Discussion Thread on Testing the Solow Model & other Growth Evidence

Chad Jones uses two different methods in predicting how the World Income Distribution (WID) will evolve in the future. The first approach he uses is based on the findings of the Solow model, which Jones then uses to say that it is up to the fundamentals of a country such as its savings rate and population growth rate, that determines how fast it will grow in the future. He plugs in the fundamentals he finds of a country from 1988 to project the future levels of income distribution, and determine whether or not there has been change and in what direction (increase or decrease). The criticism of this approach, however, is that it assumes that the fundamentals of savings rate and population growth rate, for example, are fixed at 1988 levels. 

The second approach Jones takes is not reliant on the Solow model, and is rather called the Markov Transitional Matrix. Essentially, Jones here classifies the data he has, either within a country's household income distribution as we learned in class or across different countries, as bottom tier, middle tier, or top tier. Then, depending on how much of the data falls outside of the central diagonal formed by the data, Jones determines whether the country or countries experiences an increase/decrease/stagnation in income distribution levels in the future. The criticism about this approach, however, is that it assumes that the distributions of growth rates across countries would stay constant in the future as they were in the period between 1960 and 1988, which is a very big detail to assume and would skew the findings coming out of this approach. 

Lant Pritchett also tries to explain the idea of convergence through two separate methods. Because the only evidence that existed at the time was based on OECD countries or more rich/developed countries, Pritchett attempts to influence the literature by accounting for the poorer countries. He does this by finding what the lowest level of GDP/capita is in our modern day by looking at the poorest of countries in our time. He then assigns this level to countries that data could not be found for in the past, because Pritchett states that countries could not exist if they did not even meet this threshold for GDP/capita. This number is $250. The second method he uses is to prove that assigning this $250 for poor countries in the past is not outrageous. He does this by determining the relationship between caloric consumption and GDP/capita and then finding what the basic number of calories needed to survive was, which was around 2,400 calories. Pritchett then plugged 2,400 into his relationship between caloric consumption and GDP/capita to back out what the level of GDP/capita would be in countries who only managed to meet 2,400 calories/day, which was the minimum subsistence level. The number he got for GDP/capita using this method was between $250-$280, which confirms his ability to plug this in for poor countries that did not have data in the past. A criticism of this approach is that there will never a feasible way to determine convergence, because regardless of the minimum GDP/capita you place for the poor countries, empirically the model will show that the poor countries can never and will never catch up to the richer countries. We saw this in the last example in class, when countries were placed at $100, they still weren't reaching the 2% growth rate that rich countries were reaching yearly, symbolizing that these poorer countries could not catch up regardless of their low GDP/capita levels.