Abstract
As Kazakhstan’s fiscal space narrows, the sectoral structure of public capital allocation becomes critical. This study constructs and solves a network optimization model using Kazakhstan’s official data to allocate limited state capital across sectors, maximizing economic diversification under budget constraints. The methodology integrates input-output analysis, production network analysis, and constrained optimization. The empirical basis comprises Kazakhstan’s 2020–2023 symmetric input-output tables, used to calculate output multipliers, Rasmussen linkages, and centrality indicators for 68 product types aggregated into broad sectors. Diversification is operationalized via the Herfindahl-Hirschman Index of export concentration. The optimization model incorporates diminishing marginal returns, where sector shares in a fixed budget are decision variables, and sector contributions are determined by export-adjusted weights. Calculations show that trade, business services, and metallurgy possess the greatest network leverage, while transportation occupies a middle position; meanwhile, the high centrality of resource sectors does not imply a contribution to diversification. The model’s solution systematically diverges from the actual investment structure: the extractive sector is overinvested relative to its diversification impact, while several non-extractive sectors remain underfunded. The scientific novelty lies in deriving sectoral priorities not qualitatively, but as the solution to a formalized optimization model based on measurable, reproducible network parameters.