12/20/09

A Preview of my Analysis on Who Engages in Land Grabbing and Why

I am doing research for a paper that will try to unpack the  mechanism behind  land deals and what factors might determine whether such a deal will take place.

The reason why I would look into this issue is quite evident. The aborted land deal with Daewoo in Madagascar was identified by many as one of the major factor in the current ongoing crisis (1). I will not write any further on the crisis today because it is Christmas time and I need a break from the continuing appalling decisions ( army officer Vital Camille was announced as the 3rd PM in 3 months) by the current leaders over there.

Madagascar is not clearly not the only who got involved in land deals. According to this study,
50 countries were targeted by maybe 1,000 investment groups. World Bank says 50m ha – nearly half the cropland of China –are signed away or under negotiation in Africa, Asia, LatAm since 2006. FAO says 20m ha in Africa alone. GRAIN estimates that US$100 billion have already been mobilised to pay for these deals. (World Bank says US$50 billion.)

The variables I used originated from this International Food Policy Research Institute (IFPRI) report.

1) numbers of reported land deals around the world and how many times a country engages in such deals
2) the level of democracy (based on polity index) in the countries trying to strike a deal.
3) the difference in Gross Domestic Product (GDP per capita) in the countries trying to strike a deal (source IMF).
(more..)


The report is not finished yet but I figure I will give you a preview of the regression analysis I have done so far:
(GDP difference)                                                           (Polity Index and Number of deals)




As you can see from the graph, early returns suggest that there might be a correlation between countries engaging in land deals and:
1) the difference in GDP (USD) between sellers and buyers. The bigger the difference, the more likely it is that a land deal would happen.
2) The less open a regime is, the more likely it is that they will go through with a land deal.

 This study is clearly not refined enough yet but I think an empirical study on whether land deals are really as lopsided as advertised (NYT) was necessary. I would welcome suggestions on what parameters or dummy variables I should be using to complete my analysis. will provide the full set of data I used after I submit the final version.

================================
Footnote:

(1) Joachim von Braun of IFPRI came to Princeton in October to discuss land grabbing. He provided insights on what land deal entails and recommendations on how they should be done (in brief: Transparency in negotiations, Respect for existing land rights, Sharing of benefits, Environment stainability, Adherence to trade Policy and if drought,no deal.) He also argued that it is a shame that a putative land deal that was not close to be settled provoked such a crisis. Land deals, he says, are not all bad but they must be conducted properly. He said that there was too much media hoopla (I am guilty as charged)  on this deal that was still ongoing. He wish the Ravalomanana government would have consulted with IFPRI while negotiating and be more transparent.
I agree. Ravalomanana certainly made many errors, he said so himself. But he deserved to be challenged in a proper elections and not removed by the army. Whatever we think of the crisis in 2002, there was an election. If we go by the path dependent theory (which I am a big believer of)  how the country will develop (or not) in the next five years is strongly tied to what we decide today with our insitutions and how they come about. This is why the current  crisis is dragging along for so long. The theory also suggests that if you seize power the wrong way, the chances that you will make a legitimate leader afterwards are slim to none (I am looking at you, D. Camara and A. Rajoelina).

1 comment:

  1. Anonymous7:28 PM

    I would like more from your analysis... please... quite interesting really.

    ReplyDelete