Egen-function for apportionment methods
egen [type] newvar = apport(votesvar) [if exp] [in range] [, options]
options Description ------------------------------------------------------------------------- method(keyword) apportionment method size(#|varname) seats to be allocated threshold(#|varname) barrier clause exceptions(exp) Exceptions from barrier clause by(varname) apply method by groups -------------------------------------------------------------------------
Description
The egen-function apport(varname) creates a new variable holding the number of seats on the basis of the absolute number of valid votes in varname. The apportionment can be done by using either a quota method or five different divisor methods are introduced. A detailed description of these methods is given in the Stata Journal article regarding this program.
Options
method(keyword) is used to select the apportionment method. The apportionment methods described above can be specified using one of the following keywords:
------------------------------------------------------------ Method Keyword and synonyms ------------------------------------------------------------ Hamilton hamilton hare-niemeyer remainder vinton Jefferson jefferson dhondt hagenbach-bischoff greatest Webster webster stlague majorfraction Hill hill huntington geometric Dean dean harmonic Adam adam smallest ------------------------------------------------------------ The default is method(jefferson).
size(#|varname) is used to specify the number of seats to be allocated. Either use a positive integer number or the name of a variable holding the number of seats to be allocated. In the latter case, the variable should be constant within one apportionment problem (i.e., for one election). Size defaults to 100.
by(varlist) is used when the data set holds several apportionment problems (i.e., several elections), or when seats should be allocated separately for regional subdivisions. Specifying by(varlist) is equivalent to using by varlist: as a prefix.
threshold(#|varname) is used to set the barring clause. Within the parentheses put the size for the barring clause as a percentage or specify the name of a variable holding the value for the barring clause. The variable must be constant within one apportionment problem (i.e., for one election).
exception(exp) is used to specify exceptions from the barring clause. Within the parentheses specify an expression indicating the exempted observations (see help exp).
Examples
. use uspop if year==1790, clear . egen ham = apport(pop), method(hamilton) size(105)
. use uspop, clear . egen jeff = apport(pop), method(jefferson) size(size) by(year)
Author
Ulrich Kohler, WZB, kohler@wzb.eu
Also see
Manual: [D] egen
Online: egen, egenmore (if installed)