Before moving into Machine Learning and Artifical Intelligence, I did quite a bit of work in Economics. Below is a collection of a set of papers I wrote as past of my Master’s studies.


Political Conflict and Economic Competitiveness

“Adam Smith’s chief concern was not so much with what man might occasionally achieve when he was at his best but that he should have as little opportunity as possible to do harm when he was at his worst.” - F.A. Hayek

June 2016. A key ingredient for a dynamic economy is a modicum of competition, but with weak institutions powerful elites have strong incentives to collude to generate economic rents. Weak institutions also increase the scope for political violence, as an array of powerful elites have incentives to use coercion for private gains. This paper explores the link between political conflict and the economy’s competitiveness by developing a framework where elites use coercive means to bargain over economic rents that arise through market manipulation. In the absence of a conflict, incumbents create restrictive markets to maximize their rents. A moderate but credible coercive threat against the regime forces incumbents to allow a greater mass of participants in the economy, thus making it more competitive. Economic competitiveness reflects the equilibrium of the underlying intra-elite conflict and shocks to the balance of power alter the competitiveness of the economy. Testing the framework explanatory power on England’s historical development, the paper challenges the popular argument that political and economic development in early modern English development was driven by a rising middle class. Instead, economic and political policy over a six century period follows the changing dynamics of a continuous intra-elite conflict.


Dynamic Macroeconomics Analysis

June 2015. What are the general equilibrium effects of labor protection policies? While most policy debates are concerned with the direct benefits of increased job security against greater reluctance among firms to hire, there is little work on the greater economic effects from these policies. We show in a simple Neo-Classical framework that capital markets are significantly effected negatively, leading to reduced output. Time constraints prevents us from further study, but an important implication of this result is that labor protection policies are more costly for wealthy individuals.

As the major module in a course on Macro Economic modelling and analysis, the paper was written as a full analysis of a proposed research question using macroeconomic theory to set up a dynamic model that was then built in MatLab. The basic idea of the paper is to study within the Neo-Classical model the quantitative effects of introducing labor market institutions that are intended to protect employees from real shocks to the economy. In particular we study and compare the effects of i) a constraint on firms ability to i) reduce their labor force and ii) a minimum real wage rate.


Behavioral Economics

March 2015. In most studies of the public goods game, investing is not an optimal choice. Moreover, unless individuals have other-regarding preferences, neither is costly punishment. Yet, a standard result is the positive effect of (costly) punishment mechanisms. However, only a small minority of these studies factor in the effect of earnings, with no study endogenizing the earnings decision in a repeated game. We show that in such a scenario, the conventional wisdom breaks down, as punishment can cause reduced earnings leading to a net welfare loss. Importantly, we predict that the efficiency of the punishment mechanism depends on what information players have access to. If players can see the actual contribution, punishment is effective, but when only relative information is available, effects of punishment becomes ambiguous.

As part of a course in Behavioural Economics, I wrote a full research article together with my coauthor (barr the actual experiment). Key moments were research proposal and experiment design. Ultimately, the incentives for free-riding now arise from another channel and the ability of a punishment mechanism to prevent such behaviour depends on the design of the punishment mechanism. We design an experiment, isolating the private effort effects within the classic public goods game.