CSEAMR participation in the SEA 2021 conference

Members of our research group successfully participated in XV World Conference of the Spatial Econometrics Association.

The conference was held online via Zoom platform. CSEAMR staff members gave the following presentations:

• O. Demidova, "Partial marginal effects for spatial models on the example of wage curve for Russia" (coauthor E. Timofeeva)
  Аннотация
  For interpreting the results of spatial-econometric model estimation, the average direct and indirect effects for each variable are usually calculated. However, sometimes we need to know how changes in region j affects region i. In this case, it is necessary to calculate the partial marginal effect ∂Yi/∂Xj, to find out how much such partial marginal effects differ from average marginal effects. We answered this question with a specific example of the wage curve estimation. The wage curve usually refers to the negative relationship between wages and unemployment (taking into account control variables). However, for such a large and heterogeneous country like Russia, one cannot speak of a single labor market, but rather a system of local regional markets interacting with each other. Given this fact, we have put forward two main hypotheses. Hypothesis 1. Wages in Russian regions negatively depend on the level of unemployment in the region, and the intensity of this influence is not the same for all regions. Hypothesis 2. Changes in unemployment in one Russian region affect wages not only in this region, but also in other regions, and with different intensities. To test these hypotheses, we estimated the Spatial Durbin Model using panel data for 81 Russian regions, 2005–2018. In order to draw inferences regarding direct and indirect partial marginal effects, we simulated the distribution of spatial direct and indirect effects using the asymptotic normality of ML estimates and a variance-covariance matrix implied by the ML estimates. Both hypotheses received empirical confirmation, and, as in other countries, wages in more remote regions and in predominantly agricultural regions are more sensitive to changes in the unemployment rate. For other variables the partial marginal effects were also often sufficiently different from the average marginal effect. Using partial marginal effects, we found for region j 1) the regions most affected by region j, and 2) the regions which most affect region j. This is important, for example, for assessing the consequences of government programs.
  
• A. Demyanenko, "The impact of public health spending on economic growth in Russia: a regional aspect" (coauthors O. Demidova, E, Kayasheva)
  Abstract
  The COVID-19 pandemic has paralyzed many sectors of the economy and has shown that the health systems of many countries were not prepared for the virus, which required a huge investment of human and material resources to sustain and rebuild the economy and mitigate the negative impact of the disease on the population. This challenge stressed the importance of investing in the promotion of quality of people’s life: in education and science - to develop innovative technologies and their more rapid implementation into mass production, in health - to create favorable living conditions and facilitate the development of the population. This investigates the influence of an increase in government healthcare expenditures on regional economic growth in Russia. Studies have shown that an increase in healthcare expenditures stimulates an increase of GDP through several channels. First, it improves the quality of labor force that can lead to an increase in labor productivity. Secondly, an increase in the productivity and size of the labor force leads to consumption extension and then to firms’ income growth, so there is a multiplication effect. Including the presupposition that a relationship between health expenditure and economic growth may be non-linear we formed the hypothesis of the existence of the average optimal share of health expenditure in GRP that maximizes average regional economic growth rate. In this study, we considered 6 main categories of expenditures of consolidated budgets of Russian regions: the share of expenditures on health care, physical culture and sport in the GRP (this is the main variable of interest), the share in GRP of general public expenditure, national economy expenditure, housing expenditure, education expenditure and social policy expenditure. Naturally, economic growth is influenced by factors other than public expenditure. Investment is one of the key drivers of economic growth (Solow, 1956), that is why the ratio of fixed investment to GRP was also included. A number of studies have shown that there are additional factors that influence economic growth: the level of urbanization (Henderson, 2003; Friedmann, 2006), the openness of the economy (Kamensky, Ivanova, 2011), the diversification of the region’s economy (Essletzbichler, 2007; Shediac, 2008), the quality of human capital (Maddison, 1991; Lutz, Samir, 2011). We have also taken into account investment attractiveness and the aggregate index of banking services in the region. In this research we suggest that an increase in healthcare expenditures, besides the direct effect on economic growth of a particular region, also affects economic growth of neighboring regions. The possible explanation of this is the positive impact of healthcare services received in the neighboring regions on nearby territories, joint national healthcare projects and distribution of scientific knowledge. Using the spatial Durbin model focusing on regional data of 2005-2018, it was shown that the average optimal share of health expenditures is 5,9% of GRP with an inclusion of spatial effects and 6,4% without them, outlining the importance of including interconnection variables between Russian regions in the model. The regional statistical analysis showed the failure to reach the recommended share by most Russian regions, which can be viewed as a possibility for future economic growth stimulation if there is an increase in government spending on healthcare.
  

More information is available at the conference and SEA websites.