Base independence of equivalence scales and nonlinearity of Engel curves

• Applicant: Dr. Christian Dudel
• Subject Area: Statistics and Econometrics
• Term: from 2015 to 2017
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 277165179

Equivalence scales are routinely applied in research on poverty and inequality and are used to guide policy decisions. A large number of different methods for empirical determination of equivalence scales has been proposed. The greater part of these methods makes use of household demand data and so called demand systems. Research conducted in Germany so far mostly used older methods due to data limitations. These methods imply one or both of two restrictive assumptions. First, Engel curves are assumed to be linear. Second, the so called base independence assumption is supposed to hold which requires equivalence scales to be independent of income and welfare level. The project aims to examine both assumptions using German data on household demand. Three methods are applied to reach this goal. First, the Quadratic Almost Ideal Demand System is estimated. Second, an improved version of a semiparametric test of base independence is used. Finally, a new nonparametric test is devised and applied which requires less restrictive assumptions than the other tests. Furthermore, the analyses also yield equivalence scales based on most recent data.

• DFG Programme: Research Grants
• Co-Investigators: Professorin Dr. Notburga Ott; Professor Dr. Martin Werding