Lesson 1 Flashcards
(24 cards)
definition public economics
analyzing policies and form guidelines for the government
libertarian economist vs interventionist
libertarian= the market decides what to produce (price/quality devision). Those who want pay will receive. –> closest to our standar way of thinking (perfectly competitive market)
Interventionist
- government can decide (and also intervene)
- if something goes wrong, they can fix it
exceptions to intervene in the market
- market failures (when a market is working inefficient)
- societal preferences (government wanting something different then what result would be (no use of drugs))
- justice/fairness (social security, bijstand etc. people are allowed to get atleast a bit of money)
Positive vs. normative analysis
positive= facts, effects, “what is”
- theory, empirical studies
normative= values, preferences, “what ought to be”
- getting the most efficient and welfare outcome
3 types of empirical studies
- experimental
= the ideal study
pros: randomization, control and clear what changes
cons: funding, ethical issues, less participation, difficult to generalize - observational
= observing and measuring actual behavior outside of an experiment and trying to “back-out” effects.
cons: existing data doesn’t include every variable, other correlations must be eliminated
- most common study - quasi-experimental (looks like an experiment, but doesn’t have randomization –> naturally occurring instance of observable phenomena)
attrition
the loss of study units from a sample
- only a problem if it is related to something that matters for the outcome
Hawthorne effect
people behave differently in experiments than they would in real life
external vs internal validity
external= validity of generalized inferences in scientific studies
internal= validity of inferences in scientific studies
dependent vs. independent variable
dependent= observed result of change
independent variable= factors that can affect this result
3 types of observational data
- cross-sectional data (compared different units in the same year
- problem- compares across people, and people differ - time-series data (same unit over time
- requires something to change
- don’t know how the future looks like, maybe this is already where they are going to (no control group) - panel data/longitudinal= follows units over time (many people, many years)
- able to compare across groups and within individual over Tim
problem= more expensive to collect
meaning counterfactual
what would have happened to the treated in the absence of treatment
causal vs. corrleation
- the cause must precede the effect
- the cause and effect mus be correlated
- other explanations for observed correlation must be eliminated –> this makes it a causal relationship
3 examples of correlations that might be interpreted as causal, but could just be correlation
- third factor explaining both
- reverse causation
- selection (people selected for treatment could differ, even before study already)
meaning spurious correlations
occurs when two variables are statistically related but not directly causally related. These two variables falsely appear to be related to each other, normally due to an unseen, third factor
- example: murders and ice cream bought have same patterns
3 types of quasi-experimental studies
- Differences in differences
- Regression discontinuity
- Instrumental variables
differences in differences meaning
uses 2 aspects at the same time: before & after experiment, but also 2 different locations/groups
Raw DiD= (D-B)-(C-A) –> Washington before&after (treatment) - Orlando before&after (control)
assumption: both places would have behaved the same in absence of the policy
- the further away the pre-policy levels, the harder it is to say the are the same
- nothing else can happen at the exact same time differently affecting the groups
Regression Discontinuity (RD)
= treatment is determined by measurable characteristics and a requirement to be above/below some cutoff
- example: scholarship funds above GPA above 3.5 –> most clear difference close to cut-off
things to consider
- requires a lot of the data (only use close to cutoff values)
- mush show that other variables are smooth
- control for control/treatment, but now also the cutoff variable
Instrumental variable (IV)
= a third variable that affect entry into treatment, but doesn’t affect the outcome otherwise
Use a variable — called an instrument — that:
- Affects the treatment (influences education)
- Does not directly affect the outcome, except through the treatment
- Is as good as randomly assigned
2 conditions for it to work:
- instrument correlated with treatment variable, not outcome
- instrument can’t suffer the same problem as original predicting (x) variable
- can only impact y variable, trough x variable
difficult because
- 2 things studying a probably very related
- even if done correctly: get the effect only for the subgroup affected by the instrument
pros & cons of quasi experiment (vs experiment)
- great internal validity –> external difficult, but quasi makes is closer to desired situation
- can get a causal effect
–> but expensive and hard to get approval of - no say in designing the set-up, difficult to find situations that mimic an experimental study
- national policy difficult to stay (all at once)
- generalizability depends on the natural experiment
meaning welfare economics
= the branch of economic theory concerned with the social desirability of alternative economic states
efficiency
- Edgeworth box
- pure exchange economy
- 2 people, 2 goods
Meaning Wedgeworth box
= a graphical tool used to analyze efficient allocations of two goods between two individuals. It shows:
- All possible ways to divide a fixed amount of two goods.
- Each point in the box represents one possible allocation.
- Indifference curves for each person show preferences.
- The set of Pareto efficient allocations lies along the contract curve, where the two individuals’ indifference curves are tangent.
meaning indifference curve
= shows all combinations of two goods that give a person the same level of satisfaction. The person is indifferent between any bundles on the same curve.
Downward sloping= More of one good means less of the other to maintain the same utility.
- to the left means less clothes but more food
- to the right means more clothes but less food
Pareto improvement vs. efficient
improvement = change the makes one person better of without making anyone else worse off
–> the part where the indifference curve of tom and Alice intertwine: they both have to get less of what they want, but then they are together better off (middle part)
–> when the curves cross, but are nog tangent
efficient= an allocation where you can’t make anyone better of without making another worse off –> when the curves are tangent
meaning contract curve
= combination of all Pareto efficient points (all different allocations from tom and Alice)
