economics metho

Question 1

Operationalization

Answer 1

Links theoretical concepts to measurable observations. Determines how theoretical concepts are measured - can be measured in different ways

Question 2

Operationalism

Answer 2

If a concept is measured differently, it’s not the same concept.

Question 3

The deductive-nomological (NM) model

Answer 3

A scientific theory according to logical positivists:
Explenans
Law(s) - verified (and true) synthetic a posteriori’s that are unrestricted generalization
Initial conditions - verified (and true) synthetic a posteriori’s
= Explanandum (3) - synthetic a posteriori
Can be used (ex ante) to predict the explanandum with explanans
Can be used (ex post) to explain the explanandum with explanans
Symmetry thesis: explaining (3) requires finding (1) and (2). If you have (1) and (2) you can predict (3) - Explanation and prediction are symmetric.

Question 4

Scientific Laws
Problem:

Answer 4

1 Scientific theories should be completely objective.
2 Scientific theories should have at least one scientific law.
3 But claiming that a statement is a scientific law is subjective!

Question 5

What is a scientific law?

Answer 5

1 A scientific law should be verified (and true) unrestricted generalization.
“All of the coins are 50 cent coins” → “accidental generalization” – not scientific law
2 But it should also identify casually relevant factors.
“As ice cream sales increase, the rate of drowning deaths increases.” - not scientific law

Question 6

Hume’s Problem of Induction

Answer 6

Induction: Deriving generalizations from limited observations.
Issue: Cannot guarantee universal truth (e.g., “All swans are white”).
⇒ Impossible to infer unrestricted generalization from a limited set of observations!
Conclusion: Induction is probabilistic, not certain.

Question 7

Instrumentalism

Answer 7

(Moritz Schlick):
Scientific laws are “usable guidelines,” not verified truths,
⇒ “Instruments” that are not verified but that serve as a “foundation” of scientific knowledge.
Universal statement = instrumental guideline → not meaningful (meaninful statement = certain statement)

Question 8

Confirmationism

Answer 8

(Rudolf Carnap):
Scientific laws are probabilistic, with a “degree of confirmation.”
⇒ Scientific laws are not verified but “confirmed” regularities, with a “degree of confirmation” (which is less than 100%).

Question 9

Keynes-Tinbergen Debate

Answer 9

Jan Tinbergen:
Developed first macroeconometric models (Netherlands, US).
Used mathematical models with fixed coefficients and least squares estimation.
Keynes’ Criticism:
Omitted Variable Bias:
Estimates are biased if not all causal factors are included.
Some factors are unmeasurable.
Data Mining:
Too much trial and error, too little theory.
Invariance Problems:
Estimated relationships may not hold in the future
Key Insight: Tinbergen measures, not tests, as causality must be fully known.
Keynes was critical of the mathematical modeling approach used by Tinbergen, arguing that it oversimplified the complexities of economic systems and could lead to misleading conclusions if not handled with caution

Question 10

Trygve Haavelmo

Answer 10

Economic variables are interdependent; no controlled experiments.
Economic data =”sample selected by nature”from an unknown data-generating process(DGP).
&
Economic Laws: (general to specific approach)

Question 11

Measurement-Without-Theory Debate

Answer 11

1. NBER (Burns & Mitchell):
   * Inductive, data-driven approach.
   * Focused on descriptive statistics and time-series data. + no mathematical models
2. Cowles Commission (Koopmans):
   * Theory-driven approach.
   * Used a priori mathematical models, simultaneous equations, and hypothesis testing.

Koopmans:
* Without theory, data can be misinterpreted or biased, especially in the presence of simultaneous equations bias (e.g., cause and effect can be confused).
* Regularities observed in data (e.g., business cycle patterns) may not hold in the future (invariance problem) without a theoretical explanation for their persistence.

Vining:
* Collecting and analyzing data is valuable for discovering patterns and generating new hypotheses.
* He criticized the Cowles Commission’s theoretical models as being overly abstract and simplistic, referring to them as a “pretty skinny fellow” compared to the richness of empirical data.

Question 12

Milton Friedman

Answer 12

Lawrence Klein’s large Model vs Carl Christ’s Naive Models:

Friedman’s Criticism of Klein’s Model:
* Too much trial and error, invariance problem.
* Crist’s naive models are natural benchmarks to assess the predictive power of Klein’s model.
*Asserting that a large economic model can predict the effect of a policy change, is an “act of faith” (in regard to the Klein’s model).

Question 13

Marshallian Approach vs Walrasian Approach

Answer 13

Marshallian Approach - partial equilibrium - fewer theoretical assumptions - Friedman

Walrasian Approach - general equilibrium - a priori theoretical models- Tinbergen,

Question 14

Friedman’s Methodology of Positive Economics (1953)

Answer 14

Key Idea: Theories should predict well, not necessarily have realistic assumptions.

In favour of “economic instrumentalism”?
* Whether assumptions are realistic or not does not matter, as long as theory predicts well!

Or opposed to “realisticness”?
* Assumptions should be descriptively correct in all thor detail

Assumptions (Musgrave’s Classification):
* Negligibility: Ignore insignificant factors.
* Domain: Define the conditions under which a theory is valid.
* Heuristic: Simplifying assumptions used to model complex behavior as if it follows a simple rule.

⇒ Marshallian partitioning, with negligibility assumption and domain specifications!

Question 15

Paul Samuelson

Answer 15

Disagreement with Friedman:
* Rejects “F-twist” (assumptions don’t matter if predictions are good).

Descriptivism:
* Scientists describe, not explain.
* Predictions depend on accuracy of descriptions.

Revealed Preference Theory (RPT):
* Operational alternative to ordinal utility theory.
* Critique: RPT also uses unoperationalized assumptions.

Question 16

Karl popper vs logical positivist

Answer 16

Logical Positivism:
Starts with observations → uses induction to find regularities → verifies to find scientific laws.
* Problem: Impossible to conclusively verify scientific laws.

Popper’s Approach:
Starts with problems and conjectures → uses deduction to derive implications → tests until falsified.
* Advantage: Possible to conclusively falsify an implication.
* Knowledge Growth: Science progresses by learning from mistakes.

Question 16

Descriptivism:

Answer 16

Scientists describe, not explain.
Predictions depend on accuracy of descriptions.

Question 17

Popper’s Lingo

Answer 17

Scientific theory should have Falsifiers: Observations that, if true, falsify the theory.
The potential falsifiers of a theory is called empirical content - what should not happen according to the theory.
*More precise predictions = more empirical content.

Question 18

Good theory acc. to Karl popper

Answer 18

*High empirical content (risky).
*Corroborated by empirical data

Question 19

Popper’s Methodology Steps

Answer 19

1. Logical Consistency: Is the theory logically consistent?
2. Falsifiability: Is the theory falsifiable?
3. Scientific Advance:
   * Does it explain existing observations?
   * Does it solve anomalies or problems?
4. Test Predictions:
   * If predictions are correct → theory is corroborated (adopted as a working hypothesis).
   *If predictions are wrong → theory is falsified (search for a better theory).

NO IMMUNIZING STRATEGIES

Question 20

Poppers Critique of Historicism

Answer 20

–>in the view that there are “inexorable laws of historical destiny” (as in fascism and communism).

Human history is a unique historical process.
=/= a closed and stationary system like our solar system!

⇒ An hypothesis about human history cannot be tested, and therefore cannot lead to a “law”.

2. Regularities in social life vary over time and across cultures.
   * Depends on a large number of factors, which cannot be isolated in lab experiments.
   * Predictions may be self fulfilling or self defending.
   These regularities are not really “laws” but rather “trends” or “tendencies”.

Poppers stance:
* Opposes historicism → leads to totalitarianism.
*Advocates for an open society (freedom of thought) and piecemeal social engineering (small, learnable policy steps).

Question 21

The Duhem-Quine Thesis

Answer 21

It is not possible to conclusively falsify a theory!
Because an empirical test always requires background assumptions, which may be wrong!
→ What has been falsified?
the theory?
the empirical set-up?
the masuring instrument?

Question 22

Thomas Kuhn

Answer 22

Descriptive (focuses on how science is practiced, not how it should be)

Normal science - pradigm

Anomalies -Puzzles that cannot be solved within the dominant paradigm.
Initially ignored, but may lead to a paradigm shift if persistent.

Incommensurability:
*Competing paradigms cannot be directly compared.
* Proponents of different paradigms often misunderstand each other.

Question 23

Disciplinary matrix

Answer 23

1 Symbolic generalizations: U(x), F(K,L), …
2 Metaphysical beliefs: People have utility functions,
3 Values; About the research domain, research methods, evaluation criteria, …
4 Paradigms; Standard models such as IS-LM, Akelof’s lemons model, …

Question 24

Incommensurability:

Answer 24

Competing paradigms cannot be directly compared. Proponents of different paradigms often misunderstand each other.

Question 25

Anomalies

Answer 25

(often first ignored or explained away) Puzzles that cannot be solved within the dominant paradigm. Initially ignored, but may lead to a paradigm shift if persistent.