EMET2007 Lecture 9 for Wattle

graph was for nwifeinc 50 exper 5 age 30 kindslt 6

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ation of married women (cont.) Graph was for nwifeinc = 50, exper = 5, age = 30, kindslt 6 = 1, kidsge 6 = 0 The maximum level of education in the sample is educ = 17. For the given case, this leads to a predicted probability to be in the labour force of about 50%. Negative predicted probability for educ < 3.84 But not a problem in this case because no woman in the sample has educ < 5. Lecture 9 (qualitative variables) EMET2007/6007 8 st May 2013 45 / 53 Disadvantages of the linear probability model Predicted probabilities may be larger than one or smaller than zero Marginal probability e¤ects sometimes logically impossible The linear probability model is necessarily heteroscedastic Variance of Bernoulli variable Heteroscedasticity consistent standard errors need to be computed Advantanges of the linear probability model Easy estimation and interpretation Estimated e¤ects and predictions often reasonably good in practice Lecture 9 (qualitative variables) EMET2007/6007 8 st May 2013 46 / 53 More on policy analysis and program evaluation Example: E¤ect of job training grants on worker productivity The dependent variable is: scrap = Percentage of defective items The explanatory variables include: grant = 1 if …rm received training grant, = 0 otherwise sales = The level of the …rms sales employ = The number of employees Lecture 9 (qualitative variables) EMET2007/6007 8 st May 2013 47 / 53 Actual estimated model No apparent e¤ect of grant on productivity with tgrant = Lecture 9 (qualitative variables) 0.52 = 0.431 EMET2007/6007 1.207 8 st May 2013 48 / 53 Treatment group: grant receivers Control group: …rms that received no grant Grants were given on a …rst-come, …rst-served basis. This is not the same as giving them out randomly. It might be the case that …rms with less productive workers saw an opportunity to improve productivity and applied …rst. Lecture 9 (qualitative variables) EMET2007/6007 8 st May 2013 49 / 53 Self-selection into treatment as a source for endogeneity In the given and in related examples, the treatment status is probably related to other characteristics that also in‡uence the outcome The reason is that subjects self-select themselves into treatment depending on their individual characteristics and prospects Experimental evaluation In experiments, assignment to treatment is random In this case, causal e¤ects can be inferred using a simple regression The dummy indicating whether or not there was treatment is unr...
View Full Document

Ask a homework question - tutors are online