incentive_models_descriptives.qmd

---
format: 
  html:
    df-print: paged
    embed-resources: true
    theme: lumen
    code-fold: true
    code-line-numbers: true
    code-overflow: wrap
    toc: true
    toc-location: left
knitr: 
  opts_chunk:
    warning: true
    message: false
---

# 2011-2022 Panel Data Cross Tables

```{r setup}
#| output: FALSE

# Load packages

library(tidyverse)
library(corrr)
library(glue)
library(DT)
library(flextable)
library(kableExtra)
library(crosstable)
library(scales)

# Set table formatting defaults

set_flextable_defaults(theme_fun = theme_vanilla, 
                       padding = 2,
                       #line_spacing = 1,
                       big.mark = ","
                       )

options(DT.options = list())

FitFlextableToPage <- function(ft, pgwidth = 6){
  ft_out <- ft %>% autofit()
  ft_out <- width(ft_out, width = dim(ft_out)$widths*pgwidth /(flextable_dim(ft_out)$widths))
  return(ft_out)
}

comm_ind <- read_csv("./Output/comm_ind_PINs_2011to2022_timeseries.csv") 

## set variable types 
comm_ind <- comm_ind %>%
  mutate(across(c(class, improvement_ind, comparable_props, class_1dig, has_AB_exemp, fmv_NA_flag, in_tif, incent:leave_tif, exempt_flag), as.character) ) 

```


## Descriptive Stats for All Years Together

```{r}
library(modelsummary)
datasummary_skim(comm_ind)
datasummary_skim(comm_ind, type  = "categorical")
```

## Descriptive Stats for 2011 Strata of data

```{r}
commind2011 <- comm_ind %>% filter(year == 2011)

datasummary_skim(commind2011)
datasummary_skim(commind2011, type = "categorical")

```

## Descriptive Stats for 2022 Strata of data

```{r}

commind2022 <- comm_ind %>% filter(year == 2022)

datasummary_skim(commind2022)
datasummary_skim(commind2022, type = "categorical")
```

```{r}
datasummary_balance(~incent_change, data = commind2022, title = "PINs from 2022 by Incentive Change", output = "flextable")
```

```{r}
datasummary_balance(~landuse_change, data = commind2022, title = "PINs from 2022 by Landuse Change", output = "flextable")

```



# Correlations

```{r}
datasummary_correlation(commind2022, output = "flextable")
```

# Muni Check

Park Forest is there again! 118 Munis have commercial or industrial PINs during tax year 2022.

```{r}
comm_ind |> filter(year==2022) |> reframe(pincount= n(), fmv = sum(fmv), .by = clean_name) |> arrange(clean_name)
```

# Model Things

- Drop unnecessary variables from dataframe. Helps with speed of `pdata.frame()` command.    
- Set index of panel data forthe PIN and year (unit of observation and time) 
- Fixed effects will drop anything that doesn't change: Municipality, Triad, distance to Indiana, etc.   
   
- DV: FMV Growth since 2011? 

- 5760 PINs were exempt at least 1 year in the dataframe...