Crawford, Ian A.; (1998) Cost of living indices. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text out.pdf Download (6MB) | Preview

Abstract

The correct measurement of changes in the cost of living is important for many reasons. Indeed, the many uses to which measures of inflation are put make it difficult to overstate the importance of obtaining an accurate value. For example, the rate of inflation is important in wage settlements, the indexation of social security benefits, estimates of macroeconomic growth rates and, generally, all economic analyses which require knowledge of real, rather than nominal, magnitudes. Unfortunately, whilst a true cost of living index (which is defined as the ratio of the minimum cost of reaching a given level of economic welfare under alternative sets of prices) is a fairly straightforward theoretical concept, in practice its validity and the way it should be implemented are much less clear. The overall theme of this thesis is that of trying to address some of the fundamental issues which cost of living indices face without assuming either the existence or the form of the preferences which might underlie an index. Chapter 1 introduces some of the ideas which appear later and describes some of the central issues. Chapter 2 is an empirical investigation into the question of uniqueness: the extent to which different types of households have experienced differential changes in their cost of living in the period 1979 to 1992. It also considers the question of the influences which indirect tax changes over the period may have had, and looks at the effects which the method used to calculate housing costs has on the resulting indices. While Chapter 2 takes the idea of existence for granted, Chapter 3 considers the issue in more detail. The basic question asked is; is the empirical evidence consistent with the existence of a cost of living index? This chapter suggests a method which uses nonparametric Engel curves to simultaneously improve the power of revealed preference tests, and to make the problem amenable to statistical testing. It also looks at a method which, in the event of rejections, can be used to allow for changes in the quality of goods or equivalently changes in preferences for a particular good over the period of the data. Chapter 4 concerns the question of the construction of true cost of living indices. Estimating true indices is computationally difficult and, as described above, the results may conflict with the theory. Chapter 5 is concerned with another source of bias; specifically with the problem of new goods bias in price indices. This is an area in which functional form matters a great deal. A cost of living index needs to include the price of the new good for the period before the one in which it first existed in order to account properly for the welfare effects of its introduction. Functional form matters a great deal here as in order to do this the main method has been to use parametric demand systems to extrapolate the demand curve to the point at which demand becomes zero. The results of such an exercise are likely to be heavily dependent on functional form. This Chapter describes a nonparametric method which uses revealed preference restrictions to place a lower bounds on the virtual price of a new good without the need for functional form assumptions. This is the lowest bound consistent with the (partly testable) maintained hypothesis that the data are, on average, consistent with utility maximisation. (Abstract shortened by UMI.)

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item