Compute Winning rate across the crisis indicator

The indicator focuses on companies that after the emergency outbreak win public contracts in the relevant economic market much more frequently than before the emergency.

Motivation

The red flag considers at risk companies that exceptionally increase their competitive power over the emergency outbreak, in terms of proportion of awarded contracts on the relevant economic market(s).

Scoring rule

The computation procedure returns 1 - p-value of the involved test (so that high values of the indicator correspond to high levels of corruption risk). When computing the composite, it will be dichotomised to 1 if statistical test is significant, and 0 otherwise (see normalise()).

Main target unit

This indicator targets companies.

Usage

ind_1(data, publication_date, emergency_name, stat_unit, test_type, cpvs, ...)

Arguments

data

a dataframe containing the data to use for computing the indicator.

publication_date

name of the variable in data containing the publication date of each contract.

emergency_name

string specifying the name of the emergency to consider. Examples could include "Coronavirus" or "Terremoto Centro Italia 2016-2017".

stat_unit

name of the variable in data containing the target unit ID (in this case, the company).

test_type

string specifying the statistical test to use for computing the indicator. Available options are "barnard", "fisher", or "z-test".

cpvs

character vector of CPV divisions (first two digits of CPV code) on which data are filtered out. Note: a panel of experts have already chosen which CPV divisions are most affected by which emergency.

...

other parameters to pass to generate_indicator_schema(), such as country_name (default: Italy).

Value

indicator schema as from generate_indicator_schema().

See also

ymd mutate, if_else, group_by, summarise, context, filter, rowwise, select, reexports str_sub nest

Examples

if (FALSE) {
if (interactive()) {
  mock_data_core <- mock_data_core |>
    tidyr::unnest(aggiudicatari, keep_empty = TRUE)
  ind_1(
    data = mock_data_core,
    publication_date = data_pubblicazione,
    stat_unit = codice_fiscale,
    emergency_name = "coronavirus",
    test_type = "fisher"
  )
}
}