Author

AMW & MVH

Bring in the data

Code 
library(tidyverse)
library(dplyr)
library(plm)
#install.packages("gghighlight")
library(gghighlight)

#16,354 obs
raw_data_joined <- read_csv("agency_raw_joined.csv")
panel_data <- read_csv("panel_data.csv")
# Tables from PTAXSIM
cpi <- read_csv("./Necessary_Files/cpi.csv") # has two year variables!!
eq_factor <- read_csv("./Necessary_Files/eq_factor.csv") %>% 
  select(-eq_factor_tentative)

There were 16,354 observations in the original ‘agency’ data table from PTAXSIM.

After removing SSAs, municipalities that change home rule status, taxing agencies that are not fully in Cook County, there are 10,374 agencies.

Check filter steps to get to that number for the panel data.

Code 
theme_awm <- function(){ 

    theme_minimal() %+replace%    #replace elements we want to change
    
    theme(
      
      #grid elements
      panel.grid.major = element_blank(),    #strip major gridlines
      panel.grid.minor = element_blank(),    #strip minor gridlines
      axis.ticks = element_blank(),          #strip axis ticks
      
      #since theme_minimal() already strips axis lines, 
      #we don't need to do that again
      
      #text elements
      plot.title = element_text(             #title
                   size = 14,                #set font size
                   face = 'bold',            #bold typeface
                   hjust = 0,                #left align
                   vjust = 2),               #raise slightly
      
      plot.title.position = "plot",
      
      plot.subtitle = element_text(          #subtitle
                   size = 12),               #font size
      
      plot.caption = element_text(           #caption
                   size = 9,                 #font size
                   hjust = 1),               #right align
      
      axis.title = element_text(             #axis titles
                   size = 10),               #font size
      
      axis.text = element_text(              #axis text
                   size = 9),                #font size
      
      axis.text.x = element_text(            #margin for axis text
                    margin=margin(5, b = 10))
      
      #since the legend often requires manual tweaking 
      #based on plot content, don't define it here
    )
}

panel_data <- panel_data |> filter(!minor_type %in% c("BOND", "UNIFIED", "COMM COLL", "COOK", "MISC", "MOSQUITO") )
Code 
raw_data_joined %>% 
  ggplot (aes(x= major_type)) + 
  geom_bar() + 
  theme_classic() +  
  geom_text(aes(label = ..count..), stat = "count", vjust=-.1) + 
 # coord_flip() +
  labs(y="Number of taxing agencies", x= "'major_type' Categories") + 
  scale_y_continuous(label = scales::comma)
Warning: The dot-dot notation (`..count..`) was deprecated in ggplot2 3.4.0.
ℹ Please use `after_stat(count)` instead.

Code 
raw_data_joined %>% 
  ggplot (aes(x= minor_type)) + 
  geom_bar() + 
  theme_awm( ) + 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 2) + 
  coord_flip() +
  labs(title = "Taxing Agency Count: Before Dropping any Observations", y="Number of taxing agencies", x= "'minor_type' Categories") + 
  scale_y_continuous(label = scales::comma)

Code 
raw_data_joined %>% 
  filter(year == 2022) %>%
  ggplot (aes(x= minor_type)) + 
  geom_bar() + 
  theme_awm( ) + 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 2) + 
  coord_flip() +
  labs(title = "Taxing Agency Count: Before Dropping any Observations", y="Number of taxing agencies", x= "'minor_type' Categories") + 
  scale_y_continuous(label = scales::comma)

Code 
panel_data %>% 
  ggplot (aes(x= major_type)) + 
  geom_bar() + 
  theme_classic() +  
  geom_text(aes(label = ..count..), stat = "count", vjust=-.1) + 
 # coord_flip() +
  labs(y="Number of taxing agencies", x= "'major_type' Categories") + 
  scale_y_continuous(label = scales::comma)

Code 
panel_data %>% 
  ggplot (aes(x= minor_type)) + 
  geom_bar() + 
  theme_awm( ) + 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 2) + 
  coord_flip() +
  labs(title = "Taxing Agency Count: After Dropping Observations", y="Number of taxing agencies", x= "'minor_type' Categories") + 
  scale_y_continuous(label = scales::comma)

Code 
panel_data %>% 
  filter(year == 2022) %>%
  ggplot (aes(x= minor_type)) + 
  geom_bar() + 
  theme_awm( ) + 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 2) + 
  coord_flip() +
  labs(title = "Taxing Agency Count: After Dropping Observations", y="Number of taxing agencies", x= "'minor_type' Categories") + 
  scale_y_continuous(label = scales::comma)

Recoding Variables

copy and pasted from ‘helper_recode_variables.R’ file

Code 
dropped_munis <- c("030250000", "030270000", "030300000", "030585000", "030840000", "030890000", "031150000")


exclude_hr_change <- c("030770000", "030800000","030880000", "031070000", "031190000", "031250000" )


recoded_data <- raw_data_joined %>% 
  left_join(cpi, by = c("year" = "levy_year")) %>%
  left_join(eq_factor) %>%
  mutate(
    # year = as.factor(year) 
    agency_num = as.character(agency_num),
    
    agency_num = str_pad(agency_num, 9, "left", "0"), #add missing leading zeros
    first6 = str_pad(first6, 6, "left", "0"),
    home_rule_ind = as.character(home_rule_ind), #change reference category
    minor_type = as.factor(minor_type),
    reassess_year = as.character(reassess_year)) %>%
  
  mutate(cty_total_eav = as.numeric(cty_total_eav),    # taxable eav in cook and neighboring counties
         cty_cook_eav = as.numeric(cty_cook_eav),      # taxable EAV in cook county only
         pct_in_Cook = cty_cook_eav / cty_total_eav,   # to identify taxing agencies that cross county lines
         total_final_levy = as.numeric(total_final_levy), 
         total_reduced_levy = as.numeric(total_reduced_levy), # for non-HR agencies that have their levy reduced
         av = cty_cook_eav / eq_factor_final,
         first6_w_hr = str_c(first6, "_", home_rule_ind),
         agency_w_hr = str_c(agency_num, "_", home_rule_ind)
         ) %>%
  group_by(year, cty_total_eav, home_rule_ind) %>%
  mutate(summed_levy_sharetaxbase = sum(total_final_levy)+1) %>% # grouped by tax base and year
  ungroup() %>%
  group_by(year, first6, home_rule_ind) %>%
  mutate(summed_levy_first6 =sum(total_final_levy)+1) %>% # grouped by first 6 digits in agency number
  ungroup()
Joining with `by = join_by(year)`
Code 
recoded_data <- recoded_data %>% 
  
  filter(pct_in_Cook > 0.9) %>% #keep only agencies greater than 90% in Cook
  
  mutate(
    total_final_levy_4log = total_final_levy + 1,
    #Add 1 to total_final_levy to allow ln transformation.
    
    log_eav = log(cty_total_eav), # eav within Cook AND neighboring counties.
    log_levy = log(total_final_levy_4log),
    log_av = log(av),
    
    year = as.factor(year), # for plm()
    agency_num = as.factor(agency_num) # for plm()
    
  ) %>% 
  filter(minor_type != "SSA") # drops 5332 taxing agencies (agency-year combos)

## 7921 observations remain after removing SSAs ##
Code 
recoded_data %>% 
  ggplot (aes(x= major_type)) + 
  geom_bar() + 
  theme_classic() +  
  geom_text(aes(label = ..count..), stat = "count", vjust=-.1) + 
 # coord_flip() +
  labs(y="Number of taxing agencies", x= "'major_type' Categories")+ scale_y_continuous(label = scales::comma)

Code 
recoded_data %>% 
  ggplot (aes(x= minor_type)) + 
  geom_bar() + 
  theme_awm( ) + 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 2) + 
 coord_flip() +
  labs(title = "Taxing Agency Count: Before Dropping any Observations", y="Number of taxing agencies", x= "'minor_type' Categories")+
  scale_y_continuous(label = scales::comma)

Code 
# turn it into panel data!
# two way fixed effects will be used for agency and year.
panel_data <-pdata.frame(recoded_data, index = c("agency_num", "year"))


# need to detach dplyr because conflict w/ plm lag command
detach("package:dplyr", unload = TRUE)
Warning: 'dplyr' namespace cannot be unloaded:
  namespace 'dplyr' is imported by 'tidyr', 'gghighlight' so cannot be unloaded
Code 
panel_data$lag_totallevy <- plm::lag(panel_data$total_final_levy, 1)
panel_data$lag_cty_total_eav <- plm::lag(panel_data$cty_total_eav, 1)
panel_data$lag_av <- plm::lag(panel_data$av, 1)


panel_data$eav_lag1 <- plm::lag(panel_data$cty_total_eav, 1)
panel_data$eav_lag2 <- plm::lag(panel_data$cty_total_eav, 2)
panel_data$eav_lag3 <- plm::lag(panel_data$cty_total_eav, 3)
panel_data$eav_lag4 <- plm::lag(panel_data$cty_total_eav, 4)

panel_data$reassess_lag1 <- plm::lag(panel_data$reassess_year, 1)
panel_data$reassess_lag2 <- plm::lag(panel_data$reassess_year, 2)

#boss said leave redundant code.


library(dplyr)

Attaching package: 'dplyr'
The following objects are masked from 'package:plm':

    between, lag, lead
The following objects are masked from 'package:stats':

    filter, lag
The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union
Code 
panel_data <- panel_data %>% 
  
  mutate(eav_pct_change = (cty_total_eav - lag_cty_total_eav)/ lag_cty_total_eav,
         tfl_pct_change = (total_final_levy - lag_totallevy) / lag_totallevy,
         av_pct_change = (av - lag_av) / lag_av,
         up_down  = ifelse(eav_pct_change > 0, "increased", "decreased"))  %>%
  
  mutate(# eav_pct_change = ifelse(is.na(eav_pct_change), 0, eav_pct_change),
         home_rule_ind = as.factor(home_rule_ind),
         minor_type = as.factor(minor_type),
         major_type = as.factor(major_type),
         first6 = as.factor(first6))

detach("package:dplyr", unload = TRUE)
Warning: 'dplyr' namespace cannot be unloaded:
  namespace 'dplyr' is imported by 'tidyr', 'gghighlight' so cannot be unloaded
Code 
panel_data$lag_eav_pct_change1 <- plm::lag(panel_data$eav_pct_change, 1)
panel_data$lag_eav_pct_change2 <- plm::lag(panel_data$eav_pct_change, 2)

panel_data$lag_av_pct_change1 <- plm::lag(panel_data$av_pct_change, 1)
panel_data$lag_av_pct_change2 <- plm::lag(panel_data$av_pct_change, 2)



library(dplyr)

Attaching package: 'dplyr'
The following objects are masked from 'package:plm':

    between, lag, lead
The following objects are masked from 'package:stats':

    filter, lag
The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union
Code 
schools_panel <- panel_data %>% filter(major_type == "SCHOOL") %>%  
  mutate(lag_totallevy = as.numeric(lag_totallevy))

governments_panel <- panel_data %>% 
  mutate(lag_totallevy = as.numeric(lag_totallevy)) %>% 
  filter( major_type != "COOK COUNTY" & major_type != "SCHOOL") 
#revised 11/26 to filter "COOK COUNTY"

table(governments_panel$minor_type)

      BOND  COMM COLL       COOK ELEMENTARY       FIRE   GEN ASST     HEALTH 
        47          0          0          0        445        266        116 
     INFRA    LIBRARY       MISC   MOSQUITO       MUNI       PARK     POLICE 
       195       1522        116          0       1888       1391         72 
  SANITARY  SECONDARY        SSA   TOWNSHIP    UNIFIED      WATER 
       291          0          0        266          0          0
Code 
governments_panel <- as.data.frame(governments_panel)
schools_panel <- as.data.frame(schools_panel)
all_agencies <- as.data.frame(panel_data)

You can add options to executable code like this

Code 
#install.packages("ggThemeAssist")
library(ggThemeAssist)
Warning: package 'ggThemeAssist' was built under R version 4.5.3
Code 
governments_panel %>% 
  ggplot (aes(x= major_type)) + 
  geom_bar() + 
  theme_classic() +  
  geom_text(aes(label = ..count..), stat = "count", vjust=-.1) + 
 # coord_flip() +
  labs(title = "Governmental Panel Agency Count", y="Number of taxing agencies", x= "'major_type' Categories")+ scale_y_continuous(label = scales::comma)

Code 
governments_panel %>% 
  ggplot (aes(x= minor_type)) + 
  geom_bar()+ 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 3
            ) + 
 coord_flip() +
  labs(title = "Taxing Agency Count", subtitle = "Governmental Panel Agency Count", y="Number of taxing agencies", x= "'minor_type' Categories")+ scale_y_continuous(label = scales::comma) + 
  theme(axis.ticks = element_line(linetype = "blank"),
    panel.background = element_rect(fill = NA))  + 
  theme_awm( )

Code 
schools_panel %>% 
  filter(year == 2022) %>%
  ggplot (aes(x= minor_type)) + 
  geom_bar(fill = "red")+ 
  geom_text(aes(label = ..count..), stat = "count", hjust=-.1, size = 3) + 
 coord_flip() +
  labs(title = "Taxing Agency Count", subtitle = "Schools Panel Agency Count", y="Number of taxing agencies", x= "'minor_type' Categories")+ scale_y_continuous(label = scales::comma) + gghighlight(minor_type %in% c("SECONDARY", "ELEMENTARY"))  + 
  theme_awm( ) 
Warning: Tried to calculate with group_by(), but the calculation failed.
Falling back to ungrouped filter operation...
Warning: Tried to calculate with group_by(), but the calculation failed.
Falling back to ungrouped filter operation...
label_key: minor_type