CONCEPT OF SMALL AREA 
   ESTIMATION
   Small area estimation methods are obviously used in 
     the situations, where there is a need to “borrow 
     strength” to determine the estimation using sample 
     survey, but the sample of considered subpopulation 
     isn’t large enough, what cause too large estimation 
     error. Here “small area” can be understood as smaller 
     administrative units (for example counties – in Polish 
     poviats) or specific groups extracted from the 
     population (for example specific socio-economic 
     groups). This problem can concern also mini-domains 
     or rare features, which are observed with smaller 
     frequency and because of this, the estimates of such 
     variables may cause difficulties even for larger 
     administrative units (for example regions).
     SOURCES OF KNOWLEDGE 
     RELATED TO THE SMALL AREA 
     ESTIMATION
     The small area estimation methodology is systematically 
      developed since 1980’s. Here we can mention books from 
     J.N.K Rao (2003) 
     N.T.Longford (2005) 
     Mukhopadhyay (1998)
     In Polish literature you can also find some examples of more 
      comprehensive studies of this topic. Here we can point out 
      works by 
     Bracha, Lednicki and Wieczorkowski (2003, 2004), 
     Domański and Pruska (2001), 
     Gołata (2004), 
     Dehnel (2003) 
     Żądło (2008).
   INDIRECT SMALL AREA 
   ESTIMATION TECHNIQUES
   Synthetic estimation (ratio and regression)
   The synthetic estimator is applied to the specific 
    domain/group it is assumed that the structure in 
    the larger domain/group is similar to 
    domain/group of interest. It is, however, biased
   Composite estimation
   To minimize the synthetic estimator bias one can 
    use the composite estimation technique where 
    the weighted average from the direct and 
    synthetic estimator is used
    MODEL BASED ESTIMATION 
    TECHNIQUES
     Empirical Best Linear Unbiased Predictor (EBLUP)
     Here mixed models theory is used involving fixed and random effects 
      and small area parameters can be expressed as linear combination of 
      these effects
     Empirical Bayes (EB) estimation
     In the EB approach the posterior distribution of the parameters of 
      interest given the data is first obtained, assuming that the model 
      parameters are known. The model parameters are estimated from the 
      marginal distribution of the data and inferences are then based on the 
      estimated posterior distribution
     Hierarchical Bayes (HB) estimation
     In the HB approach prior distribution of the model parameters is 
      specified and the posterior distribution of the parameters of interest is 
      than obtained . Inferences are based on the posterior distribution. 
      Here, for example, value of considered parameter is obtained using 
      posterior mean and its precision is obtained from the posterior variance
   APPLICATION OF SMALL AREA 
   ESTIMATION TECHNIQUES 
   TO POLISH HOUSEHOLD BUDGET SURVEY
   Models were obtained for regions (voivodships) 
     and counties (poviats). 
   For regions various income and expenditure 
     models was prepared. For example model for 
     available income was obtained using regional 
     accounts data
   For counties also models for income variables was 
     prepared. Here Polish Tax Register (POLTAX) for 
     obtaining the auxiliary data was used