REAT (version 3.0.2)

howard.xcl: Howard-Newman-Tarp excess colocation (XCL) index

Description

Calculating the excess colocation (XCL) index by Howard, Newman and Tarp for two industries

Usage

howard.xcl(k, industry, region, industry1, industry2, no.samples = 50, e_k = NULL)

Arguments

k

a vector containing the IDs/names of firms \(k\)

industry

a vector containing the IDs/names of the industries \(i\)

region

a vector containing the IDs/names of the regions \(j\)

industry1

Regarded industry 1 (out of the industry vector)

industry2

Regarded industry 2 (out of the industry vector)

no.samples

Number of samples for the counterfactual firm allocation via bootstrapping

e_k

Employment of firm \(k\)

Value

A single value of \(XCL\)

Details

The Howard-Newman-Tarp excess colocation index (\(XCL\)) is standardized (\(-1 \le CL \le 1\)). The rationale behind is that the CL index (see howard.cl) is compared to a counterfactual (random) location pattern which is constructed via bootstrapping. Processing time depends on the number of firms and the number of samples.

References

Howard, E./Newman, C./Tarp, F. (2016): “Measuring industry coagglomeration and identifying the driving forces”. In: Journal of Economic Geography, 16, 5, p. 1055-1078.

See Also

howard.cl, howard.xcl2, ellison.c, ellison.c2

Examples

# NOT RUN {
# example from Howard et al. (2016):
firms <- 1:6
industries <- c("A", "B", "A", "B", "A", "B")
locations <- c("X", "X", "X", "Y", "Y", "X")

howard.xcl(firms, industries, locations, industry1 = "A", 
industry2 = "B")
# }