Examination 201209 Flashcards
Question 1)
Fredrikssons artikel om storlek på skolklasser.
a) Explain intuitively how the maximum class size rule may help you in estimating the causal effect of class size.
b) What are the key identifying assumptions?
Klasser ska vara max 30. Är det 30+1 elever kommer klassen splitas till två osv…
Detta skapar variation i klasstorlekar. Några kommer vara stora och andra kommer vara små. Denna variationen är exogen för studenterna oc hderas föräldrar givet att dessa inke kan klaga och byta för enkelt.
Detta skapar en fuzzy RDD.
Efter som de visar att regeln inte följs till punkt och pricka är det en fuzzy design och inte en sharp design.
Assumptions:
* No bunching at the treshhold (ja enligt PPL=
——- Show dencity graph
——- McCrary dencity test
- No discontinuities in the distribution of the forcing variable (?)
- No discontinuities in the predetermined variables at the treshhold. = Characteristics are balanced around the treshold.
- Inte sensitive för olika bandbredder eller polynom.
Then when there is a Fuzyy RDD, it is a IV so also the following assumptions most apply
- Relevant instrument
- Exclusion restriction.
- Exogenius. Cov(z,u) = 0 -> Random assignedn
Question 1c)
The paper by Fredriksson et al. (2013), among other things, reports the results of regressing: (i) teacher experience; (ii) average class size in grades 1-3; and (iii) average class size in grades 4-6, on an indicator for being above any of the thresholds (and relevant control variables).
The Table below reports the results from these regressions (insignifikanta resultat)
What does this evidence tell you about the validity of the research design? In particular, which identifying assumption does the evidence presented in each column shed light on?
This shows that there are no jumps in the pre-.determined variables. If it was the case, the instrument wouldn’t be indipendent.
Enligt papperet är det ”exclusion restriction” som testas här. Samt om det är korrelerat att gå i samma klasstorlek i andra årskurser. Den sista kollumnen är First stage!
”An important question is whether the instrument only af- fects the outcomes via its effect on class size. If this is not the case, we can only causally interpret the RF estimates shown in Table III. To provide evidence on the validity of the exclusion restriction, we examine if districts respond in other ways to the class size rule. Results are presented in columns (1)–(6) of Table IV. In column (1), we examine whether the probability of being assigned to remedial training is affected by the instrument. If schools respond to the instrument, we would expect it to be lower in districts that have surpassed one of the thresholds. We find no such evidence, however. Column (2) examines if the probability of being assigned to an age-integrated class is affected by the instrument. Again, we find no evidence that this is an issue.”
Question 1d)
The Fredriksson et al. paper also reports the result of regressing a child’s cognitive ability at age 13 on mother’s years of schooling. This yields an estimate of 0.069 (standard error: 0.006).
Does this imply that you must control for mother’s years of schooling when estimating the effect of class size using the maximum class size rule? Why or why not?
Han skriver typ att det inte är förvånande att så är fallet och att den i princip inte behöver inkluderas eftersom det är ett random experiment där mothers education bör vara jämt fördelat på båda sidorna, men att han kan göra det för att öka efficiency.
Han säger att variabeln inte kan inkluderas om den inte är balanserad… jag skulle nog hävda att den måste inkluderas då?
Question 2a)
Baker, Gruber, and Milligan (2008) analyze the introduction of highly subsidized, universally accessible child care in Quebec, Canada. What are the key results from this study on child outcomes?
De studerar introduktionen (expansionen?)av child care i qubec. De ser att detta leder till dåliga utfall hos barnen välmående men inte har någoin effekt på deras koginiva förmågor.
Barnen var unga.
Genom att använda en DiD visar de att barnen som påverkades av policyn bvar mindre sannolika att ha bra fysiologisk hälsa (de var sjuka mer tror jag) och att de var mer hyperaktiva och mer oroliga.
Question 2b)
Havnes and Mogstad (2011) examine a reform from late 1975 in Norway, which led to a large scale expansion of subsidized child care; see figure 1 below.
Describe how Havnes and Mogstad use this expansion to examine the effects of childcare on the children’s outcomes. What is the identifying assumption?
Blabla. Den här kan jag .
Assumption är att de utan så kraftig expansion skulle ha utvecklats i samma takt som de med kraftig expansion.
Question 2c)
Havnes and Mogstad (2011) and Baker, Gruber, and Milligans (2008) conclusions about the effects of childcare on children’s outcomes differ. Which are the key potential reasons according Havnes and Mogstad?
Kontrasten äör att Havers och mogstad hittr en positiv kong run effekt.
Formal informal vs parents….
Cost of socialisation… Separation anexiety.
De fick mer sjukdommar pga mer interaktioner.
Maternal labor supply.
Question 3b)
Discuss whether it is useful to answer both positive and normative questions.
Positive questions aim to describe the underlying mechanisms of agent behavior and are thus usefull to use to provide explanations for the results obtained from normative questions. However, the advantage of the SS method is that positive questioins do not have do be answered prior to normative ones.
Question 3 c)
Explain, in as much details as you can, one application of the Sufficient Statistics. Provide the final formula.
One application relates to the welfare effect of a tax increse on som good X1. Then oaccording tothe method of Sufficient statistics we can just look at how the EQ quantity changes du to the tax….. Slut på papper…
Question 3d)
Discuss how the Sufficient Statistics may deal with situations where agent is not perfectly optimizing her utility, i.e. there exists some irrational behaviors.
In the structural framwork it is rather difficult to identify the undergoping mecanisms and variables and it can lead to many interpretations resultingnin different predictions of a single policy. The SS method is then helpfull in the sense that it avoids those deep parameters but can still be used to evaluate the effects of policys.
”The sufficient-statistic approach can be applied even when one is uncertain about the positive model that generates observed behavior, such as in recent studies in the behavioral economics literature that document deviations from perfect rationality. In such cases, welfare analysis based on a structural model may be impossible, whereas the more agnostic sufficient-statistic approach permits some progress. For instance, Chetty et al. (2009) derived formulas for the deadweight cost of taxation in terms of price and tax elasticities in a model in which agents make arbitrary optimization errors with respect to taxes.”
Question 4a)
Grönqvist estimerar den här ekvationen för försäljningen av p-piller:
Sales_ct = βSubsidy_ct + μ_c + μ_t + ρ(μ_c x t) + ε_ct
Tolka β och berätta vad den identifierande variationen är som Grönqvist använder?
B fångar effekten of the subsidy on sales in county C at time T
The identifying assumnption is that there are no shocks present at the conty-level that occour at the same time as the subsidy policy change that affefts the relative outcomes betweennthe cohorts.
The variation used for this study is at the regional-level, cohort-level and across time as the policy was implemented acrtoss differentr conties over time and had a different targeted group in terms of age.
Question 4 b)
Grönqvist (2012) also runs a regression which explores the relationship between future subsidies and current sales. What is his motivation for estimating that model?
Describe how a negative, zero, and positive estimate of a future subsidy on current sales should be interpreted and how these three scenarios (respectively) affect the interpretation of β in the equation above. (10)
The regression is to ensure that sales are not increasing absent of the subsidy introduction.
If zero effect: Reasuring. B should have a causal interpretation.
If negative: Could suggest a negativ trend. This would be consistent with the fact tat oral conceptives wernt subsidized and that therefpr decrsed sakles could bias B.
If positive could suggest a positive trende. SOme other shock. Bias
Question 4 c)
Assume that Grönqvist had access to price information of the pills for the relevant regions and cohorts, and then proceeded with an instrumental variable (IV) strategy using the subsidies as an instrument for pill prices. According to Grönqvist, would such an IV-estimate solely capture the impact of pill prices on e.g. teenage pregnancy? Why/why not?
Grönqvist claims that is is ambiguous how prices of contraceptives affects tenage sexualk behavior. Those who doint discount the future to mush would probably get the conterceptivrs regardless or abstain, avoiding pregnancy.
On the other hand, cheaper contraceptives could indice more sexual activity, ultimatly increasing pregnancy risk. Thus it hinges on whether contraceptives are price elastic or not.
” is not obvious that the demand for contraception is price elastic. Women who consider the cost of pregnancy as very high may either choose to completely abstain from sex or always pay the price of getting the pill. In the last case subsidizing the pill will not lead to a behavioral response but only result in an income transfer from the government to the individual. ”’
”having access to inexpensive contraception could mean that women raise their level of sexual activity, increasing the likelihood of a pregnancy. This makes the net effect on fertility ambiguous. A lower relative price for the pill can also affect abortions if women substitute between the pill and other less efficient contraceptive technologies in order to avoid unwanted births.”
