Zemgale planning region divides into two intra-regions (Zemgale 1 and Zemgale 2) by natural, soil and other agricultural resources. Each intra-region has different size of farms. The aim of the research was to find the structure of rural economy in different size farms or rural enterprises of Zemgale 1 and Zemgale 2 intra-regions. The research showed that very small farms dominate grow-plant sectors production. Medium size farms and enterprises have solid high proportion of grow-plant sector production in total production value. This proportion is even higher in biggest agricultural enterprises and companies (100 is less than or equal 250 ESU). In general author has stated that in Zemgale 1, which is the biggest agricultural production territory in Latvia, grow-plant production sectors dominate.

The evaluation of EU Member States’ co-founded programmes was assigned particular importance in recent years. The core question to be answered in programme evaluation is whether the stated objectives are accomplished by particular intervention. Until recently, the use of ‘naïve’ estimates was common in the evaluations of EU Member States Rural Development Programmes. The use of these estimates leads to potentially substantial selection bias resulting from using the outcomes of non-participants as a proxy for the possible outcomes of participants in the case of non-participation. The effectiveness of interventions on outcomes of interest can be better evaluated by propensity score matching (PSM). The objective of the study is to determine the net average effects from semi-subsistence farm support measure of the Latvian Rural Development Programme 2007 – 2013. To reach the research objective, propensity scores based on the most important characteristics were calculated for participants and non-participants of the measure and average treatment effects for gross value added were evaluated by matching methods. The research results show that the positive programme effects evaluated by ‘naïve’ estimators are overestimated in comparison with the results obtained by more rigorous PSM method.

The aim of this research is to evaluate the spatial differences of the agricultural economy of Latvia during the First Republic by using spatial autocorrelation calculations. Morans’s I coefficient of spatial autocorrelation is used to describe the spatial relations of the analyzed indicators. The obtained results show that Moran’s I for more than half of the obtained parameters is greater than 0.5, which indicates a rather high spatial autocorrelation. It was observed that variations of the spatial autocorrelation values can be explained by what affects these parameters. Spatial autocorrelation is significantly higher for those parameters whose value depends mainly on the variety of natural factors (land use, specific crop share). Moran’s I value for those parameters whose values depend on the actions by the local government in rural parishes is considerably lower (for example, budget allocation, etc.) A large portion of the examined parameters characterizing the level of parish economic development show the biggest differences when comparing Latgale with the rest of Latvia. In the case of such regional differences, Moran’s I is usually within 0.5–0.7, indicating moderately close spatial autocorrelation. For many indicators regarding agriculture and land use many southern parishes within Jelgavas and Bauskas county show a very considerable difference from the rest of Latvia showing much higher rate of agricultural production.

The aim of this research is to evaluate the spatial differences of the agricultural economy of Latvia during the First Republic by using spatial autocorrelation calculations. Morans’s I coefficient of spatial autocorrelation is used to describe the spatial relations of the analyzed indicators. The obtained results show that Moran’s I for more than half of the obtained parameters is greater than 0.5, which indicates a rather high spatial autocorrelation. It was observed that variations of the spatial autocorrelation values can be explained by what affects these parameters. Spatial autocorrelation is significantly higher for those parameters whose value depends mainly on the variety of natural factors (land use, specific crop share). Moran’s I value for those parameters whose values depend on the actions by the local government in rural parishes is considerably lower (for example, budget allocation, etc.) A large portion of the examined parameters characterizing the level of parish economic development show the biggest differences when comparing Latgale with the rest of Latvia. In the case of such regional differences, Moran’s I is usually within 0.5–0.7, indicating moderately close spatial autocorrelation. For many indicators regarding agriculture and land use many southern parishes within Jelgavas and Bauskas county show a very considerable difference from the rest of Latvia showing much higher rate of agricultural production.