The Useful Economic Model: Social Surplus Modeling and the Inductive Approach

  • Published on

  • View

  • Download

Embed Size (px)


Grad Students Honoring Fred Lee session at 12th International Conference


<ul><li> 1. The Useful EconomicModelSocial Surplus Modeling and the InductiveApproachNicola MatthewsUMKC</li></ul> <p> 2. Contents: Introduction Modeling and Credibility Nature of the Model Role of Ontology-Epistemology Model Credibility Social Surplus Models The Useful Model 3. Intro. Modeling is both an orthodox and heterodox approachand is dated back to the Classical economists of the1700s Modeling is the what conventional economists do: If itisnt modeled, it isnt economics, no matter howinsightful (Colander, 2004) The practice and type of modeling in standardeconomic theory is non-controversial The practice and type of modeling in heterodoxeconomic theory is controversial 4. Cont. Intro3 Heterodox Perspectives Towards Modeling (Lee, 2014)1. Models are mathematically overly complex and do not reflect thereal world, thus they should be avoided as a choice of method2. Models are credible, if they are built on heterodox theory3. Models are credible and can contribute to the advancement ofeconomic theory, if they are empirically grounded 5. What Is an Economic Model? A model is an abstract intellectual representation of a system, aminiature world A system is set of interacting and interdependent elements which togetherconstitute a whole; a system may be nested and its size can vary Systems are typically complex and difficult to analyze Closed-modeling and Linear modeling limit the amount of assistance themodel can provide Thus, modeling is a systems approach towards science that is, it is aMETHOD towards science An advantage of using economic models as a method is that they canbe manipulated to reveal structural relationships 2 General Approaches Toward Model construction Deductive Inductive 6. Conventional, Deductive Economic Models Currently one dominant form of reasoning in conventional modelconstructiondeduction Begin with a series of presuppositions based on reductive and idealizedframework (ex. individualism, rational acting, profit seeking, etc.) Create axioms based on presuppositions Deduce model from these premises; choice of math technique assumed prior topresuppositions Note: there is no discovery process in the formation of the model, althoughthere is a minimal amount of induction Original assumptions are not derived in a vacuum but based partially onobservation Problem: used reductionism and idealization (like a caricature, blow up onefeature) Model becomes distorted Reductionist modeling 7. Alternative, Inductive Economic Model Inductive reasoning towards model construction Begin with a series of presuppositions based on cumulativeobservations (capitalist state, agency, class analysis) Make observations, collect data, categorize Sort, combine and filter data based on ``recipe makingframework Build model on categorical data; primary math techniquedepends on applicability and assessed after data collection Discovery process begins at the start of model construction; thetype of structures are not assumed a priori Limited distortions Constructivist modeling One should subordinate deduction to induction, anddiscover the empirical regularities first...(Kaldor, 1985) 8. Role of Ontology-Epistemology in ModelBuilding The world and the universe are real and can bemeasured but.. All science is practiced through the human habitualmind and its associated culture/history/politics, makingdirect (appearances) and indirect (essence)measurements imprecise How do we know this? Failure of theories and theexistence of paradigms 9. Role of Ontology-Epistemology in ModelBuilding Obstacles towards precise science Neuraths Boatscience conducted ON Neuraths boat, cant bringthe boat to land for complete overhaul The active element of science is how we rebuild and navigatethe boat not that the objective world exists (passive) No such thing as a view from nowhere (Barker and Feiner, 2004;Nelson, 1996) Multiple sets of conditions which may yield same or very similarresults (Lowe, 1936) Duhem-Quine Theoryimpossible to test a single hypothesis inisolation = logical-positivism untenable Theory dependence of observationthe actual process ofobserving, collecting, recording and categorizing data is notfree from subjectivity-bias as it is dependent upon theories todo so (Hanson,1965 [1958]) (more relevant to inductivemodeling) Deductive reasoning weak, one break in chain, entire theoryinvalid (Peirce, 1868) 10. Model Credibility Conventional Deductive Models Credibility relies on predictive power; Positivism No criteria to model building apart from linked-chaindeductive reasoning Subjective benchmarks for prediction (Lee, 2014) Model is intended to support the underlying axioms andmath, NOT to be used as a method in theorydevelopment Result: Reductionist Model-Theory Duhem-Quine Thesis = logical positivism not crediblewhether the data is correlated to the real world or not Singular deductive approach is weak Suffers from subjectivity-bias; the choice of tool prior to thedevelopment of axioms; the model behaves as the theorynot as a method 11. Model Credibility Heterodox Inductive Models Credibility relies on real world observations andmeasurements Model is intended to aid in the discovery process andassist in theory development Result: Constructivist Model-Method But still.. Theory Dependence of Observation, NeurathsBoat No Guarantee that observations, measurements andcategorization represents ultimate reality of the objectiveworld 12. 3 Heterodox Perspectives (Lee, 2014)1. Models are mathematically overly complex and do not reflect thereal world, thus they should be avoided as a choice of method2. Models are credible, if they are built on heterodox theory3. Models are credible and can contribute to the advancement ofeconomic theory, if they are empirically grounded 13. Surplus Models: Sraffa Sraffian Model Presuppositions: Circular production (basic commodities go into all production, indecomposable) Homogenous labor Fixed technical coefficients Convergence (uniform rates of profit) Real wages (commodity numeraire) Closed System Value determined by social system of production Fixed Technology Given Surplus (difference between the social product and the total amount of intermediateinputs) Class Analysis Capitalist State No Agency Results: Inverse relation between wages and profits Role of Prices = are exchange ratios of commodities relative to the numeraire, coordinateeconomic activity 14. Sraffian Surplus Model Credibility of Sraffian Model Model is semi-deductive, semi-inductive Quasi model-theory, model-method Sraffa closes his system by assuming a given surplus, a giventechnical condition and no agency Surplus is a residual, meaning there is no explanation for the leveland composition of the social product Able to solve for prices in the system independent of Supply &amp;Demand Critique of the Marginal Theory of Value and Distribution Model devised to support theory not to discover it Model is a theoretical exposition Taking a host of givens, certain properties fall out of the model 15. Surplus Models: Frederic S. Lee Lees Model Presuppositions: Circular production (basic commodities go into all production, productionindecomposable); input-output Heterogeneous labor Fixed technical coefficients No convergence (parameters are not fixed) Nominal wages Open System Value determined by social system of production Surplus not given Class Analysis Capitalist State State Money Agency Results: Surplus = determined by capitalist investment decisions and government spendingdecisions No inverse relation between wages and profits Role of Prices = serve to reproduce the system (going-concern prices), not tocoordinate the system Matrix Algebra 16. Lees Heterodox Surplus Model Credibility of Lees Model Model is primarily inductive; constructed from the ground-up Surplus is not a residual, meaning there is an explanation for the leveland composition of the social product Investment and government decisions determine surplus, surplus determinesthe social product Price theory explains how prices are determined (based on budgetedoutput) and their role (going-prices) but nothing more Model created to assist in the discovery process and theorydevelopment Model credible in that it was built from real world observations andmeasurements where structures and patterns emerged But.still do not know for certain (Theory Dependence ofObservation) Yetin the right direction induction 17. Induction vs. Deduction Lees Heterodox Model has built upon the SocialSurplus Approach Philosophy ought to imitate the successful sciences in itsmethods, so far as to proceed only from tangiblepremises which can be subjected to careful scrutiny, andto trust rather to the multitude and variety of its argumentsthan to the conclusiveness of any one. Its reasoningshould not form a chain which is no stronger than itsweakest link, but a cable whose fibers may be ever soslender, provided they are sufficiently numerous andintimately connected (Peirce, 1868) 18. Model as Method, Model as Abduction =Model as Useful, Model as Tool Sometimes we do not know how to weave the fibers in the thread; canuse abduction Abduction is a type of inference that generates a conjectured solutionto a problem that is not readily traceable (Peirce, 1933) It is ``a sudden advance towards the solution...[it is] a bright idea, a goodidea, a happy thought, a brain-wave (George Polya, 1945, p.146) Abduction is an interruption of habitual thinking Abductive inference can be triggered by: Reorganizing, reshuffling, decomposing and recombining, through lateralthinking, through creative thinking, through brainstorming or even throughthe use of analogies Michael Polanyi (1958) and Charles Wright Mills1(1959) some of theabove technique as ``ransack[ing] our memory for any similar problem"(p.128) and ``get[ing] a comparative grip on the materials respectively 19. Humble the Model = Model as Tool General perceptions of Models and Modeling At Bestviewed as a sense of accomplishment, rigorous At Worstviewed as a theory To use model as a tool for theory development, it must besubdued, it must be humbled Opens the door to inquiry, specifically inquiry of thepresuppositions Facilitates the abductive process Empirically grounded modeling can be credible but only to adegree Empirically grounded modeling has value as a method;applicable mathematical techniques do not negate this 20. EndThank You </p>