Economic modeling using evolutionary algorithms: the effect of a binary encoding of strategies

We are concerned with evolutionary algorithms that are employed for economic modeling purposes. We focus in particular on evolutionary algorithms that use a binary encoding of strategies. These algorithms, commonly referred to as genetic algorithms, are popular in agent-based computational economics research. In many studies, however, there is no clear reason for the use of a binary encoding of strategies. We therefore examine to what extent the use of such an encoding may influence the results produced by an evolutionary algorithm. It turns out that the use of a binary encoding can have quite significant effects. Since these effects do not have a meaningful economic interpretation, they should be regarded as artifacts. Our findings indicate that in general the use of a binary encoding is undesirable. They also highlight the importance of employing evolutionary algorithms with a sensible economic interpretation.     

Wages,implicit contracts,and the business cycle: Evidence from a European panel

We study the joint behavior of hours and wages over the business cycle in a unique panel of 13 European countries, and document significant history dependence in wages. Workers who experience favorable market conditions during their tenure on the job have higher wages, and work fewer labor hours. Unobserved differences in productivity, such as varying job quality, or match-specific productivity are not likely to explain this variation. The results instead point to the importance of contractual arrangements in wage determination. In economies with decentralized bargaining practices, such arrangements resemble self-enforcing insurance contracts with one-sided commitment (by the employer). On the other hand, in countries with strong unions and centralized wage bargaining, wage behavior is better approximated by full-commitment insurance contracts. The co-movement of hours and wages further confirms a contractual framework with variable worker hours. Despite the strong prevalence of contracts in Europe, however, the elasticity of labor supply is considerably smaller compared to the U.S. labor market.     

Mechanical engineering and issues on teaching mechanical engineering design in Turkey

Students with relatively low affinity and/or weak ability for science and mathematics have been presently acknowledged into Mechanical Engineering (ME) departments around the world. This has resulted in apparent lowering of fully competent graduate engineers, due to insufficient comprehension and its concomitant cognitive issues. Now, the complexity and the demand of ME education have been extremely unbalanced with recruited students. This conflict between the education process and less knowledgeable students has been adequately acknowledged, but remedies for this global issue are not yet available. This is no longer sustainable. The paper gives insight into (individual and combined) plausible reasons for fewer fully competent graduate engineers, taking the specific case study in Turkey. It proposes a generic approach, which can be extended to courses of any university/degree subject. Findings on student learning are provided using a hierarchical decomposition. A proposed remedy for this issue has been thoroughly evaluated, and accountability measures, qualitative data and a survey has been conducted for the ME design course taught in Turkey. The findings indicate that the new recruits are not fully to blame for the conflict, because there have also been other reasons for the issues within teaching. It identified multiple instances of the reasons such as unnecessary complexity of textbooks, unsolved contradictions even between the technical component design standards and etc. If these are considered by educators, this will help to reduce the perceived degree of teaching difficulty, and have a positive effect on the quality of graduates. It may even assist in attracting higher ability students into ME.     

-learning agents in a Cournot oligopoly model

Q-learning is a reinforcement learning model from the field of artificial intelligence. We study the use of Q-learning for modeling the learning behavior of firms in repeated Cournot oligopoly games. Based on computer simulations, we show that Q-learning firms generally learn to collude with each other, although full collusion usually does not emerge. We also present some analytical results. These results provide insight into the underlying mechanism that causes collusive behavior to emerge. Q-learning is one of the few learning models available that can explain the emergence of collusive behavior in settings in which there is no punishment mechanism and no possibility for explicit communication between firms.