Another potential concern is that with only 440 observations in most

Another potential concern is that with only 440

This preview shows page 14 - 16 out of 136 pages.

Another potential concern is that, with only 440 observations in most specifications and thirteen regions, the clusters used for the standard errors may be too small. I have re-defined alternative clusters corresponding roughly to the United Nations’ division of Africa into regions – Southern Africa (African Hunters, South African Bantu), Western Africa (Guinea Coast, Western Sudan, Nigerian Plateau, Moslem Sudan), Central Africa (Central Bantu, Equatorial Bantu, Eastern Sudan) and Eastern Africa (Northeast Bantu, Upper Nile, Ethiopia/Horn, Indian Ocean). Re-estimating the main results, I show in Table 7 that the results are now more statistically robust. The coefficient estimate falls less with the addition of these broader fixed effects than with fixed effects added for the regions as defined in the Ethnographic Atlas . Finally, the inclusion of the major river dummy and distances from the coast, Lake Victoria, and slave trade ports may be capturing elements of trade based on features other than ecological diversity. Similarly, while the inclusion of the date of observation is intended to control for both remoteness and the possible impacts of European in- fluence, it may be endogenous to state centralization. I show in Table 7 that excluding these variables barely affects the results. It is also possible that the results here are driven by outliers. In Table 8, I control for this possibility by dropping influential observations from the sample. I estimate the main results by OLS with the full set of controls and without fixed effects. I then com- pute the leverage and dfbeta (for ecological diversity) statistics for each observation. In column 1 of Table 8, I drop all observations with leverage greater than 2( df + 2) /N . In column 2, I remove any observations with absolute dfbeta greater than 2 / N . The main result survives both of these procedures, though the former is marginally insignificant when fixed effects are included. It is also possible that the results are spuriously iden- tified by variation within a single African region. In columns 3 through 6, I drop each of the “South African bantu,” “Ethiopia/horn,” ‘Moslem sudan” and “Indian Ocean” in turn. These are the regions in which most states are concentrated. The results again are robust to each of these, though some are again marginally insignificant with fixed
Image of page 14
ECOLOGY, TRADE AND STATES IN PRE-COLONIAL AFRICA 15 T ABLE 9. Robustness: Reverse causation [Table 9 here] effects. It is not the contrast between a handful of states and their neighbors that is driving the results. 4.4. Possible reverse causation. It is also possible that stronger states shape the en- vironment; McCann (1999) describes, for example, the careful regulation of forest re- sources in Ethiopia before the twentieth century. To control for this possible reverse causation, I employ the standard deviation of temperature within an ethnic group’s ter- ritory as an instrument for its Ecological diversity. This is intended to capture varia-
Image of page 15
Image of page 16

You've reached the end of your free preview.

Want to read all 136 pages?

  • Spring '17
  • JAMES FENSKE

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern

Stuck? We have tutors online 24/7 who can help you get unstuck.
A+ icon
Ask Expert Tutors You can ask You can ask ( soon) You can ask (will expire )
Answers in as fast as 15 minutes
A+ icon
Ask Expert Tutors