Economics Dictionary of ArgumentsHome
| |||
|
| |||
| Economic models: Economic models are simplified representations of real-world economic processes used to analyze, predict, and understand economic phenomena. They employ assumptions and mathematical frameworks to illustrate relationships between variables, aiding in decision-making and policy formulation. See also economics, Models, Model theory, Simulation, Economic Theories, Experiments, Method._____________Annotation: The above characterizations of concepts are neither definitions nor exhausting presentations of problems related to them. Instead, they are intended to give a short introduction to the contributions below. – Lexicon of Arguments. | |||
| Author | Concept | Summary/Quotes | Sources |
|---|---|---|---|
|
Jonah B. Gelbach on Economic Models - Dictionary of Arguments
Parisi I 33 Economic models/Gelbach/Klick: Econometric studies come in two basic flavors: structural and reduced form. Structural modeling involves writing down an explicit mathematical and statistical representation of the determinants of individual, firm, or organizational behavior, such that these relationships can be captured with a finite collection of parameter estimates Reduced form: Reduced form work instead involves attempting to estimate more generally defined contextual objects such as the average treatment effect of past implementations of policy changes. (...) it is possible that one doesn’t learn as much from reduced form estimation as from valid structural estimation. Thus, the choice between structural and reduced form approaches can involve trading off the need to make stronger assumptions (structural work) against the prospect of learning less information (reduced form work) that could prove to be valuable. Parisi I 34 Omitted variables: The fundamental challenge in this context is omitted variable bias. >Empirism/Economic theories. That is, when attempting to isolate the causal effect of policy P on outcome Y through, say, the use of multiple regression analysis, it is necessary to rule out the possibility that any estimated effect is driven by unobserved (or at least uncontrolled for) variables that happen to be correlated with P. Terminology: This general omitted variable bias problem goes by many names (e.g., endogeneity, selection effects, reverse causality, simultaneity, etc.),(...). Suppose we are interested in how changes in a policy P affect some continuous outcome variable Y. Traditional solution: A traditional way to model the relationship between these variables was to assume that there is a parametric function F that relates them structurally, through a combination of assumptions on individual behavior, organizations’ cost functions, and market forces (or other aggregating forces) relating them to each other, such that Y = F(P;τ,ε), where τ is a parameter and ε is an unobserved term. The causal effect of a policy change from P1 to P2 is thus to shift Y from F(P1;τ,ε) to F(P2;τ,ε). If we assume that F is linear in P and ε, then the structural relationship between Y and P is captured by the equation Y = Pτ+ε together with the claim that when ε is held fixed, a change in P’s value from P1 to P2 will induce a change of τ units in Y’s value. On this account, the parameter τ measures the causal effect on Y of a one-unit change in P. If P and ε are uncorrelated, then the OLS estimator is consistent for this causal effect. On the other hand, if P and ε are correlated, then the OLS estimator will differ from τ even in large samples. (...) Parisi I 37 Policies: The key to policy-relevant empirical work, then, involves two questions. First, is it reasonable to assume that ε and P are mean-independent, or that there is a linear structural relationship between Y and P, with P and ε uncorrelated? The second key question is how to estimate causal effects when it is not reasonable to assume that either situation (A) or (B) holds. An enormous amount of modern empirical work is focused on answering this question. Random assignment: One approach to solving the problem of dependence between ε and P is to assign policy levels to units randomly. This approach, common in studies involving the effects of medical and psychological interventions, is frequently used in empirical economics (...).The advantage of random assignment is that it directly imposes the mean independence of ε and P, so that τ may be regarded as the causal effect of the policy, at least within the particular population studied experimentally. For this reason, it is common in the empirical economics literature to consider randomized controlled trials (RCTs) the conceptual benchmark against which other study types are measured. Parisi I 38 Randomized controlled trials/problems: This is surely too strong a claim, as Heckman and Smith (1995)(1) and Deaton (2010)(2) have ably discussed, because RCTs do have potentially important drawbacks. One drawback is that not all questions are susceptible to study using RCTs. RCTs cannot measure what are sometimes called “general equilibrium effects,” that is, effects that a policy change has to behavior outside the study’s domain of impact. >Randomized assignment/Economic theories. 1. Heckman, James J. and Jeffrey A. Smith (1995). “Assessing the Case for Social Experiments.” Journal of Economic Perspectives 9(2): 85–110. 2. Deaton, Angus (2010). “Instruments, Randomization, and Learning about Development.” Journal of Economic Literature 48(2): 424–455. Gelbach, Jonah B. and Jonathan Klick „Empirical Law and Economics“. In: Parisi, Francesco (ed) (2017). The Oxford Handbook of Law and Economics. Vol 1: Methodology and Concepts. NY: Oxford University Press._____________Explanation of symbols: Roman numerals indicate the source, arabic numerals indicate the page number. The corresponding books are indicated on the right hand side. ((s)…): Comment by the sender of the contribution. Translations: Dictionary of Arguments The note [Concept/Author], [Author1]Vs[Author2] or [Author]Vs[term] resp. "problem:"/"solution:", "old:"/"new:" and "thesis:" is an addition from the Dictionary of Arguments. If a German edition is specified, the page numbers refer to this edition. |
Gelbach, Jonah B. Parisi I Francesco Parisi (Ed) The Oxford Handbook of Law and Economics: Volume 1: Methodology and Concepts New York 2017 |
||
> Counter arguments against Gelbach
> Counter arguments in relation to Economic Models
