Usin Macro-data to analyze micro effects. Checking the Representativeness of a Sample.

<p style="text-align: justify;">The Graduate work investigates how macro data that can be viewed as aggregates of micro data can be used with combination of survey data to improve estimates of microeconometrics models.</p><p style="text-align: justify;">The object of the research are methods of unknown parameters estimation that can help us to combine in microeconometric model the information from two different sources: macro- and microeconomic data. The subject of the study is to analyze the statistical properties of the estimated coefficients for binary choice model with integrated macroeconomic information.</p><p style="text-align: justify;">The aim is to use practical approaches to estimate coefficients of the choice model that models the probability of being employed conditionally on different individuals&rsquo; characteristics. The methodological basis for the implementation of this goal is the generalized method of moments. We use the statistical package Eviews 8 and the programming environment Visual Basic for Application to estimate the coefficients of the model.</p><p style="text-align: justify;">The micro data is represented by RLMS sample: &laquo;The Russia Longitudinal Monitoring Survey - HSE&raquo; (RLMS-HSE), conducted at the end of 2012 year. The macroeconomic information is presented by the data from the Federal State Statistics Service (GKS). We conduct the statistical tests to verify whether our macro and micro data are compatible. We compare the RLMS sample conditional means with the values ​​provided by GKS to check the representativeness of micro sample.</p><p style="text-align: justify;">The study leads to the conclusion that macroeconomic information included in the binary choice model, can improve statistical properties of the coefficients estimates. The direction for further research is to investigate the estimation approaches for binary choice model with integrated macro information without the assumption that the macro information represents the true value of the population parameters.</p>