This paper explores the integration of machine learning into economics and social sciences, assessing its potential impact and limitations. It introduces fundamental machine learning concepts and principles, highlighting the differences between the two disciplines, particularly the focus on causal inference in economics and prediction in machine learning. The paper discusses diverse applications of machine learning, from extracting insights from unstructured data to creating novel indicators and improving predictive accuracy, while also addressing challenges related to data quality, computational efficiency, and data (...) ownership. It emphasizes the importance of standardization, transparency, and ethical considerations in prediction tasks, recognizing that machine learning is a powerful tool but cannot replace economic theory. Ultimately, researchers remain optimistic about the transformative potential of machine learning in re-shaping research methodologies and generating new insights in economics and social sciences. (shrink)

Within the discussion of ethics and economics some have considered designing a code of ethics for economists. But the idea of such a code is potentially problematic from a pluralist standpoint. Some possibilities are discussed here to show that any code concerning the behaviour of economists presumes a particular view of human nature and thus of professionalism. Further, issues of socio-economic power in the profession pose problems for the interpretation and implementation of some possible principles, notably those referring to standards (...) of competence and truth-seeking. It is therefore concluded that any code of ethics should take the form of general guidelines, with primacy given to the ethics of pluralism: tolerance, even-handedness and open-mindedness, on which the interpretation of all other ethical considerations rests. Read the Open Peer Discussion on this paper » Read the discussion of this paper on the WEA's Ethics conference ». (shrink)