Politicians of the developing world, pay attention. Experimental evidence of road paving on voter behavior in Mexico:
Families living along streets that were treated with pavement were 0.304 points more satisfied with the local government than those in the control group (on a 4-point scale).
Satisfaction with the State and Federal government also increased significantly, by 0.168 and 0.140 points. In other words, the direct effect of the local policy is about twice as large as the indirect effects.
…the indirect effect does not occur along party lines, but reflects a generalized improved perception for all other branches of government.
Finally… the return for the implementing politician in terms of vote share is 7 percentage points (20% increase in share) if an unpaved electoral section gets fully paved under the politician’s watch.
The paper is here.
Read why roads might be a good investment. The question is, are the vote-rewarding roads the same as growth-promoting roads?
Wouldn’t you expect paved roads to be systematically correlated with a range of other goods like streetlighting, access to health services (the roads will help, most likely), links to other paved roads?
Ranil – I think this is looking at the total causal effect of roads (or, in this case, asphalting), – and the whole point is to identify just that effect (so it isn’t that roads are part of some bigger scheme, the roads are the treatment).
Of course, the total effect may include stuff that follows (like access to health, etc) not just the warm fuzzy feeling you get from having a chunk of asphalt outside our door.
It’s nice to see research like this, but who is it really informing? In most SSA countries, for example, I think the governments would file this under `duh’.