Evolution of consumption distribution and model of wealth distribution in China between 1995 and 2012

•The PDFs (Pe(x)=K(x−μ)e−(x−μ)22σ2) of consumption appear identical with PDFs of income in China.•The consumption of majority population is strongly influenced by their income.•The residue PDFs between 1995 and 2012 are close to a Gaussian distribution.•Gaussian distribution of residue PDFs derive from high propensity of savings.•High propensity of savings is attributed to the rapidly increasing payments.

We study the evolution of the distribution of consumption of individuals in the majority population in China during the period 1995–2012 and find that its probability density functions (PDFs) obey the rule Pc(x)=K(x−μ)e−(x−μ)22σ2. We also find (i) that the PDFs and the individual income distribution appear to be identical, (ii) that the peaks of the PDFs of the individual consumption distribution are consistently on the low side of the PDFs of the income distribution, and (iii) that the average of the marginal propensity to consume (MPC) is large, MPC¯=0.77, indicating that in the majority population individual consumption is low and strongly dependent on income. The long right tail of the PDFs of consumption indicates that few people in China are participating in the high consumption economy, and that consumption inequality is high. After comparing the PDFs of consumption with the PDFs of income we obtain the PDFs of residual wealth during the period 1995–2012, which exhibit a Gaussian distribution. We use an agent-based kinetic wealth-exchange model (KWEM) to simulate this evolutional process and find that this Gaussian distribution indicates a strong propensity to save rather than spend. This may be due to an anticipation of such large potential outlays as housing, education, and health care in the context of an inadequate welfare support system.