-------------------------------------------------------------------------------
help for isopoverty                                          Joao Pedro Azevedo
                                                                  Samuel Franco
-------------------------------------------------------------------------------

isopoverty income [weight] [if exp] [in range] [ , varpl(varname) pline(number) poverty inequal stepgrw(number) stepinq(number) mininq(number) maxinq(number) mingrw(number) maxgrw(number) frontier int(number) target(number) ]

fweights, and aweights are allowed; see help weights.

Description

isopoverty generates data that can be used to plot the Inequality-Poverty, Growth-Poverty and the Iso-Poverty curves.

stepgrw or stepinq have to be stated otherwise the programme does not work. If the option stepgrw is specified the ado automatically generates the Growth-Poverty data, likewise, if the option stepinq is spefied the program it will generate the Inequality-Poverty data. If both parameters are specified both procedures will be implemented.

Please note that in order to provider a greater flexibility on the graph options this ado does not automatically produce any figures. The final output is stored as a matrix in the Stata memory (please see ereturn), in order to retrive it to the data set the use will have to use the svmat command. An illustration of a figure produced with data from this ado can be found in the examples below.

Inequality -> r(ineqreduc) Growth -> r(growth) Iso-Poverty -> r(frontier)

WARNING! The Iso-Poverty is quite computationally intensive. We strongly advice the user to first estimate the Inequality-Poverty curve and then the Growth-Poverty curve, in order to find out the best cut off points for the data that she or he is using.

Options

varpl(varname) provide the variable name containing the values poverty line.

pline(#) sets the value of the poverty line. The default poverty line is computed as half the median of varname.

poverty is the default output and includes the head count ratio and the extremme poverty head count ratio, several other poverty measures can be easily included in this output.

inequal includes the Gini and Theil inequallity measures in the output.

Options Inequality

mininq(#) states the minimum inquality reduction. Default value 0.

maxinq(#) states the maximum inequality reduction. Default value 1.

stepinq(#) specifies the number of increments that will be ploted between the minium and maximum values of the inequality variation.

Options Growth

mingrw(#) states the minimum growth. Default value 0.

maxgrw(#) states the maximum growth. Default value 1.

stepgrw(#) specifies the number of increments that will be ploted between the minium and maximum values of the growth variation.

Options Frontier

frontier this option has to be stated if the user wants to estimate the Iso-Poverty line, otherwise the ado will estimate the data for the Growth-Poverty and the Inequality-Poverty curvers separtely.

target specify the poverty rate as percentage of the population (head count) from the Iso-Poverty curve being estimated. Default value 0.

int specify the interval of the poverty rate that is acceptable. This is a usefull option the user is using large increments on the both the poverty and inequalty variation. Default value 0.

Examples

. isopoverty rdpc, stepinq(50)

. isopoverty rdpc, stepinq(50) varpl(lp)

. isopoverty rdpc [fw=peso], stepinq(50) . mat temp=r(ineqreduc) . svmat double temp . graph twoway line temp2 temp1 if temp2>0, ytitle("Poverty - Head Count") xtitle("Inequality") name(temp1, replace)

. isopoverty rdpc [fw=peso], stepinq(50) varpl(lp)

. isopoverty rdpc, stepinq(50) inequal

. isopoverty rdpc, stepinq(50) stepgrw(10) mininq(0) maxinq(.50) mingrw(0) maxgrw(1.50) target(25) int(2) frontier varpl(lp)

. isopoverty rdpc [fw=peso], stepinq(200) stepgrw(10) mininq(0) maxinq(.30) mingrw(0) maxgrw(1.50) target(25) int(2) frontier varpl(lp)

Authors

Joao Pedro Azevedo jazevedo@ipea.gov.br

Samuel Franco sam@ipea.gov.br