Bereket Kebede School of International Development University of East Anglia London (UEA London) 10 December 2011 10:00 a.m. – 5:00 p.m.

Epistemological Issues in International Development

Epistemologies in economics

Bereket KebedeSchool of International Development

University of East AngliaLondon (UEA London)

10 December 201110:00 a.m. – 5:00 p.m.

1. An excursion into a history of economics2. Dominant epistemologies3. Mathematics and formalism4. Observational data and econometrics 5. Experimental methods

Outline

Publication of A. Smith’s Wealth of Nations in

1776 (‘birth of economics’) and subsequent classical economists (Ricardo, Malthus, Marx, Mill, etc.)

Economics was not yet a really distinctive and unified discipline (political economy, moral philosophy, etc.)

The marginalist revolution and the rise of neo-classical economics (Jevons, Menger, Walras, Marshall) – dominant school after end of 19th century

1. An excursion into a history of economics

Economics as we know it, based on

neoclassical school, emerged after WWII (the ‘formalist revolution’) Hick, Value and Capital (1939) Samuelson, Foundations of Economic Analysis

(1947) Arrow, Social Choice and Individual Values

(1951) von Neumann and Morgenstern, Theory of

Games and Economic Behaviour (1953) Debreu, Theory of Value (1959)

1. An excursion into a history of economics (cont’d)

With the dominance of modern conventional

economics, issues of methodology and epistemology were left in the background Economic epistemologists (philosopher rather

than economists) separate from standard economists

Recent revival of interest in epistemological and philosophical problems of economics

Development economics: a separate discipline or a branch of main stream economics?

1. An excursion into a history of economics (cont’d)

Positivism: the application of methods of natural

sciences to the study of social reality (Bryman, 2004) Phenomenalism: knowledge confirmed by the senses

can genuinely be warranted Deductivism: theory to generate testable hypotheses

to identify laws; e.g., economic theory Inductivism: gathering of facts to provide basis for

laws; e.g., survey data and econometrics Objective: science should be conducted as value free Positive vs. normative: the latter prescriptive; e.g.,

efficiency vs. income distribution

2. Dominant epistemologies

Post-positivism: knowledge based on

probabilistic propositions rather than certainty; approximate truth; the ‘uncertainty principle’

Methodological individualism – both an organising strategy for neoclassical economics and an ontological commitment

Recent interest in alternative epistemological approaches in economics: e.g., Lavoie, ed., Economics and Hermeneutics

2. Dominant epistemologies (cont’d)

“Those who have handled sciences have been either men of experiment or men of dogmas. The men of experiment are like the ant, they only collect and use; the reasoners resemble spiders, who make cobwebs out of their own substance. But the bee takes a middle course: it gathers its material from flowers of the garden and of the field, but transforms and digests it by a power of its own … Therefore from a closer and purer league between these two faculties, the experimental and the rational (such as has never yet been made), much may be hoped.”

Francis Bacon, 1620

3. Mathematics and formalism

The reasoners/spiders/theoretical

economists? – use simplified and formal models

The bees/empirical economists? – use mainly econometrics to analyse data testing theories

(Wo)Men of experiments/experimentalists? – both random controlled trials (RCTs) and experimental economics

3. Mathematics and formalism (cont’d)

Modern mainstream theoretical economics uses

simplified, abstract, formal mathematical models Starting from simple axioms and assumptions

builds an internally consistent logical system Theoretical models are understood as simplified

representations of important real economic relationships

Idealisation as a price worth paying to understand important economic relationships (ontological problem?)

3. Mathematics and formalism (cont’d)

“Ultra-empiricism”- postulates, assumptions,

axioms and basic principles in economic models should be testable

Friedman (1953) – “ultra-empiricism” would eliminate much useful theorising Basic assumptions of theoretical models neither

should be true nor testable Usefulness of models depends on predictive

reliability (pragmatism, instrumentalism, indirect empiricism)

Judged on fit with parallel theorising

3. Mathematics and formalism (cont’d)

Formalism: an approach (mathematical or

otherwise) to theorising that aims at making explicit the logical structure of a theory; set-theoretic formalism widely accepted in economic theory Formal models as simplified representation of

economic reality; also remember Friedman Formal models as training tools; improves

perception and analytical skills of economists

3. Mathematics and formalism (cont’d)

Formalism and mathematics still important in

recent critiques of conventional economics New Institutional Economics: instead of

individual agency the role of institutions emphasised; a dominant part of the school retains and prioritises formalistic methods

Behavioural Economics: using insights from psychology, relaxes the assumption of the hyper-rational economic agent of conventional economics; mostly widened the scope of formal economic models

3. Mathematics and formalism (cont’d)

Complexity and computational economics:

economic interactions are too complicated for capturing in neat closed models E.g., agent-based computational economics:

economies as dynamic systems of interacting agents; agents can range from active data-gathering decision-makers with sophisticated learning capabilities or as passive with no cognitive functioning

Difficult to analytically solve the outcome of these complex interactions; left for computer algorithms including artificial intelligence

3. Mathematics and formalism (cont’d)

Most empirical analysis in economics in

general and development economics in particular uses observational rather than experimental data – need for quasi-experimental method, mainly econometrics

‘Recreating’ an experimental set-up without experimental data – controlling for confounding variables and orthogonal residuals

4. Observational data and econometrics

Tension between theory-led and data-

led econometrics (the LSE school); difference on how knowledge is to be attained rather than ontological differences in economic entities

Single econometric tests rarely make significant changes; accumulated weight of evidence is required to bring paradigm shift (a la Lakatos)

4. Observational data and econometrics (cont’d)

Economics considered as non-experimental

discipline at least until the 1980s Recent trends:

Randomised control trials (RCTs): identifying project/treatment effects by randomising interventions

Lab experiments (experimental games): ‘players’ making decision under incentivised and controlled situation (laboratory)

5. Experimental methods

Increasing interest in the epistemology of

experimental economics (e.g., Santos, 2010) The social epistemology of experiment (SEE):

Experimental facts generated through material and social processes

Material processes: ‘phenomenal model’ (experimenter’s view of how the world works), ‘instrumental model’ (view on how the experimental apparatus works), ‘material procedure’ (actually how the experimental apparatus works)

Social processes: the experimental subjects and the experimentalist

5. Experimental methods (cont’d)

The direct participation of human subjects is

the major source of epistemic value; but participants may ‘resist’ expectations of economists

Experimental games are amenable to scrutiny by wide audience

SEE pays special attention to scientists predisposition to obtain confirmation of prior beliefs (‘confirmation bias’)

5. Experimental methods (cont’d)

Boettke, Peter J.; Christopher J. Coyne; John Davis; Francesco Guala; Alain Marciano; Jochen Runde and Margaret Schabas. 2006. "Where Economics and Philosophy Meet: Review of the Elgar Companion to Economics and Philosophy with Responses from the Authors." Economic Journal, 116(June), F306-F25.

Bryman, Alan. 2004. Social Research Methods. Oxford: Oxford University Press.

Davis, John B.; Alain Marciano and Jochen Runde eds. 2004. The Elgar Companion of Economics and Philosophy. Cheltenham, UK, Northampton, MA, USA: Edward Elgar.

Friedman, Milton. 1953. Essays in Positive Economics - the Methodology of Positive Economics. Chicago: University of Chicago Press.

References

Hoover, Kevin D. 2006. "The Methodology of Econometrics," T. C.

Mills and K. Patterson, Palgrave Handbook of Econometrics: Econometric Theory: Econometric Theory. New York: Palgrave Macmillan,

Hoover, Kevin D. 2010. "Pragmatism, Perspecitval Realism and Econometrics," Durham, NC: Duke University,

Lavoie, Don ed. 1990. Economics and Hermeneutics. London and New York: Routledge.

Sagal, Paul T. 1977. "Epistemology in Economics." Journal for General Philosophy of Science, 8(1), 144-62.

Santos, Ana Cordeira dos. 2010. The Social Epistemology of Experimental Economics. London and New York: Routledge.

References