Question:

You have been asked by an aid organization to examine whether child hunger in the developing world is a more severe problem in autocratic states than in democratic states. You also suspect that education for girls could be an important factor in decreasing the level of child hunger. You employ a dataset on 98 developing countries, including the following variables:

Child hunger : percent of underweight children

Autocracy : a dummy variable where autocracy is coded as 1 and democracy as 0.

Female school enrollment : percent of girls enrolled in primary and secondary schooling (measured before the dependent variable to ensure time order)

1. Below you see some descriptive statistics for the dependent variable, child hunger, when splitting the sample of countries between autocratic and democratic states. Do you receive any support for your hypothesis that child hunger is higher in autocratic states?

   Table 1. Descriptive statistics for child hunger
   	Mean	Std dev
   In autocracies only	22.52	13.29
   In democracies only	16.73	12.60

2. In table 2 you find bivariate linear regression results for the effect of autocracy on child hunger. Give a substantive interpretation of the coefficent of autocracy, as well as the intercept (the constant). Can you say that this effect is statistically significant at the 95 percent level?

   Table 2. OLS for child hunger, bivariat
   Variable	Coefficient	t-value	p-value
   Autocracy	5.7923	2.19	0.031
   Constant	16.7320	8.44	0.000

   Note: The critical values for t at the 95% significance level are ±1.98
3. In the next step, you include autocracy and female school enrollment in the same regression (see table 3). What happens to the results when you go from a bivariate regression to a trivariate regression? Comment on the coefficients as well as the p-values.

   Table 3. OLS for child hunger, trivariat
   Variable	Coefficient	t-value	p-value
   Autocracy	0.1159	0.05	0.961
   Female school enrollment	-0.44355	-6.47	0.000
   Constant	59.5704	8.73	0.000

   Note: The critical values for t at the 95% significance level are ±1.98
4. Draw a causal diagram to illustrate these causal effects. You do not have to write out the effect of the arrows, but it is important that you argue for the direction of the effects. Also comment on whether the effect discovered in 3b can be described as spurious or indirect, based on your causal diagram. (Note that there is no exact right or wrong answer concerning the direction of the arrow between the two independent variables. The important thing is that you draw correct conclusions based on the direction that you propose.)

Price: $2.99

Solution: The downloadable solution consists of 4 pages
Deliverable: Word Document