BOFIT Seminar - Juha Tervala (University of Helsinki) - Inefficiency Unveiled: A Bayesian DSGE Analysis of Public Investment in China

Co-author: Timothy Watson (Australian National University)


This study employs a Bayesian Dynamic Stochastic General Equilibrium (DSGE) model, calibrated with Chinese data, to assess the impact of public investments, which account for about 16.2% of China's GDP. Despite an expected public capital stock of 256% of GDP, based on a 4.6% depreciation rate and a 0.73 efficiency rate of public investment (IMF's estimate for emerging economies), the actual figure stands at only 152%. This significant discrepancy underscores the inefficiencies in public investments, with only 43% of public investment expenditures enhancing the capital stock. Using the Bayesian DSGE model, the output elasticity (productivity) of public capital is estimated to be just 3%. Simulations with these efficiency and productivity measures indicate that the output multiplier of public investment is merely 0.7. Public investments significantly increase public debt and adversely impact GDP and productivity in the medium term.


Online BOFIT seminars are open to all researchers interested in the subjects covered. Those wishing to attend a seminar are kindly asked to register in advance, by filling in the Online Registration Form.

The registration for each seminar is open until 9:00 am the day of the seminar. You will receive a link to join the seminar by email at the latest one hour before the seminar is scheduled to begin.