The purpose of this paper is to illustrate some of the dangers inherent in use of statistical tests as a criterion for deleting variables from regression models. The deletion of variables from regression models based on t or F tests of regression coefficients has been a procedure widely followed by applied economists and other researchers. When economic theory does not provide an adequate conceptual basis for rigorous a priori specification of the regression model, one approach to model specification has been to include in the regression equation all variables thought to be “somehow” related to the dependent variable of interest. Subsets of variables with statistically significant coefficients are identified, with the aid of a stepwise regression routine. Truncated models consisting of only those variables with statistically significant regression coefficients are sometimes presented in the published research without reference to the initial data dredging that took place.

A number of economic policies are believed to influence an individual's decision of how many hours or weeks to work. Among these policies are welfare programs, income maintenance plans, and unemployment insurance. To date, questions of agricultural labor response to economic incentives have been analyzed by resorting to aggregate data and models, typically utilizing state or U.S. time series data. While this does provide needed information for analysis of some policies, aggregate data and models are deficient in isolating substitution and income effects. These are necessary for analysis of particular programs affecting only income or affecting the individual's budget constraint in a discontinuous way. In particular, aggregate models cannot approach the question of a backward bending supply curve, since aggregate data include not only variations in duration of employment but also variations in labor force participation.

The USDA's Food Stamp Program (FSP) is a major item in the department's budget. In effect from 1939 to 1943 and revived as a pilot program in 1961, FSP has grown until, in 1973, it provided nearly $4 billion in food stamps to an average of 12 million persons per month. About 55 percent of the $4 billion is federal subsidy. The program is continuing to expand as a result of a congressional mandate that FSP be in effect nationwide after June 30, 1974. Because of the FSP's growth, questions are now being asked about the program's impact on demand for food in the United States.

In its pre-World War II inception, FSP was developed as an alternative to direct distribution of commodities to relief families. Although the objective of improving food consumption among needy households was recognized, FSP was viewed primarily as a method for stimulating demand for farm products.

There has been some perplexity among economists over the failure of interregional wage differentials to approach zero over time in an economy characterized by labor mobility. Johnson [7], and Sjaastad [14], among others, have hypothesized declining wage differentials among regions and have puzzled over contrary empirical results. It has generally been assumed that labor will flow toward regions paying the highest wage rate. This equilibrating framework has dominated economic thought on this problem area and has directed research along narrow market-oriented lines.

The hypothesis of this paper is that nonmarket considerations, specifically psychic costs, are a major force in preventing a market-directed flow of human resources. Moreover, “nonoptimal” allocation of human resources results from differences in workers' perceptions of utility between various regions.

Teaching rural development must reflect the breadth and complexity of the real world situation. Time limitations may restrict the teacher primarily to presentation and discussion of the field's many facets, leaving little time to consider how these diverse elements interact. Even if careful course planning and strict adherence to a course outlined provide classroom time for such consideration, only advanced graduate students are likely to have sufficient expertise to readily understand the relationships, real or hypothesized, resulting from these interactions.

Alternatives in the classroom include giving these relationships only cursory, descriptive treatment, or digging into them with analytical fervor, thereby causing many students a great deal of chagrin. A third alternative is demonstrating interactive aspects of rural development with a game. Such a game is discussed in this paper.

Two major programs — allotment-quota and price support — have been in force to support peanut farming since 1952. The purpose is to restrict production of peanuts while supporting the price of peanuts produced. Production is restricted through the allotment-quota program, determined by the Secretary of Agriculture, and converted to a national acreage allotment. However, minimum national acreage allotment has been fixed at 1.61 million acres. The price support program provided that price be supported no lower than 75 percent of the parity price of peanuts. Under the programs, peanuts produced by the grower are sold on the market at the support price for edible uses and excess quantities are sold to the government at the support price. Peanuts purchased by the government are stored and later sold for crushing at the going market price. Because peanuts for crushing command a much lower price than the government acquisition price, the purchase-and-resale operation results in a net loss representing public cost of the peanut price support program.