Assignment 10

In this chapter, statistics are primarily used as a tool to analyze social behaviors as they pertain to MFIs.  For instance, one of the more intriguing stats at the end of the chapter is how a 10% increase in loans leads to 9% increase in profits for the loaner.  However, they also say that this requires the loans to be paid off and said that due to several events, 70% of the loans remained unpaid.  They claimed that these stats were caused by people losing confidence in paying the lenders due to things such as local newspapers leading to a complete lack of trust.  The authors made the claim that since no one else was paying, the entire system essentially collapsed.

This certainly makes sense and does have some evidence thanks to the statistics.  However, one of the issues I had with the piece is that since behaviors were so easily influenced by other people, I have to question why it’s so difficult to restore confidence in the lenders.  According to the book, confidence was mainly shattered through bogus reports in local papers and that was enough to completely break the trust of the population.  So why does the book make it seem like its nigh impossible to restore confidence?  If the population is so easily swayed one way, why can’t they just find some way to restore confidence by using local papers, politicians, or even grassroots organizations to improve confidence?

The article “The Evolution of the Returns to Education for 21- to 35-Year-Olds in Canada and across Provinces:”  provides an overview of how returns to education have shifted from 1991 to 2006.  The paper runs through a series of models and ultimately comes to the conclusion that returns have increased drastically over the years for nearly all groups, with high school educations having the most prominent increase.

1)  This paper does not use a linear regression as I have and instead uses a log in order to estimate a variety of educational categories.  It also categorizes education into more categories than I did.

2)  I know my regression has several issues with it, notably some multicollinearity or heteroskedacity  (probably both).  I didn’t think of taking the log of wages before regressing so I’ll experiment with that  That’s as easy as making a variable for log(wages) and then replacing my previous variable with that one.

Assignment 8

For my regression, I regressed a person’s average wage vs a variety of variables including education (Separated into high school graduate, no high school, and a PHD), race, whether they were unionized, and their marital status.  The first big issue was that my r^2 was almost 1 at 9.9 which means I’ve run into an issue somewhere since I doubt my regression is that on the nose.  In addition, I had several insignificant variables, primarily the union and marriage ones so I’ll likely remove those from the final model.  Another oddity I noticed was that in regard to the education variables (the ones I’m interested in) only being a HSG had a positive coefficent while the others were negative.  While this makes sense for a lack of education, its more dubious with the phd.  I intend on adding another college level variable to see if that changes everything, maybe just a bachelors.  

Assignment 7

Looking at the article here, I noticed that both the author and Poor Economics offer somewhat different ideas regarding the effects of education compared to one’s future prospects.  For instance, PE claims that ultimately, years of education offer proportional returns of education while the article claims that the scale is more exponential.  I personally find the article in the Times more convincing regarding America as a whole while PE better regarding third world countries.  In America, where quality education is in relatively large supply and jobs have a higher demand of education, I can understand returns being more exponential while the inverse is true in the third world.  Since education is more dubious there and jobs on the whole less demanding, it stands to reason that education returns are more linear.  Neither side uses a great deal of statistics in that regard and would be better off including something more definitive.  However, in terms of logical discourse, I find merit in both sides of the argument depending on the location of the education in question.

Note: I was unable to leave comments on the website



Bourbeau, Emmanuelle, Pierre Lefebvre, and Philip Merrigan. “The Evolution of the Returns to Education for 21- to 35-Year-Olds in Canada and Across Provinces: Results from the 1991-2006 Analytical Census Files.” Canadian Public Policy 38.4 (2012): 531-49. Print.

Carneiro, Pedro, James J. Heckman, and Edward J. Vytlacil. “Estimating Marginal Returns to Education.” American Economic Review 101.6 (2011): 2754-81. Print.

Fasih, Tazeen, et al. Heterogeneous Returns to Education in the Labor Market. The World Bank, Policy Research Working Paper Series: 6170, 2012. Print.

Jensen, Robert. “The (Perceived) Returns to Education and the Demand for Schooling.” Quarterly Journal of Economics 125.2 (2010): 515-48. Print.

Long, Mark C. “Changes in the Returns to Education and College Quality.” Economics of Education Review 29.3 (2010): 338-47. Print.

Park, Seonyoung. “Returning to School for Higher Returns.” Economics of Education Review 30.6 (2011): 1215-28. Print.

Stanek, Kevin C., William G. Iacono, and Matt McGue. “Returns to Education: What do Twin Studies Control?” Twin Research & Human Genetics 14.6 (2011): 509-15. Print.

Yakusheva, Olga. “Return to College Education Revisited: Is Relevance Relevant?” Economics of Education Review 29.6 (2010): 1125-42. Print.


Introduction:  Returns of education has been a topic that has been widely discussed throughout economics and has never gotten a clear answer on what its effects are.  The problems with calculating any result is that there are so many variables that ultimately do not behave in an objective manner.  A person can have all the advantages he can ask for, with a stellar education and connections, but he can still end up not getting a job befitting his status due to a myriad of other factors such as current openings, prejudiced managers, or similar.  Outlying results like that (or the inverse) can ultimately skew the data on the returns of education which makes the true results difficult to pin down.  While compensating for these outliers is virtually impossible, I will simply have to ignore them for the purposes of this paper.  Instead, I will analyze the number of years a person was at school, his race, his economic background, and compare it to the wage he ultimately earns as well as the cost of his education.  I expect there to be a relatively strong, positive correlation between the amount of time a person spends on his education and his wage.  However, I also expect that past a certain point, the person will no longer receive additional wages from his education as diminishing returns start to step in as well as an increasing dependency on connections rather than solely his degree.

Assignment 5

The thesis of this chapter is that data is often misleading and finding the correct statistics of a given topic (such as this chapter’s focus on drug dealing) requires an extreme amount of work and can lead to unforeseen conclusions based on previous conventional wisdom.

Statistics: Conventional wisdom about the homeless (86), Atlanta’s missing police reports (88), Drug revenues/wages (97-100), Stockings (107)

These statistics are important because they essentially provide the grounding that makes the story believable.  The first two are there in order to demonstrate how easily statistics can simply be made up or skewed in order to get a result that is palatable for the public or whatever other enterprise they’re trying to convince.  Now that the reader is good and skeptical about soft or conventional statistics the story begins proper about JT and the drug dealers.  The author goes through great lengths to confirm that the sources (the notebook for instance) are essentially from the horse’s mouth and thus trustworthy (as far as drug reports can be) and from people who are authorities on the subject.  Finally they conclude the story with the stocking anecdote giving precedence as to why such a profession might be desirable despite the appalling conditions and set backs.  All in all, such a layout goes a long way to making the story work as something other than an extended anecdote and the first three stats help this wonderfully.  The stockings are less vital, since the logical discourse earlier is perfectly adequate, but it doesn’t take  anything away from the chapter since more evidence is never a bad thing.


Assignment 4

My topic is essentially what exactly are the returns to education: in other words, I want to see exactly how much extra pay someone would earn depending on many years of school they went through.  My motivation going in was essentially that of curiosity.  There’s a lot of money that goes into a college education and knowing whether or not such an education actually gives someone a decent advantage is quite a fascinating topic.  Hopefully, I’ll find a decent set of statistics that can lead to some lasting conclusions.  While I understand that the topic is fairly complex with a lot of factors, I intend on just ignoring the personal factors an individual would have and instead focus on the overall figures.  To that end I’ve already found a good set of numbers that categorizes people’s pay based on how many years of education they’ve had.  I’m currently looking for the average cost of extended education such as college and graduate school but so far the results haven’t been to my liking.