I am an Associate Professor in the School of Economics at the University of Leicester, Leicester, LE1 7RH, United Kingdom.
My email address is mhs15@le.ac.uk

Completed Working Papers

Biased Information and Opinion Polarisation 

First Version: 26th May, 2025.     [Download]

Why do people form polarised opinions after receiving the same information? Why does disagreement persist even when public information is abundant? We show that a Bayesian model with potentially biased public signals can answer these questions. When agents are uncertain and disagree about the bias in the signals, persistent disagreement and opinion polarisation can readily emerge. This happens because uncertainty surrounding the bias induces agents with diverse initial beliefs to form drastically different posterior estimates. Prolonged exposure to these signals can in some cases drive the agents’ opinions further away from each other and also further away from the truth.

Information Quality, Disagreement and Political Polarisation 

(Joint with Emre Aytimur)  Games and Economic Behavior, R&R   [Download]

First Version: 17^th May, 2024.

How does the quality of information received by voters affect political polarisation? We address this long-standing question using an election competition model in which voters have to infer an unknown state from some noisy and biased signals. Their policy preferences are shaped by the posterior belief, which is unknown to the parties when they choose their platforms. The greater the uncertainty faced by the parties, the greater the incentive to polarise. We show that better information can either promote or suppress polarisation, depending on the gap between voters' and politicians' beliefs (disagreement). We also examine the welfare implications of polarisation.

• Mathematical Appendix
• MATLAB codes
• Working paper version

Precautionary Saving Behaviour under Ambiguity

First Version: 2^nd August, 2022.   [Download]

This paper analyses a two-period model in which a consumer faces a future income risk but is uncertain about its probability distribution. We derive three sets of sufficient conditions under which a consumer with generalised recursive smooth ambiguity (GRSA) preferences will save more under ambiguity than in a deterministic environment. Our results show how precautionary saving is jointly determined by attitudes toward atemporal risk, ambiguity and intertemporal substitution. We also find a close connection between risk prudence under non-expected utility and precautionary saving under GRSA preferences.

Research Policy and U.S. Economic Growth

First Version: August 14, 2013.   [Download]

This paper examines quantitatively the effects of R&D subsidy and government-financed basic research on U.S. economic growth and consumer welfare. To achieve this, we develop an endogenous growth model which takes into account both public and private research investment, and the differences between basic and non-basic research. A calibrated version of the model is able to replicate some important features of the U.S. economy over the period 1953-2009. Our model suggests that government spending on basic research is an effective policy instrument to promote economic growth. Subsidizing private R&D, on the other hand, has no effect on economic growth.

Concave Consumption Function and Precautionary Wealth Accumulation

[Download]

This paper examines the theoretical foundations of precautionary saving behavior in a canonical life-cycle model where consumers face uninsurable idiosyncratic labor income risk and borrowing constraints. We begin by characterizing the consumption function of individual consumers. We show that the consumption function is concave when the utility function has strictly positive third derivative and the inverse of absolute prudence is a concave function. These conditions encompass all HARA utility functions with strictly positive third derivative as special cases. We then show that when consumption function is concave, a mean-preserving increase in income risks would encourage wealth accumulation at both the individual and aggregate levels.

Publications

How Does Political Uncertainty Affect the Optimal Degree of Policy Divergence? (with Emre Aytimur and Aris Boukouras)

Social Choice and Welfare, Published Online.

We examine how the optimal degree of policy divergence between two policy platforms in an election is affected by two types of aggregate uncertainty: policy-related and candidate-specific. We show that when the candidate-specific uncertainty is sufficiently large, policy convergence becomes optimal. We also show that when these two types of uncertainty co-exist, only purely office-motivated parties result in policy convergence, in other words, any level of policy motivation of parties results in some policy divergence, making policy motivation undesirable when candidate-specific uncertainty is sufficiently large.

• Working Paper Version

Source of Economic Growth in Models with Non-Renewable Resources (with Hongsilp Sriket)

Journal of Macroeconomics, vol. 62, June 2022.

This paper re-examines the conditions under which endogenous economic growth can emerge in neoclassical models with non-renewable resources. Our analysis is based on a general production function which encompasses the Cobb-Douglas specification. We show that endogenous growth is possible only when the elasticity of substitution between effective labour input and effective resource input is constant and equal to one. If this does not hold (as some empirical studies suggested), then economic growth is solely driven by an exogenous technological factor. We also show that the assumption on this elasticity will affect the model's policy implications in regard to resource taxation.

• Mathematical Appendix

Diversity and Economic Performance in a Model with Progressive Taxation (with Wei Wang)

Economic Journal, vol. 129, 2019.
Is a more heterogeneous population beneficial or harmful to long-term economic performance? This paper addresses this and other questions in a dynamic general equilibrium model where consumers have different labour productivity and time preference. We show how differences in the cross-sectional distribution of these characteristics can affect the economy via two channels. The first one involves changing the composition of the labour force; and the second one involves changing the cross-sectional distribution of marginal tax rate. We show how these channels are, respectively, determined by the shape of the labour supply function and the curvature of the marginal tax function.

• Working paper version
• MATLAB codes

Standard Risk Aversion and Efficient Risk Sharing

Economics Letters, vol. 173, 2018.
This paper analyzes the risk attitude and investment behavior of a group of heterogeneous consumers who face an uninsurable background risk. It is shown that standard risk aversion at the individual level does not imply standard risk aversion at the group level under efficient risk sharing. This points to a potential divergence between individual and collective investment choices in the presence of background risk. We show that if the members' absolute risk tolerance is increasing and satisfies a strong form of concavity, then the group has standard risk aversion.

• Technical Appendix
• MATLAB codes

Asset Bubbles in an Overlapping Generations Model with Endogenous Labor Supply (with Lisi Shi)

Economics Letters, vol. 123, 2014.

This short paper examines the effects of asset bubbles in an overlapping generations model with endogenous labor supply. We derive a set of conditions under which asset bubbles will lead to an expansion in steady-state capital, investment, employment and output. We also provide a specific numerical example to illustrate these results.

Time Preference and the Distributions of Wealth and Income

Economic Inquiry, vol. 52, 2014.
This paper examines the connection between time preference heterogeneity and economic inequality in a deterministic environment. Specifically, we extend the standard neoclassical growth model to allow for (i) heterogeneity in consumers' discount rates, (ii) direct preferences for wealth, and (iii) human capital formation. The second feature prevents the wealth distribution from collapsing into a degenerate distribution. The third feature generates a strong positive correlation between earnings and capital income across consumers. A calibrated version of the model is able to generate patterns of wealth and income inequality that are very similar to those observed in the United States.

• Working paper version
• MATLAB Codes

A Quantitative Analysis of Suburbanization and the Diffusion of the Automobile* (with Karen A. Kopecky)

International Economic Review, vol. 51 (4), 2010.
Suburbanization in the U.S. between 1910 and 1970 was concurrent with the rapid diffusion of the automobile. A circular city model is developed in order to access quantitatively the contribution of automobiles and rising incomes to suburbanization. The model incorporates a number of driving forces of suburbanization and car adoption, including falling automobile prices, rising real incomes, changing costs of traveling by car and with public transportation, and urban population growth. According to the model, 60 percent of postwar (1940-1970) suburbanization can be explained by these factors. Rising real incomes and falling automobile prices are shown to be the key drivers of suburbanization.

*This paper was previously circulated with the title “Suburbanization and the Automobile.”

Working paper version

Finite State Markov-Chain Approximations to Highly Persistent Processes (with Karen A. Kopecky)

Review of Economic Dynamics, vol. 13 (3), 2010.

The Rouwenhorst method of approximating stationary AR(1) processes has been overlooked by much of the literature despite having many desirable properties unmatched by other methods. In particular, we prove that it can match the conditional and unconditional mean and variance, and the first-order autocorrelation of any stationary AR(1) process. These properties make the Rouwenhorst method more reliable than others in approximating highly persistent processes and generating accurate model solutions. To illustrate this, we compare the performances of the Rouwenhorst method and four others in solving the stochastic growth model and an income fluctuation problem. We find that (i) the choice of approximation method can have a large impact on the computed model solutions, and (ii) the Rouwenhorst method is more robust than others with respect to variation in the persistence of the process, the number of points used in the discrete approximation and the procedure used to generate model statistics.

• Longer version
• Computer Codes