Мировая теневая экономика, мнение турецкого бухгалтера.

The estimated sizes of the SE rely on two-sector (formal and shadow economies) dynamic general equilibrium modelling (DGE). To do this, the values of various currently observed variables, such as consumption, investment, formal output (ie GDP), employment, taxes, government expenditures, and the growth rate of GDP are fed into the model. The model then yields an SE size consistent with the ‘ingredients’ of the model. Data series for all these variables are available for all the years up to the year 2011. Once the values for the shadow economy have been determined as a percentage of GDP on a country-by-country and year-by-year basis (up to 2012), forecasting the size of the SE for any country i is done by estimating the following econometric regression equation in a time-series (year-by-year) framework

Here, for country i, in year t, the dependent (left-hand-side) variable, IS, is SE size as a percentage of the official GDP. Several variables have been used on the right-hand side (denoted by X) to take in to account the various determinants of SE. These variables can be grouped under three categories: institutional, economic and demographic. Institutional variables include: political stability, ethnic tensions, law and order, democratic accountability, quality of bureaucracy and corruption control. These variables are intended to represent several dimensions of institutional quality within an economy. Measures for all these variables are available in the International Country Risk Guide (ICRG) of Political Risk Services. ICRG reports estimates for these variables up to 2013 and forecasts of all these variables are constructed up to 2025 using and AR (m) model with m=2. That is, all the variables have been forecasted using their first and second lag, as well as their growth rates in the previous 10 years

Economic variables include: consumption expenditures (as percentage of GDP), investment (as percentage of GDP), real GDP per capita (in thousand USD), unemployment (%), taxes (total tax revenue as percentage of GDP), government expenditure (as percentage of GDP), growth rate of GDP (as a percentage rise per year), internet users per thousand people (proxy for technology). Demographic variables include: Population growth rate (as a percentage rise per year), and percentage of young population in the total population. These two groups of variables have been obtained from the World Development Indicators of the World Bank up to 2012. Then, as with institutional quality variables, forecasts have been constructed up to 2025 using an autoregressive of order m, ie AR (m) model with m=2 for each variable. Once the above-specified econometric regression equation has been run by including all these determinants of SEs (within X) the link between these variables and SE size could be established. Specifically, the estimated coefficients for all these variables represent the effect of these variables on the SE up to 2012. Next, forecasted values of the determinants of SE to 2025 are combined with the estimated coefficients to construct forecasts of the SE size. For the forecast of the world as a whole, several other explanatory variables, for which estimates are available up to 2025, have also been included in X in the above equation. These are the percentage of people living in heavily populated and deprived areas and of people living in areas with periodic water shortages, share of the market for industrial robotics, the size of global 3D printing market as percentage of the global market, the market share of the ‘maker economy’ as well as the barter economy, and finally the share of virtual currencies in the market.