estimate_demand: Estimate a Demand Model From Aggregate Market Shares
In lachlandeer/demandr: Nested Logit Demand Estimation with Aggregate Market Shares

Description Usage Arguments Value Examples

Estimate a Demand Model From Aggregate Market Shares

1
2
3

estimate_demand(df, market_ids, product_id, market_share, outside_share,
  exog_charac, price, nest_shares = NULL, instruments = NULL,
  productFE = TRUE, marketFE = "both", supply_side = FALSE)

`df`	Dataframe to containing data to be used in estimation
`market_ids`	List of variables contaning market identifiers
`product_id`	Variable containing product identifier
`market_share`	Variable containing product market shares, as a character string
`outside_share`	Variable outside shares, as a character string
`exog_charac`	Vector of exogenous product characteristic variables, as a character vector
`price`	Variable containing market price, as a character string
`nest_shares`	Vector of nest share variables, as a character vector
`instruments`	Vector of instrument names, as a character vector - NOT CURRENTLY IMPLEMENTED
`productFE`	Logical variable, `TRUE` leads to estimation with product fixed effects
`marketFE`	Character string of which market fixed effects to contain, takes value "both", "geog" or "time"
`supply_side`	Logical variable, `TRUE` leads to estimation with the constraints from a Bertrand-Nash Oligopoly Model - NOT CURRENTLY IMPLEMENTED

Returns an object of class "lm" when instruments = FALSE

# Regression Without Nests
mkts <- create_markets(geog_id = "market_id", time_id = "year")
price_var     <- c("price_euro")
exog_var      <- c("kilowatts", "fueleff_avg", "width", "height", "domestic")
mkt_share     <- c("mkt_share")
outside_share <- c("outside_share")
## Create Shares
df   <- mkt_share_from_sales(eurocars,
                               quantity   = 'qty_sold',
                               population = 'population')
df2  <- create_shares(df, market_id  = mkts,
                          mkt_share  = 'mkt_share')
## Run regression
reg <- estimate_demand(df             = df2,
                        market_ids    = mkts,
                        product_id    = 'model_id2',
                        market_share  = mkt_share,
                        outside_share = outside_share,
                        exog_charac   = exog_var,
                        price         = price_var,
                        marketFE      = "both",
                        productFE     = TRUE)

# Regression With One Level Nest
mkts <- create_markets(geog_id = "market_id", time_id = "year")
price_var     <- c("price_euro")
exog_var      <- c("kilowatts", "fueleff_avg", "width", "height", "domestic")
mkt_share     <- c("mkt_share")
outside_share <- c("outside_share")
## Create Shares
df   <- mkt_share_from_sales(eurocars,
                               quantity   = 'qty_sold',
                               population = 'population')
df2  <- create_shares(df, market_id  = mkts,
                          mkt_share  = 'mkt_share',
                          nest_id    = 'class_id')
## Run regression
reg <- estimate_demand(df             = df2,
                        market_ids    = mkts,
                        product_id    = 'model_id2',
                        market_share  = mkt_share,
                        outside_share = outside_share,
                        exog_charac   = exog_var,
                        price         = price_var,
                        nest_shares   = c("within_nest"),
                        marketFE      = "both",
                        productFE     = TRUE)

# Regression With Two Level Nest
mkts <- create_markets(geog_id = "market_id", time_id = "year")
price_var     <- c("price_euro")
exog_var      <- c("kilowatts", "fueleff_avg", "width", "height", "domestic")
mkt_share     <- c("mkt_share")
outside_share <- c("outside_share")
## Create Shares
df   <- mkt_share_from_sales(eurocars,
                               quantity   = 'qty_sold',
                               population = 'population')
df2  <- create_shares(df, market_id  = mkts,
                          mkt_share  = 'mkt_share',
                          nest_id    = 'class_id',
                          subnest_id = 'domestic')
## Run regression
reg <- estimate_demand(df             = df2,
                        market_ids    = mkts,
                        product_id    = 'model_id2',
                        market_share  = mkt_share,
                        outside_share = outside_share,
                        exog_charac   = exog_var,
                        price         = price_var,
                        nest_shares   = c("within_nest", "within_subnest"),
                        marketFE      = "both",
                        productFE     = TRUE)