at policy recommendations is a task even more daunting. Faced with the inherent difficulties of modeling equilibrium labor market events given the limited amount of data to which researchers have access, much recent research has been performed outside of an explicit behavioral framework, with researchers pursuing the more limited objective of carefully describing the observed effects of recent minimum wage changes using quasi-experimental methods [see Card and Krueger (1995) for a summary of these studies and a comprehensive, critical survey of most of the previous research done in this area]. In our view, these recent studies have been particularly useful in indicating that the “textbook” competitive model of the labor market, which has been used as an interpretive framework for the bulk of empirical work performed using aggregated time series data, may have serious deficiencies in accounting for minimum wage effects on labor market outcomes when confronted with disaggregated information.
While the quasi-experimental results have raised a number of interesting challenges to orthodox theory, few cogent models have been advanced that are consistent with the results that have been found. Some of the explanations for these empirical findings (e.g., lack of significant employment losses, impacts on the wage distribution above the minimum) do not seem to be testable given our current data resources. In fact, it appears difficult to operationalize many of the explanations proposed even for the more modest purpose of empirical implementation.
Experience from around the world indicates that poor credit quality coupled with weak credit management practices continue to be a dominant factor in bank failures and banking crises. Many of the credit losses suffered by banks, thrifts and insurance companies in the United States in the early 1990s have resulted from excessive portfolio concentrations of loans in the real estate industry (residential mortgages, commercial real estate mortgages and commercial real estate loans). More specifically, US banks loaned enormous amounts of money to commercial real estate companies, for the period 1989-1994, based on optimistic projections of rental income growth and increased asset values (Browne & Case, 1992). “When the (real estate) bubble burst, banks had to charge off around $34 billion in real-estate-related loan losses” (Caouette et al.,1998; FDIC, 1997). A similar crisis in the US sub-prime loan market is currently taking place but it is too early to comment on the likely causes and assess the effects and the consequences of this crisis.
European countries such as Switzerland, Sweden and the UK – as well as Japan (Siebert, 2002: 116-119) and East Asia (Hilbers et al., 2001; Collyns & Senhadji, 2002; Quigley, 2001) – experienced similar crises in the 1970s and 1990s. Historical data show that there is a very close relationship between the over-borrowing of the real estate companies, the real estate bubbles and the banking crises (see BCBS April 2004). The interlinkages between banks and real estate companies involve an inherent transfer of credit risk.
There are two competing approaches to statistical inference: classical and Bayesian. The fundamental difference between them is the notion of probability. In the classical framework, the probability of an event is defined by the limit of its relative frequency. Estimators and test procedures are evaluated in repeated samples. In the Bayesian framework, probability is defined by a degree of belief.
The Bayesian approach makes it possible to incorporate a belief about the hypothesis being tested and its alternative in the form of a prior-odds ratio. When we look at the data, we get a posterior-odds ratio, which summarizes all the evidence (prior and sample) in favor of the hypothesis or its alterna-tive. The posterior-odds ratio can be interpreted as the ratio of the probability that the hypothesis is valid to the probability that its alternative is valid.
