undidr

Difference-in-differences for unpoolable data.

https://github.com/ebjamieson97/undidr

Repository

Difference-in-differences for unpoolable data.

README.Rmd

---
output: github_document
---



```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
```

# undidR


[![R-CMD-check](https://github.com/ebjamieson97/undidR/actions/workflows/R-CMD-check.yaml/badge.svg)](https://github.com/ebjamieson97/undidR/actions/workflows/R-CMD-check.yaml)
[![Codecov test coverage](https://codecov.io/gh/ebjamieson97/undidR/graph/badge.svg)](https://app.codecov.io/gh/ebjamieson97/undidR)
[![CRAN status](https://www.r-pkg.org/badges/version/undidR)](https://CRAN.R-project.org/package=undidR)


The **undidR** package provides the framework for implementing difference-in-differences with unpoolable data (UNDID) developed in [Karim, Webb, Austin, and Strumpf (2024)](https://arxiv.org/abs/2403.15910v2). UNDID is designed to estimate the average treatment effect on the treated (ATT) in settings where data from different silos cannot be pooled together (potentially for reasons of confidentiality). The package supports both common and staggered adoption scenarios, as well as the optional inclusion of covariates. Additionally, **undidR** incorporates a randomization inference (RI) procedure, based on [MacKinnon and Webb (2020)](https://doi.org/10.1016/j.jeconom.2020.04.024), for calculating p-values for the UNDID ATT.

See the image below for an overview of the **undidR** framework:
```{r schematic, echo=FALSE, fig.align='center', out.width='90%'}
knitr::include_graphics("vignettes/figures/undidR_schematic.png")
```

## Installation

You can install the stable release version of `undidR` from [CRAN](https://CRAN.R-project.org/package=undidR) with:

```{r, eval=FALSE}
install.packages("undidR")
```

## Examples

For a set of highly detailed examples see the package vignette using:
```{r, eval=FALSE}
vignette("undidR", package = "undidR")
```

The following code chunks show some basic examples of using **undidR** at each of its three stages.

### Stage One: `create_init_csv()` & `create_diff_df()`
```{r example}
library(undidR)
init <- create_init_csv(silo_names = c("73", "46", "54", "23", "86", "32",
                                       "71", "58", "64", "59", "85", "57"),
                        start_times = "1989",
                        end_times = "2000",
                        treatment_times = c(rep("control", 6),
                                            "1991", "1993", "1996", "1997",
                                            "1997", "1998"))
init

init_filepath <- normalizePath(file.path(tempdir(), "init.csv"),
                               winslash = "/", mustWork = FALSE)
empty_diff_df <- create_diff_df(init_filepath, date_format = "yyyy",
                                freq = "yearly",
                                covariates = c("asian", "black", "male"))
head(empty_diff_df, 4)
```

### Stage Two: `undid_stage_two()`
```{r}
silo_data <- silo71
empty_diff_filepath <- system.file("extdata/staggered", "empty_diff_df.csv",
                                   package = "undidR")
stage2 <- undid_stage_two(empty_diff_filepath, silo_name = "71",
                          silo_df = silo_data, time_column = "year",
                          outcome_column = "coll", silo_date_format = "yyyy")
head(stage2$diff_df, 4)
head(stage2$trends_data, 4)
```

### Stage Three: `undid_stage_three()`
```{r}
dir_path <- system.file("extdata/staggered", package = "undidR")
results <- undid_stage_three(dir_path, agg = "silo", covariates = TRUE,
                             nperm = 501)
results
```

Owner

GitHub Events

Packages