14

Dev2 (Subject) / Chapter 6 (Lesson)

This lesson was created by Dessinha.

Learn lesson

This lesson is not released for learning.

back  |  next 1 / 1

• 1.) Human Capital Theory (Becker) - Education / Human Capital? 2.) Wage in perfect competitive markets 3.) Unemployment in perfect markets? 1.)Education as an investment in human capital. The costs of investment are both direct (school fees) and indirect (foregone earnings). Returns emanate from increased labor productivity 2.) W/P = MPL (Workers are paid marginal product of labour) 3.)Perfect market framework assumes all unemployment is voluntary. In reality, education may help you get a job (or not).  
• 1.) Labour demand - competitive -skilled/unskilled labour 2.) Rate of return to Education? What do we want to measure? 1.) Firms pay more for skilled labor because it's more productive. In a perfectly competitive market with price-taking, prot maximizing firms, the wages of skilled (s) and unskilled (u) labor are equal to their respective marginal products. 2.)Marginal increase in earnings due to additional year of schooling. This is the private, pecuniary return to schooling - ignoring social externalities, etc.
• 1.) How to measure - problems with "constructed" data? 2.) Difference in measurement between perfect world and unperfect - observed characteristics - formula 1.) We need a counterfactual (what would you have earned if you hadn't enrolled in program)- a) before and after comparison b) cross sectional comparison! Individuals differ in observed and unobserved ways - with survey data we control for observeable differences. The unobserveable could lead to OVB so the estimated effect wouldn't be causal. 2.) ln yi = a + bSi (perfect world: b indicates the percentage increase in earnings from an additional year of schooling)    ln yi = a(0) + a(1)X(i)' + bSi (X(i)' : observable characteristics that effect earnings other than education, we control for them this way)  
• 1.) How to deal with unobserveables? One possibility IV, which conditions to satisfy in case of education? 2.) Other ways besides IV? 1.)Decision to go to school might be an unobserved characteristics between those, that did go and the others.The IV  must be (partially) correlated with education.  It must be `excludable', that is, not correlated with earnings after controlling for education. 2.) Natural Experiment , Randomization
• 1.) Duflo (2001): Estimate rates of return based .... 2.) How? Data? 3.) Results: 1.) on interaction of education and policy changes 2.) School building programme in Indonesia: Between 1973-4 (and 78-9), 61,000 schools were built    Men born between 1950-1972 from 1995 census: birth date, salary level and level of education in 1995   3.)Enrolment rate increased from 69% to 83% between 1973 and 1978 Number of schools built in each region was inversely related to number of children not enrolled in 1972. On average, estimates indicate that programme increased years of education between .25-.4 years. Economic returns to education range from 6.8-10.6 %  
• 1.) Variation in intensity of program for given individual: 1.) Variation in number of schools built in each region. Cohort of child: the younger the child (<12) in 1972, the more it would potentially benet from program.  Dufflo argues that conditional on district of birth and cohort fixed effect, -being born in a high intensity region and cohort- is as good as randomly assigned.
• 1.) Use variation in education predicted by programme to .... 2.) How does programme "work" / affect returns ? 3.) Instrument: 1st stage - 2nd stage 1.) estimate returns to schooling 2.) Evolution of wages and education across cohorts would not have varied systematically from one region to another in absence of programme (strong assumption. Programme aected wages only through increasing attainment (and not directly). 3.) 1st: Educational level of person i in region j and cohort k 2nd: wages/outcome (while result of 1st stage is independent variable)  
• 1.) Facts about enrollment: Pre-primary, primary, Secondary 1.) Pre-primary enrolment: 18 countries with less than 10% of children enrolled in 2007. 55 with less than half the children enrolled in 2007   Primary: 86.9 % of primary age children enrolled worldwide. All regions except Sub-Saharan Africa and South Asia have net enrolment rates between 90 and 95%   Secondary: 58.8% enrolled globally in 2007, increase of 5.6% since 2001. Only about a quarter of children enrolled in secondary school in SSA. ---> slightly improvement even in SSA    
• 1.) Facts about achievement and quality: 2.) What about men and women concerning literacy? 3.) How has been this data collected? What would be better? 1.) Adult literacy rates improved from 76.2% in 1990 to 83.8% 2007 world wide. Global primary completion rate is 87 %. SSA went from 50 to 68%. Primary pupil teacher ratios are 30 on average. Highest in Sub-Saharan Africa and South Asia, went from 40 to 50 in between 1985 and 2004. 2.) More adult men literate than women in all regions 3.) Data are mainly based on asking adults about their reading ability. Yes/No question. Better method: assess reading ability (and numeracy etc.) directly.   [Literacy Assessment and Monitoring Programme (UNESCO) currently being piloted. Combination of household survey and educational assessment]
• 1.) Bold, Kimenyi, Mwabu and Sandefur (2010) about literacy of school children in Kenya (using monitoring stuff) - results: 2.)Bold, Gauthier, Svensson, Wale (2011) for Tanzania& Senegal (4th grade) 1.) Based on a one-on-one reading test, 79% of children in Standard 1 cannot read a single word in a sentence. Even in Standard 2, 57% of the sampled children who took the tests in literacy were not able to read a single word in a passage. 2.) 33 % of Senegalese fourth graders can read a sentence. 6 % of Tanzanian fourth graders can read a sentence.
• Advertisement
• 1.) Deininger (2003): "Does cost of schooling affect enrolment by the poor? Evidence from Uganda 2.) Bold, Kimenyi, Mwabu and Sandefur (2011): "Does Abolishing Fees reduce School Quality? Evidence from Kenya 1.) Uganda abolished fees for primary schools in 1997. Enrolment in primary schools increased from 62% to 83.6%. Enrolment of girls nearly doubled. Parental income is less important in determining enrolment choices following the reform. BUT: Trade-o between quantity and quality. 2.)Infux of poorer students into public primary schools. Flight of richer to the private sector. Very small increase in net enrolment in the public sector. Private sector tripled in size. demand eect as evidence of decline in educational quality. Find that decline in educational quality can be entirely explained by change in composition of students.
• 1.) Hypothetical Link: FPE to school quality - 3 channels linking fpe to students performance 1.) Increased funding (increased gov't funding, net effect on total funds ambiguous due to abolition of local fundraising) Changes in the pool of students (Fees abolished - more students can access education - these might be different from others in socio-economic background, ability etc..) Increased centralization -loss of local accountability (Parents no longer pay for the school, so may lose sense of ownership)
• 1.) Effect of abolished fees on private/public schools? 2.) Why did the rich leave? How to find out by method? 3.) Change in cost for private/public? 1.) Private: more enrolled , public less than before, in total more students enrolled  2.) Net nancial resources declined, pupil-teacher ratios increased. The socio-economic composition of peers became less attractive --> Estimate social multiplier in the enrolment decision 3.)Under FPE, fees for public primary have fallen to zero, and total costs nearly halved. Cost of private education has more than doubled, now more than 20 times more expensive than public ed.
• 1.)Kremer and Miguel (2001) "Worms: Education and Health Externalities in Kenya": what? 2.) Results 3.) Method 1.)Examine impact of deworming medicine on school attainment (and achievement) on basis of randomized controlled trial in Western Kenya. 2.) Reduced school absenteeism in treatment schools by 7%, which is a 1/4 reduction. No eect on test scores. Document large externalities from treated to untreated children both within and across schools. Among untreated (within school) dierence between infection rates is 21%. Among treated 27%. Being located within 3 kilometers of treatment school reduces infection rate by 12%. 3.) 2 Treatment groups (from 1998 deworming and from 1999 deworming) and 1 control group Control variables to control for differences within and across schools

back  |  next 1 / 1