For this problem set we will use data from the 2008 Mexico National Rural Household Survey (Encuesta Nacional a Hogares Rurales de Me?xico, acronym: ENHRUM) to study inequality and poverty in rural Mexico. You will only need a few basic Excel skills (plus a mastery of the inequality and poverty indexes discussed in class).

Start by opening up the Excel file called "ENHRUM." This file contains data from a random sample of rural Mexican households (N=927). The variables (columns) include:

(A) hhid: A household identifier (you don't have to worry about this?it's a wacky number)

(B) hhsize: A household's size (number of people living in the house 9 months or more in 2007)

(C) y: a household's total income in 2007,excluding government transfers. You will notice that some (the poorest) households actually had zero income in 2007, mostly due to agricultural and livestock losses.

(D) t:the amount of money transferred from the government to a household as part of a social protection program.

Incomes are all in 2007 pesos. The nominal exchange rate was 20.6 pesos per US$. The cost of living is relatively low in Mexico. The PPP adjustment factor for Mexico is 9.904.

Use the data in the spreadsheet and what you have learned so far in class to answer the following questions about income in rural Mexicoexcluding government transfers.

What is the average per-capita income in U.S. dollars for Mexico's rural households in 2007 excluding government transfers using the nominal exchange rate then adjusting for PPP? [Hint: you will have to first calculate income in per-capita terms. It will also help if you sort the data from smallest to largest by income per capita.]

We know Mexico is an "upper-middle" income country. If rural Mexico were its own country, would it be an upper-middle country, too? Which exchange rate?the nominal one or PPP adjusted?do you use to answer this question, and why? [Hint: recall the definition of "upper-middle" from Taylor and Lybbert, p. 68-70.]

1

Next, we want to determine the level of inequality and the level of poverty in rural Mexico. creating a new variable (column) called rankpcy. This will be the household's income rank in terms of per-capita total income in U.S. dollars, PPP adjusted, excluding government transfers. The household with the least amount of per capita income should have a value of rankpcy equal to 1. The household with the second least amount of per capita income should have a value of rankpcy equal to 2, and so on.The richest household should have a value of rankpcy equal to 927.

Then, using rankpcy create new variable called F(pop), which will be the cumulative distribution of the population (i.e., the ratio of the household value ofrankpcycompared with the total number of households). The way to check if you have done this correctly is thatthe richest household should have a value of F(pop) equal to 1.

Finally, create new variable called F(pcy), which will be the cumulative distribution of income (i.e., the ratio of income earned by a given household, plus the income of all households that are poorer, compared to the sum of income earned by all households). The way to check if you have done this correctly is thatthe richest household should have a value of F(pcy) equal to 1.

Create graph the shows the Lorenz Curve for this sample of rural Mexican households. This can be done by graphing the data in F(pop) and F(pcy). What does the shape of the Lorenz Curve tell you about the distribution of income in rural Mexico?

Calculate the Gini coefficient of inequality for rural Mexican households, excluding government transfers. Is it high, medium, or low? Explain. [Hint: recall the mathematical equation for the Gini coefficient in Taylor and Lybbert, p. 98-101. Also note that while the book refers to the cumulative distribution of income, we really want to use the cumulative distribution of population in this calculation.]

In the year of our survey, the per-capita food poverty line in rural Mexico was 7,596 pesos. Use this food poverty line to calculate all three versions of the Foster-Greer-Thorbeck (FGT) index. In all three cases, treat the unit of analysis as the household, not the individuals in the households. Make sure you use income per capita in pesos.

5. Calculate the FGT when ?? = 0 (Headcount Poverty Index) and briefly explain what it means. [Hint: recall the formula in Taylor and Lybbert, p. 83-4.]

6. Calculate the FGT when ?? = 1 (Poverty Gap Index) and briefly explain what it means.

7. Calculate the FGT when ?? = 2 (Severity of Poverty Index) and briefly explain what it means.

Mexico has two main government programs that transfer income to rural households. 2

PROCAMPO, which pays a set amount per acre to farmers who grew basic grains in a base year prior to the elimination of guaranteed prices. Output price guarantees are no longer allowed under the old North American Free Trade Agreement (NAFTA), and this did not change with the new United States-Mexico-Canada Agreement (USMCA). The idea here was to replace the price support with a direct income payment to farmers, to compensate them for lower grain prices. Direct income payments are allowed, because unlike price supports, they are not expected to affect production. Most PROCAMPO payments go to corn farmers, big and small.

OPORTUNIDADES, a conditional welfare program that gives payments to poor women provided that their children are enrolled in schools and in the local clinic. This is the successor program to PROGRESA that we discussed in class. The idea is to combat poverty in the short run (by giving money to women) and in the long run (by making sure kids get the human capital they need to be productive and make an income when they get older). Because of the evidence generated by the first RCT that showed PROGRESA was a success, it has become a model for welfare programs in less-developed countries.

8. How would youexpecteach one of these public transfer programs to affect (i) inequality and (ii) poverty in rural Mexico? Why?

Add government transfers to household income in your spreadsheet. This requires you to repeat many of the previous steps. How do government transfers change each of the following? (Please calculate the number and briefly explain in words).

Per-capita income. Would this change rural Mexico's position in the World Bank's categorization of countries by income level?

The Lorenz Curve. Create new Lorenz Curve with transfers. Has the distribution of income in rural Mexico changed? Is it more equal or less?

The Gini coefficient of inequality. By this measure, are government transfers effective at reducing inequality in rural Mexico?

The Headcount Poverty Index. By this measure, are government transfers effective at reducing poverty in rural Mexico?

The Poverty Gap. By this measure, are government transfers effective at reducing poverty in rural Mexico?

The Severity of Poverty Index. By this measure, are government transfers effective at reducing poverty in rural Mexico?

In percentage terms, is the effect the same using these three indexes? Where are government transfers having the biggest effect?

3