ELABORATION OF EDUCATION COMPONENT-BASED MODELS AND ALGORITHMS FOR FORECASTING SOCIO-ECONOMIC DEVELOPMENT UP TO 2050

The paper analyzes growth forecasts made for 10 world's largest economies up to 2050. Forecasts were made with PricewaterhouseCoopers exogenous model and VIK endogenous model developed by the authors with the latter model dominantly based on the education component. Compared are errors of GDP at PPP growth forecasts made in 2006 - 2017 with the well-known PwC model and in 2015 with the VIK model. The difference of 0.7% between total GDPs at PPP of 10 economies in 2050 forecasted with PwC and VIK models testifies to adequacy of the latter. As to some countries, the relative variation of the forecasts reaches the acceptable error threshold being 39-52% for Russia, Brazil and Nigeria. Revealed are significantly different PwC forecasts over the years, in particular the forecast made in 2017 differs from that made in 2015 by 18% due to the reference growth rate exogenous decrease from 2% to 1.5%. Deviation between VIK and PwC forecasts falls within the deviation between PwC forecasts over the years. The identified forecast variations do not change principally the order of leading economies in the long-term period of 33 years though the order of economies of similar GDPs, places from 5 to 9, is hard to determine. The study focuses on elaboration of endogenous models and algorithms for forecasting socio-economic development with detailed educational characteristics of human capital being involved. The results of the study may be applied to determine the level of confidence in long-term socio-economic forecasts and may be useful when developing new simulation methods and algorithms for socio-economic forecast.

человеческий капитал, образование, экономическая динамика, ВВП, моделирование, трудовая деятельность, прогнозирование, эндогенный, погрешность