The goal of ispv is to retrieve and load data from the Czech database of average earnings (ISPV, Informační systém o průměrném výdělku).
Installation
You can install the development version of ispv like so:
remotes::install_github("petrbouchal/ispv")Example
The package does two things, either for national or regional data:
- retrieve links to Excel files on the ISPV site for a given year
- read those data and return in a clean tibble ready for analysis
Here is an example for regional data:
library(ispv)
links <- pv_list_reg(2020)
paths <- file.path(getwd(), links$name[c(1, 2, 15, 16)])
purrr::walk2(links$url[c(1, 2, 15, 16)], paths, curl::curl_download)
data <- pv_reg_monthlypay_isco4(paths)
data
#> # A tibble: 545 × 23
#> kraj_id_ispv isco4_full fte_thous pay_median pay_d1 pay_q1 pay_q3 pay_d9 pay_mean bonus_perc supplements_perc compensation_pe…
#> <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 Pra 1120 Nejvyšší… 1.25 147534. 39480. 66898. 3.04e5 4.48e5 214252. 28.0 0.46 10.4
#> 2 Pra 1211 Řídící p… 2.84 100871. 44964. 68750. 1.59e5 2.20e5 131047. 19.3 0.73 10.0
#> 3 Pra 1212 Řídící p… 0.900 100313. 45651. 64583. 1.66e5 2.37e5 129529. 18.6 0.53 11.0
#> 4 Pra 1219 Ostatní … 1.62 76178. 31050. 49090. 1.31e5 1.86e5 102520. 20.9 1.3 10.5
#> 5 Pra 1221 Řídící p… 4.88 91330. 38618. 56198. 1.47e5 2.35e5 119951. 24.9 0.71 10.6
#> 6 Pra 1223 Řídící p… 0.989 99731. 45760. 67100. 1.43e5 2.04e5 118815. 18.3 0.82 11.0
#> 7 Pra 1321 Řídící p… 1.44 67123. 20246. 44127. 1.12e5 1.57e5 87536. 24.4 0.82 10.6
#> 8 Pra 1323 Řídící p… 2.07 75469. 31909. 49077. 1.04e5 1.38e5 83886. 21.6 0.52 10.3
#> 9 Pra 1324 Řídící p… 2.40 72326. 35602. 48826. 1.10e5 1.54e5 88585. 17.7 1.83 13.3
#> 10 Pra 1330 Řídící p… 2.81 115329. 52052. 73825. 1.62e5 2.29e5 134422. 19.6 1.3 9.64
#> # … with 535 more rows, and 11 more variables: hours_per_month <dbl>, file <chr>, sfera <chr>, period <chr>, year <chr>,
#> # quarter <chr>, kraj_id <chr>, kraj_name <chr>, kraj_id_nuts3 <chr>, isco4_id <chr>, isco4_name <chr>This is a long format data frame, so is easy to analyse and visualise:
library(ggplot2)
ggplot(data_select, aes(pay_mean, kraj_name, fill = sfera)) +
geom_col() +
facet_wrap(~sfera, ncol = 1) +
theme_minimal()
About the data
The data is special in several ways:
- it contains pay and staff counts broken by ISCO-4 level for each region (kraj), further separated into the “wage sphere” and “salary sphere” - broadly private and public sector, respectively
- pay is broken down into regular pay, bonuses, other compensation and other components
This is probably the most detailed labour market data available publicly.
See details on the official page: English and Czech
The data is only available to the public in Excel files, separately for each year and region. This package makes it possible to download and read the data in bulk.
The public sector part of the data is based on individual-level returns provided by each public sector organisation. The private sector part is based on a survey.
The data items are documented in the relevant data loading functions, i.e. currently only pv_load_reg.