| Summary: | In micro-economic theory, individual labour supply functions are derived from utility functions. Economic agents maximise their utility under rime and budget constraints. The maximisation conditions indicate the optimal choice between the level of consumption and labour supply. Taking into account the rime constraint produces two-dimensional supply functions representing the participation decision and the quantity of labour supplied conditional on participation. Beginning with a simple static model of individual labour supply behaviour the theoretical part of this study presents several extensions which account for the intertemporal aspects and the family context of labour supply decisions as well as the impact of discontinuous budget constraints. These specifications are then used to derive corresponding econometric models which are tested using Finnish cross-section data from three different sources. In the empirical applications only observations on married women are sampled. A sample of married women is particularly interesting for the study of labour supply behavior since it requires that the participation dimension of the labour supply decisions is fully taken into account. While the subsample of participants is relevant in analysing the quantity of supply, it must be seen as auto-selected. Now the elimination of observations on the basis of an endogenous decision criterion will produce biased estimates of supply parameters if such potential bias is not corrected by econometric methods. The sample selection thus also affects the specification of the econometric model. The specification of econometric models and choice of estimation techniques are based on a careful data analysis aimed at locating potential biases caused by sample selection and data defects as well as endogeneity and simultaneity problems. Several limited dependent variable methods are used in estimations to take account of censoring and truncation (which arise for reasons stated above), simultaneity of endogeneous variables and endogeneity of sample selection criteria. In addition to the usual logit, probit and tobit methods more sophisticated versions of these are also used alike with two and three stage estimation and simultaneous and instrumental variable techniques. Static labour supply models are dynamised under several anticipation hypotheses. The econometric specifications of these intertemporal supply functions are derived and estimated using cross-section data. In general, the estimation of multiperiodic labour supply functions requires panel-type data. The study applies statistical tests to evaluate estimated supply functions and compares different supply models derived from various theoretical starting points. These results are then used in the further development of the theoretical and econometric analysis. The final section derives a system of simultaneous equations which although requiring the specification and estimation of both supply and demand functions also provides an elegant way of dealing explicitly with the problem of involuntary unemployment even at microlevel.
|
|---|