Transport economics is a lively field. I have written papers on who uses public transit in the United States and I've studied the effectiveness of new rail systems built in the U.S (and their gentrification effects) and I've studied the complementarities between public investment in transit systems and private investment in nearby real estate and retail stores in China.
One point that I thought was established was the "fundamental law of traffic congestion". When roads are not priced, there is a classic tragedy of the commons such that each driver who seeks to maximize his own speed ignores the negative congestion externality he imposes on others. This means that when areas build new roads, average road speeds do not increase because at the extensive margin more people seek to make trips on these "empty roads" and this eventually slows everyone down again. I pose this because Science magazine has just published a new piece arguing that HOV lanes increase speeds in Indonesia's cities. Recall that HOV lanes allow you "special access" to a lane if there are multiple people in the vehicle. Ignoring transaction costs, I would have thought that HOV lanes would eventually move at the same speed as the other lanes. Why? Suppose that there are 5 drivers. 4 of the drivers can board 2 cars (with 2 in a car) to use the HOV lane. At first, this policy induced carpooling should increase traffic speeds because there are now fewer cars on the road but this isn't an equilibrium.
The Fundamental Law of Traffic Congestion would predict that there should be a set of people who are on the extensive margin who would take a car trip if the roads were less congested. Some of the members of this set will start driving if roads speeds increase and this "arbitrage" explains why the fundamental law is a fundamental law.
Now, since cars are expensive, there is a fixed cost to being at the Extensive Margin (i.e not driving on the highway now but having a car and thus having the option to drive). So, perhaps because cars are expensive and LDC residents on average are poor --- perhaps the Fundamental Law of Traffic Congestion does not hold in LDCs but will hold in the future as such nations grow richer and have more cars?
This would be an interesting hypothesis to test!