Analysis of Regional Economic Network Structures Based on Estonian VAT Transaction Data
This paper has illustrated main characteristics and tendencies for Estonian firms’ network and analyzed main factors that influence the emergence and persistence of trading relationship between firms.
One of the important results is that spatial proximity is significant for the formation of partnership among firms. Estonian enterprises tend to interact mostly within its geographic unit and location in the same county significantly increases the chances of between-firm ties. Also, all geographic units are connected to Tallinn and Northern Estonia that have a high volume of operating enterprises and are major connection centers for firms that are more embedded in import and export operations. Tartu, as expected, has its own developed network and has fewer connections to Tallinn than to its own geographic unit. Regional differences in terms of productivity and value chain length are partially explained by sectoral structures of the regions and network characteristics. Also, network characteristics may influence productivity.
The analysis of tie formation and tie persistence shows that geographical distance plays an important role in tie formation process but is less significant when a tie between firms has already been established. Individual firm characteristics are significant for both processes and social aspects (such as attending same network events by employees, previous history of successful interactions, etc.) in tie persistence analysis might be taken into account in the future research. Embeddedness in sub-networks (or communities) also shows positive impact on average performance of enterprises and further initiatives to encourage domestic firms’ cooperation should be considered.
The influence of geographical distance shows that seller-buyer search conditions right now are limited by regional boundaries and cost-minimizing outsourcing strategy is likely to be dependent on spatial proximity of potential trading partners. The encouragement of distant interactions may allow better seller-buyer conditions and more efficient partnerships and joint projects. Another issue is the impact of export orientation on the value chain length and possible spatial shocks that are associated with it. The productivity analysis shows that there is a small negative correlation between the number of foreign buyers and productivity. Possible alternatives of concentrating more high-tech production stages within economy should be considered to avoid falling into “middle income trap”.
Foreign or state-owned companies perform comparatively better than domestic firms and productivity analysis shows that enterprises are more likely to have higher productivity when they have more foreign suppliers. Numerous studies focus on learning effects from interactions with foreign companies but the effects of such cooperation should be viewed at regional scale. Based on previous research and current analysis results, it can be concluded that the effects of geographical location of foreign and domestic companies as well as the efficiency of integration into local networks should be taken into account in policy making.
The scope of the research can be further extended to a more thorough analysis of regional differences and application of additional data sets. Particular issues that emerged during the research are data availability and data cleaning. VAT tax declaration is more reliable source of firms’ interactions than data from questionnaires but the data set is constructed based on raw declarations that are subjects to misreporting and absence of particular data. All such data has been excluded from the analysis in order to avoid any misspecification issues but the introduction of more data provides opportunities for further research. The problem of official and actual addresses has been solved by excluding several industries from the data. This issue can also be solved by comparing employees’ working addresses from TÖR (Töötamise registrisse or working registry) and comparing it to business registry data on firm location. It can allow accounting for larger network that includes all industries and companies that can have several locations within the country with more pronounced regional effects. Another important issue that can be reseached is the performance of domestic and foreign companies. Due to favorable investment and business opportunities there is quite a large fraction of foreign companies in the economy. Presence of a succesfull multinational or foreign enterprise can significantly influence the regional performance and create competitive environment for domestic firms. By using foreign direct investment for the analysis of innovation flows, the influence of foreign enterprise in the market can be assessed. Thus, there are perspectives for further regional analysis.
Author: Irina Martyanova University of Tartu