Package 'nopp'

Title: Nash Optimal Party Positions
Description: Estimation of party/candidate ideological positions that correspond to a Nash equilibrium along a one-dimensional space as in Curini and Iacus (2017) <doi:10.18637/jss.v081.i11>.
Authors: Luigi Curini, Stefano M. Iacus
Maintainer: Stefano M. Iacus <[email protected]>
License: GPL (>= 2)
Version: 1.1.3
Built: 2024-11-03 03:49:52 UTC
Source: https://github.com/siacus/nopp

Help Index

Nash Optimal Party Positions

Description

Estimation of party/candidate ideological positions that correspond to a Nash equilibrium along a one-dimensional space

Details

Package: nopp
Type: Package
Version: 1.0
Date: 2012-06-26
License: GPL (>= 2)

nopp is a package for R which enables to compute party/candidate ideological positions that correspond to a Nash Equilibrium along a one-dimensional space. It accommodates alternative motivations in (each) party strategy while allowing to estimate the uncertainty around their optimal positions through two different procedures (bootstrap and MC).

Author(s)

Luigi Curini, Stefano M. Iacus

Maintainer: Luigi Curini <[email protected]>, Stefano M. Iacus <[email protected]>

References

Adams, James F., Samuel Merrill III, and Bernard Grofman (2005). A Unified Theory of Party Competition. Cambridge: Cambridge University Press

Merrill, Samuel III, and James Adams (2001), Computing Nash Equilibria in Probabilistic, Multiparty Spatial Models with Nonpolicy Components, Political Analysis, 9, 347–61

Curini, L., and Iacus, S.M. (2017) Nash Optimal Party Positions: The nopp R Package, Journal of Statistical Software, 81(11), 1–25

Nash Optimal Party Positions

Description

Nash Optimal Party Positions

Usage

equilibrium(start, model, data, tolerance = 1e-05, max.iter = 100, 
coal = 0, alpha = 0, margin = NULL, fixed = NULL, gamma = 0, 
boot = 0, MC = 0, self.var = "self", prox.var="prox", 
position=NULL, votes=NULL, quadratic=TRUE, conf.level = 0.95)

Arguments

start

initial party positions. Numerical vector. Optional.
model

the mlogit model analysis
data

the data set
tolerance

tolerance in the convergence of Nash equilibrium. Default 1e-5
max.iter

max iteration to convergence in Nash equilibrium. Default 100
coal

a list specificing electoral coalitions. See Details.
alpha

the weight of coalition vote-share in party utility function. Default = 0. See Details.
margin

a list specifing the vote share margin to be maximized of a party/coalition against other party/coalition. See Details.
fixed

a list of fixed party positions. See Details.
gamma

the weight among nash and fixed arty position. Default=0. See Details.
boot

number of boostrap replications. See Details.
MC

number of Monte Carlo replications. See Details.
self.var

character: name of self-placement of respondent. See Details.
prox.var

character: name of party-placement variable. See Details.
position

a named list: of perceived position of parties. See Details.
votes

a named list: of actual vote share at election. See Details.
quadratic

a logical value: if FALSE the linear utility function is used to calculate the proximity. See Details.
conf.level

significant level for empirical Monte Carlo or bootstrap confidence intervals.

Details

See vignette.

Value

an object of class nash.eq

Note

See the vignette for detailed explanations and other working examples.

Author(s)

Luigi Curini, Stefano M. Iacus

References

Adams, James F., Samuel Merrill III, and Bernard Grofman (2005). A Unified Theory of Party Competition. Cambridge: Cambridge University Press

Merrill, Samuel III, and James Adams (2001), Computing Nash Equilibria in Probabilistic, Multiparty Spatial Models with Nonpolicy Components, Political Analysis, 9, 347–61

Curini, L., and Iacus, S.M. (2017) Nash Optimal Party Positions: The nopp R Package, Journal of Statistical Software, 81(11), 1–25

See Also

See Also as plot.nash.eq

Examples

## Not run: 
data(italy2006)

str(italy2006)
italy2006[1:2,1:14]

election <- set.data(italy2006 , shape="wide", choice="vote",
varying=c(5:14), sep="_")
str(election)

m <- mlogit(vote~prox+partyID | gov_perf+sex+age+education, 
election, reflevel = "UL")
summary(m)

true.pos <- list(FI=7.59, UL=3.50, RC=1.95, AN=8.08, UDC=5.66)
true.votes <- list(FI=.24, UL=.40, RC=.10, AN=.18, UDC=.08)
# model 1: comparison against true votes and party positions
nash.eq <- equilibrium(model=m, data=election, pos=true.pos, 
 votes=true.votes)
nash.eq

par(mfrow=c(3,1))
plot(nash.eq)
par(mfrow=c(1,1))

# model 2: colation behaviours
coal1 <- list(FI=1, UL=2, RC=2, AN=1, UDC=1)
alpha1 <- list(FI=0.5, UL=0.5, RC=0.5, AN=0.5, UDC=0.5)
nash.eq <- equilibrium(model=m, data=election, coal=coal1, 
 alpha=alpha1)
nash.eq

# model 3: colation behaviours
coal1 <- list(FI=1, UL=2, RC=2, AN=1, UDC=1)
alpha1 <- list(FI=0.7, UL=0.8, RC=0.1, AN=0.5, UDC=0.9)
nash.eq <- equilibrium(model=m, data=election, coal=coal1, 
 alpha=alpha1)
nash.eq

# model 4: rivals tends to separate each other
nash.eq <- equilibrium(model=m, data=election, margin=list(FI="UL", UL="FI"))
nash.eq

# model 5: fixed position averaged with Nash equilibrium solution
nash.eq <- equilibrium(model=m, data=election, fixed=list(RC=1), gamma=0.2)
nash.eq

# model 6: rivals tends to separate each other with 
# fixed position averaged with Nash equilibrium solution
nash.eq <- equilibrium(model=m, data=election,  
margin=list(FI="UL", UL="FI"), fixed=list(RC=1), gamma=0.2)
nash.eq

# model 7: coalition and fixed position averaged with 
# Nash equilibrium solution
coal1 <- list(FI=1, UL=2, RC=2, AN=1, UDC=1)
alpha1 <- list(FI=0.7, UL=0.8, RC=0.5, AN=0.5, UDC=0.5)
nash.eq <- equilibrium(model=m, data=election,  coal=coal1, 
 alpha=alpha1, fixed=list(RC=1), gamma=0.2)
nash.eq

# model 8: Bootstrap analysis
set.seed(280715)
nash.eq <- equilibrium(model=m, data=election, boot=10)
nash.eq

# model 9: Monte Carlo simulation
set.seed(280715)
nash.eq <- equilibrium(model=m, data=election, MC=10)
nash.eq

## End(Not run)

2006 Italian General Election survey

Description

2006 Italian General Election survey, with quadratic ideological proximity.

Usage

data(italy2006)

Format

A data frame with 438 observations on the following 18 variables.

country

country name

id

id of respondent

vote

a factor with levels FI UL AN UDC RC for each party voted

self

self-placement of respondent on a 0 to 10 left-right scale

prox_FI

see Details.

prox_UL

see Details.

prox_AN

see Details.

prox_UDC

see Details.

prox_RC

see Details.

partyID_FI

see Details.

partyID_UL

see Details.

partyID_AN

see Details.

partyID_UDC

see Details.

partyID_RC

see Details.

sex

gender variable 1 = female

age

see Details.

education

see Details.

gov_perf

see Details.

Details

In this survey respondents were asked to indicate which party they voted for in the 2006 Election. The data concerns 5 parties: UL (Ulivo), RC (Communist Refoundation party), FI (Forza Italia), AN (National Alliance) and UDC (Union of Christian Democrats).

prox_* quadratic ideological distance between the respondent and a party * placement

partyID_* binary variable equals to 1 if the respondent declares to feel herself close to party *

age : 1 = "18-24 years", 2 = "25-34", 3 = "35-44", 4 = "45-54", 5 = "55-64", 6 = "65 +"

education : 0 = "up to primary school", 1 = "incomplete secondary", 2 = "secondary completed", 3 = "post-secondary trade", 4 = "university undergraduate degree inc", 5 = "university undergraduate degree comp"

gov_perf : 1 = "very good job", 2 = "good job", 3 = "bad job", 4 = "very bad job"

Source

CSES - Comparative Study of Electoral Systems: https://cses.org/).

Examples

data(italy2006)
head(italy2006)

2006 Italian General Election survey

Description

2006 Italian General Election survey, with linear ideological proximity.

Usage

data(italy2006.lin)

Format

A data frame with 438 observations on the following 18 variables.

country

country name

id

id of respondent

vote

a factor with levels FI UL AN UDC RC for each party voted

self

self-placement of respondent on a 0 to 10 left-right scale

proxlin_FI

see Details.

proxlin_UL

see Details.

proxlin_AN

see Details.

proxlin_UDC

see Details.

proxlin_RC

see Details.

partyID_FI

see Details.

partyID_UL

see Details.

partyID_AN

see Details.

partyID_UDC

see Details.

partyID_RC

see Details.

sex

gender variable 1 = female

age

see Details.

education

see Details.

gov_perf

see Details.

Details

In this survey respondents were asked to indicate which party they voted for in the 2006 Election. The data concerns 5 parties: UL (Ulivo), RC (Communist Refoundation party), FI (Forza Italia), AN (National Alliance) and UDC (Union of Christian Democrats).

prox_* linear ideological distance between the respondent and a party * placement

partyID_* binary variable equals to 1 if the respondent declares to feel herself close to party *

age : 1 = "18-24 years", 2 = "25-34", 3 = "35-44", 4 = "45-54", 5 = "55-64", 6 = "65 +"

education : 0 = "up to primary school", 1 = "incomplete secondary", 2 = "secondary completed", 3 = "post-secondary trade", 4 = "university undergraduate degree inc", 5 = "university undergraduate degree comp"

gov_perf : 1 = "very good job", 2 = "good job", 3 = "bad job", 4 = "very bad job"

Source

CSES - Comparative Study of Electoral Systems: https://cses.org/).

Examples

data(italy2006.lin)
head(italy2006.lin)
## maybe str(italy2006.lin) ; plot(italy2006.lin) ...

2006 Italian General Election survey

Description

2006 Italian General Election survey - wide format

Usage

data(italy2006.wide)

Format

A data frame with 524 observations on the following 15 variables.

country

country name

id

id of respondent

vote

a factor with levels FI UL AN UDC RC for each party voted

self

self-placement of respondent on a 0 to 10 left-right scale

FI

see Details.

DS

see Details.

AN

see Details.

DL

see Details.

UDC

see Details.

RC

see Details.

pID

see Details.

sex

gender variable 1 = female

age

see Details.

education

see Details.

gov_perf

see Details.

Details

In this survey respondents were asked to indicate which party they voted for in the 2006 Election. The data concerns 5 parties: UL (Ulivo), RC (Communist Refoundation party), FI (Forza Italia), AN (National Alliance) and UDC (Union of Christian Democrats). The dataset is in wide format.

variable from FI to RC identify the placement of those parties, on a 0 to 10 left-right scale, as perceived by the respondent.

pID is a variable that identifies the partisanship of the respondent (where 0=stands for no partyID, 1 = FI partyID, 23 = UL partyID, 3 = AN partyID, 4 = UDC partyID, 6 = RC partyID)

age : 1 = "18-24 years", 2 = "25-34", 3 = "35-44", 4 = "45-54", 5 = "55-64", 6 = "65 +"

education : 0 = "up to primary school", 1 = "incomplete secondary", 2 = "secondary completed", 3 = "post-secondary trade", 4 = "university undergraduate degree inc", 5 = "university undergraduate degree comp"

gov_perf : 1 = "very good job", 2 = "good job", 3 = "bad job", 4 = "very bad job"

Source

CSES - Comparative Study of Electoral Systems: https://cses.org/).

Examples

data(italy2006.wide)
head(italy2006.wide)
## maybe str(italy2006.wide) ; plot(italy2006.wide) ...

Show the NEWS file

Description

Show the NEWS file of the nopp package.

Usage

noppNews()

Value

None.

Plot function for Nash equilibrium object

Description

Plot function for Nash equilibrium object

Usage

## S3 method for class 'nash.eq'
plot(x,...)

Arguments

x

a nash.eq object
...

additional arguments passed to the inner plot function

Details

See vignette.

Author(s)

Luigi Curini, Stefano M. Iacus

References

Curini, L., and Iacus, S.M. (2017) Nash Optimal Party Positions: The nopp R Package, Journal of Statistical Software, 81(11), 1–25

See Also

See Also as equilibrium

Examples

## Not run: 
data(italy2006)
election <- set.data(italy2006 , shape="wide", choice="vote", varying=c(5:14), sep="_")
m <- mlogit(vote~prox+partyID | gov_perf+sex+age+education, election, reflevel = "UL")

true.pos <- list(FI=7.59, UL=3.50, RC=1.95, AN=8.08, UDC=5.66)
true.votes <- list(FI=.24, UL=.40, RC=.10, AN=.18, UDC=.08)

# comparison against true votes and party positions
nash.eq <- equilibrium(model=m, data=election, pos=true.pos, votes=true.votes)
nash.eq
par(mfrow=c(3,1))
plot(nash.eq)



# bootstrap confidence intervals
nash.eq <- equilibrium(model=m, data=election, boot=10)
nash.eq
plot(nash.eq)

par(mfrow=c(1,1))

## End(Not run)

Prepares data for Nash equilibrium

Description

Prepares data for Nash equilibrium

Usage

set.data(data, shape="wide", choice, varying, sep="_")

Arguments

data

the data set
shape

either wide or long. Default wide. See Details.
choice

the variable indicating the choice made: it can be either a logical vector, a numerical vector with 0 where the alternative is not chosen, a factor with level 'yes' when the alternative is chosen.
varying

the indexes of the variables that are alternative specific. See Details.
sep

the seperator of the variable name and the alternative name (only relevant for a wide data.frame). See Details.

Details

For general examples see the vignette. The arguments shape, choice, varying and sep as as in the mlogit.data function.

Value

A mlogit.data object, which is a data.frame in long format, i.e. one line for each alternative. It has a index attribute, which is a data.frame that contains the index of the choice made ('chid'), the index of the alternative ('alt') and, if any, the index of the individual ('id'). The choice variable is a boolean which indicates the choice made. This function use reshape if the data.frame is in wide format. It also has the attribute call for further data manipulation in the bootstrap task of equilibrium.

Note

See the vignette for detailed explanations and other working examples.

Author(s)

Luigi Curini, Stefano M. Iacus

References

Curini, L., and Iacus, S.M. (2017) Nash Optimal Party Positions: The nopp R Package, Journal of Statistical Software, 81(11), 1–25

Examples

## Not run: 
data(italy2006)

str(italy2006)
italy2006[1:2,1:14]

election <- set.data(italy2006 , shape="wide", choice="vote",
varying=c(5:14), sep="_")
str(election)

m <- mlogit(vote~prox+partyID | gov_perf+sex+age+education, 
election, reflevel = "UL")
summary(m)

true.pos <- list(FI=7.59, UL=3.50, RC=1.95, AN=8.08, UDC=5.66)
true.votes <- list(FI=.24, UL=.40, RC=.10, AN=.18, UDC=.08)
# model: comparison against true votes and party positions
nash.eq <- equilibrium(model=m, data=election, pos=true.pos, 
 votes=true.votes)
nash.eq

## End(Not run)