Using an actuarial model to assess own-source revenue streams in county governments using hypothetical data

Abstract

The paper employs an actuarial and stochastic modelling framework to assess the performance, stability and downside risk properties of their own-source revenue (OSR) streams in county governments on the basis of hypothetical but policy-realistic data calibrated to the Homa Bay County, Kenya. A Gamma-based Generalized Linear Model of revenue sources, with likelihood-based estimation, and backed by back-testing to measure predictive power and Monte Carlo simulation to create probabilistic new revenue streams are used. The tail-risk exposure is quantified using risk-sensitive indicators such as Value-at-Risk (VaR) and Conditional Value-at-Risk (CVaR), and a risk-adjusted optimisation model is figured out to compute the optimum allocation of enforcement effort among revenue streams. The findings indicate that land rates and business permits have high expected returns, positive growth coefficients, small back-testing error rates, and simulated revenue dispersion than other OSR streams. Such sources are thus more actuarial and fiscal reliable. On the contrary, the volatility of parking fees and market fees is higher, the forecast performance is lower, and the range of downside risk, in its turn, is larger, meaning that both are more prone to changes in revenues. The results of the optimisation further show that by giving priority on enforcement on land rates and business permits maximise the revenue expected to be generated and minimise the downside fiscal risk. The research paper has shown that actuarial and stochastic techniques can be beneficial in improving the OSR assessment and fiscal risk management within devolved governmental frameworks. The model offers strict quantitative foundation of revenue prioritisation, conservative budget planning, and evidence-based enforcement strategy and can be implemented to larger levels of inter-county comparative or policy implementation in future studies.