Sunday, August 9, 2015

Peru’s mining mita

Reading the notable book from Acemoglu and Robinson, "Why nations fail" I came across the concept of mita and the effect of it on economic growth. Mita was an extensive forced mining labor system in effect by the Spanish Empire in some regions of today´s Peru and Bolivia between 1573 and 1812. Fascinatingly, a paper wrote in 2011 by Melissa Dell (or here) found that:
Regression discontinuity results indicate that a mita effects lowers household consumption by 25% and increases the prevalence of stunted growth in children by around 6% points. Mita’s influence has persisted through its impacts on land tenure and public goods provision. Today, those regions are less integrated into road networks and their residents are substantially more likely to be subsistence farmers and with lower educational attainment.
Undoubtedly, it’s been known that historical institutions are key on current economic outcomes. This paper not only confirms this statement but finds the routes through which those institutions can have this persistent effect on growth. Aka public goods provision and land tenure. The key element of this research is that the mita required indigenous communities to send a constant share of their adult population to work in the silver and mercury mines and conscripts changed discretely at the boundary of the subjected region: on one side, all communities sent the same percentage of their population, while on the other side, all communities were exempt. This is probably an arbitrary and short informed decision of the regional government ef the Spanish Empire that allows the researcher to use as an external shock.

Friday, July 31, 2015

Machine learning visual experiment


R2D3 is an experiment in expressing statistical thinking or machine learning with interactive visual design. In machine learning, computers apply statistical learning techniques to automatically identify patterns in data. These techniques can be used to make highly accurate predictions.
In this case the authors Stephanie and Tony design  a machine learning process in order to know if a house with unknown location is from San Francisco or New York based on other parameters such us altitude, price, sq feet,…



In most cases they seem to apply a logistic regression where high or low values of one variable such as high altitude are more likely to be from one city rather than the other. By sequentially applying this process with all variables and using recorded data the authors can get a very accurate estimate of location for any given house, provided the explanatory variables are available.

Saturday, July 4, 2015

Coloured time series

Here comes an idea (I saw it in Flowingdata.com) to graph many time series together in a well-designed and simple way.

The thing is that when you try to put many graphs together, the Y-axis becomes too short and changes are really hard to appreciate.


These kind of graphs can be converted into colour graphs which are much easier to understand. The Y-axis is fixed to a given level for all graphs. If one of the time series goes above that level, a darker area start at that point from the Y=0. Negative numbers can be drawn as reddish.

It is much clearer when a time serie is up or down by looking at colours.




Monday, June 15, 2015

Determinants of Trade in Parts and Components

There is a broad agreement that growth in world trade has outpaced the growth in global output in recent decades but the key aspect in global trade is production fragmentation. The production process has been split in an increasing number of micro-processes that rises the interaction between the different members internationally, giving rise to a sharp increase in global trade.

The question is, What factors determine trade in parts and components? This is the focus of a recent study.

Theoretically, each stage of the production process should be allocated following the comparative advantage of each country.

Empirically, the author selected parts and components trade data for 26 countries and from 1990 to 2010. From their regression results, they found that trade of parts and components are positively and significantly related to the size of the economy as well as the quality of labour. Second, infrastructure and institutional qualities matter. Third, trade in parts and components are at least somewhat different from trade in all goods. Trade in parts and components are much more sensitive to the quality of labour. In addition, the extent of the production network is responsive to a narrower subsets of measures of quality of infrastructure/institutions.


Once again human capital and institutions are the key to trade and output growth.

Tuesday, May 26, 2015

Forecasting

I found quite amusing to discover several papers about forecast performance and the ability to estimate future events.

It has repeatedly said that forecasting shouldn’t be an economist job. Future is unforeseeable for
everyone and none of the economist tools can help there.  Regardless of what is said, many of us
can’t stay away from forecasting. Two recent papers have been analysing the forecasting power of different people.

http://www.apa.org/pubs/journals/releases/xap-0000040.pdf

http://www.cis.upenn.edu/~ungar/papers/forecast_AAAI_MAGG.pdf

Aparently being open-minded, a knowledge on probabilistics, interacting with others and discuss your forecasts with other analysts increases significatively your chances of being right.

Saturday, May 16, 2015

Stata in Google Scholar

In the academic world the use  of analytical/statistical tools is essential to do research. The data analysis and empirical research has grown exponentially in academic publications as PCs became universally affordable, -- there is an interesting graph showing the growth of empirical research in the economic literature (vs theoretical) that unfortunately I couldn’t find.


The most widely used of those statistical tools is R, as the graph below shows. However, Stata is surprisingly matching R. Strangely enough Stata is not that ubiquitous outside the academic world.




Source: http://www.r-bloggers.com/statas-academic-growth-nearly-as-fast-as-rs/

Wednesday, April 15, 2015

Statistics done wrong

Statistics Done Wrong (a website and a book) is a very useful guide to the most common statistical errors explained in a simple and strightforward way.

"Statistics Done Wrong is a guide to the most popular statistical errors and slip-ups committed by scientists every day, in the lab and in peer-reviewed journals. Many of the errors are prevalent in vast swaths of the published literature, casting doubt on the findings of thousands of papers. Statistics Done Wrong assumes no prior knowledge of statistics, so you can read it before your first statistics course or after thirty years of scientific practice."

http://www.statisticsdonewrong.com/index.html