README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(grattantheme)

base_coefs <- c("NAIRU")

```


# Recreating McKell Institute's super-wages model

This repo attempts to recreate the McKell Institute's super-wages model presented in their 2019 research paper [_'Does higher superannuation reduce workers’ wages?'_](https://mckellinstitute.org.au/app/uploads/Does-higher-superannuation-reduce-wages.pdf). 

It uses the statistical package R and publicly-available data. Three R scripts are used in order:

1. `01_get_data.R`: Retrive and tidy data from the ABS, OECD and RBA.
2. `02_prepare_data.R`: Combine data and generate variables for the model set.
3. `03_run_regressions.R`: Fit four models on each of AWOTE, AENA (wages) and AENA (social).

```{r, message = FALSE, warning = FALSE}
source("R/01_get_data.R")
source("R/02_prepare_data.R")
source("R/03_run_regressions.R")
```

They compile relevant data and produce tables of regression results for four model specifications.

The results from these regressions are summarised in the charts below. 

**Model 1: Recreation of McKell models**

```{r, echo = FALSE}
plot_model1 <- plot_summs(mck_awote_model, mck_aena_wages_model, mck_aena_social_model,
                          model.names = c("AWOTE",
                                          "AENA\n(wages)",
                                          "AENA\n(social)"),
                          colors = c(grattan_red, grattan_orange, grattan_yellow))

plot_model1
```


**Model 2: McKell model with pre-SG award**

```{r, echo = FALSE}
plot_model2 <- plot_summs(mck_awote_model_with_award2, 
                          mck_aena_wages_model_with_award2, 
                          mck_aena_social_model_with_award2,
                          model.names = c("AWOTE",
                                          "AENA\n(wages)",
                                          "AENA\n(social)"),
                          colors = c(grattan_red, grattan_orange, grattan_yellow))

plot_model2
```

**Model 3: McKell model with lagged SG one-quarter change**

```{r, echo = FALSE}
plot_model3 <- plot_summs(mck_awote_d1_lag2_model, 
                          mck_aena_wages_d1_lag2_model, 
                          mck_aena_social_d1_lag2_model,
                          model.names = c("AWOTE",
                                          "AENA\n(wages)",
                                          "AENA\n(social)"),
                          colors = c(grattan_red, grattan_orange, grattan_yellow))

plot_model3
```

**Model 4: The previous model with RBA specifications**

```{r, echo = FALSE}
plot_model4 <- plot_summs(rba_awote_model,
                          rba_aena_wages_model, 
                          rba_aena_social_model, 
                          model.names = c("AWOTE",
                                          "AENA\n(wages)",
                                          "AENA\n(social)"),
                          colors = c(grattan_red, grattan_orange, grattan_yellow))
plot_model4
```