Estimation of Structural Business Statistics for Small Firms by Using Administrative Data

The need of handling missing response and potential bias in the estimation of business statistics has brought to exploit new possibilities offered by administrative sources in the estimation of Structural Business Statistics (SBS) for small-medium enterprises. The joint use of fiscal agency sources and balance sheet data with survey data, has led to a more integrated approach in the production process. The purpose of this work is to describe the new integrated approach and of measuring the reduction of the sampling error with the new estimation process by distinguishing a source effect from the non response effect. The paper is articulated in the following steps: (i) analysis of the coverage of the survey units through administrative sources by stratification variables, such as the economic activity, the number of persons employed and the legal form; (ii) comparison of meaning and content of the variables from balance sheets and tax data with the corresponding SBS variables; (iii) imputation of unit non-response through administrative data and re-weighting of the final sample to obtain a new estimate for 2007; (iv) evaluation of the discrepancy between the old and the new estimates and evaluation of the non response effect.