California's Proposition 99: The Synthetic DiD Panel

Downloads

Each dataset is available as a labeled Stata .dta and its source file.

⇩ Download all data (ZIP)stata_codebook.do

Run stata_codebook.do in Stata once to attach long-form per-variable notes to the .dta files.

Load directly in code

Every file loads straight from GitHub (raw URLs). Swap the file name to load any dataset.

Stata

* Stata 14+ : `use` reads an https URL directly
global BASE "https://raw.githubusercontent.com/cmg777/starter-academic-v501/master/content/post/stata_sdid/data/"
use "${BASE}prop99_example.dta", clear
describe
notes

Python

!pip install -q pyreadstat
import pandas as pd
BASE = "https://raw.githubusercontent.com/cmg777/starter-academic-v501/master/content/post/stata_sdid/data/"
df = pd.read_stata(BASE + "prop99_example.dta")

# load every dataset at once
files = ["prop99_example"]
data = {f: pd.read_stata(BASE + f + ".dta") for f in files}

# pyreadstat (richest metadata) reads LOCAL files -> download first
import pyreadstat, urllib.request
urllib.request.urlretrieve(BASE + "prop99_example.dta", "prop99_example.dta")
df, meta = pyreadstat.read_dta("prop99_example.dta")

Copy and paste this snippet in Google Colab app. https://colab.research.google.com/notebooks/empty.ipynb

R

# R : haven::read_dta auto-downloads an https URL
library(haven)
BASE <- "https://raw.githubusercontent.com/cmg777/starter-academic-v501/master/content/post/stata_sdid/data/"
df <- read_dta(paste0(BASE, "prop99_example.dta"))

Overview & sources

Companion data for a hands-on Stata tutorial on synthetic difference-in-differences (SDID), applied to re-evaluate California's Proposition 99 — the 1988 ballot measure that raised the cigarette excise tax by 25 cents a pack and funded an anti-smoking campaign. The file is the canonical strongly balanced panel distributed with the sdid package (originally from Abadie, Diamond & Hainmueller 2010, and used by Arkhangelsky et al. 2021): 39 US states observed annually from 1970–2000 — 1,209 observations — with annual cigarette sales in packs per capita as the sole outcome. California is the single treated unit; the policy bites from 1989 onward. The post writes DiD, synthetic control, and SDID as one weighted two-way fixed-effects regression and estimates the ATT of Proposition 99 with the sdid command, cross-checking synthetic control against synth2.

Data sources

Cite this data

Please cite this dataset as follows.

APA

Mendez, C. (2026). Synthetic Difference-in-Differences (SDID) in Stata: Re-evaluating California's Proposition 99 [Data set]. https://carlos-mendez.org/post/stata_sdid/

Arkhangelsky, D., Athey, S., Hsiao, D. A., Imbens, G. W., & Wager, S. (2021). Synthetic Difference-in-Differences. American Economic Review, 111(12), 4088–4118. https://doi.org/10.1257/aer.20190159 Abadie, A., Diamond, A., & Hainmueller, J. (2010). Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California's Tobacco Control Program. Journal of the American Statistical Association, 105(490), 493–505. https://doi.org/10.1198/jasa.2009.ap08746

BibTeX

@misc{mendez2026statasdid,
  author       = {Mendez, Carlos},
  title        = {Synthetic Difference-in-Differences (SDID) in Stata: Re-evaluating California's Proposition 99},
  year         = {2026},
  howpublished = {\url{https://carlos-mendez.org/post/stata_sdid/}},
  note         = {Data set}
}

@article{arkhangelsky2021synthetic,
  author  = {Arkhangelsky, Dmitry and Athey, Susan and Hsiao, David A. and Imbens, Guido W. and Wager, Stefan},
  title   = {Synthetic Difference-in-Differences},
  journal = {American Economic Review},
  volume  = {111}, number = {12}, pages = {4088--4118}, year = {2021},
  doi     = {10.1257/aer.20190159}
}
@article{abadie2010synthetic,
  author  = {Abadie, Alberto and Diamond, Alexis and Hainmueller, Jens},
  title   = {Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California's Tobacco Control Program},
  journal = {Journal of the American Statistical Association},
  volume  = {105}, number = {490}, pages = {493--505}, year = {2010},
  doi     = {10.1198/jasa.2009.ap08746}
}

Variable explorer search & filter all 4 variables

Type to filter by name or label, or use the chips to filter by type. Each row shows a mini distribution. Click a header to sort.

Cross-file variable index

Which file each variable appears in (● = present).

Construction & formulas

Every estimator in the post is the same weighted two-way fixed-effects (TWFE) regression over packspercapita (Y) and the treatment indicator treated (W), changing only the weights — the unifying view of Arkhangelsky et al. (2021).

• Synthetic DiD objective: (τ̂, μ̂, α̂, β̂) = argmin Σ_i Σ_t (Y_it − μ − α_i − β_t − W_it·τ)² · ω̂_i · λ̂_t — a TWFE regression weighted by unit weights ω̂_i and time weights λ̂_t. α_i = state fixed effect; β_t = year fixed effect; τ = the ATT.
• DiD: the special case with uniform ω and λ (plain TWFE) — credibility rests entirely on parallel trends versus all controls.
• Synthetic control: keeps optimized unit weights ω but drops the time weights λ and the unit fixed effects α — so it must match California's pre-period level and trend.
• Unit weights ω: chosen so the weighted controls track California's pre-period path (with intercept ω₀ and a ridge penalty ζ² for stability).
• Time weights λ: chosen so weighted pre-period years line up with the post-period — here SDID places all pre-period weight on 1986–1988.

Estimand (ATT). τ = (1 / (N_tr·T_post)) · Σ_{i: W_i=1} Σ_{t > T_pre} [ Y_it(1) − Y_it(0) ] — the effect of Proposition 99 on California over the post-1988 period, where the counterfactual Y_it(0) is never observed and each method imputes it differently. Here N_tr = 1 (California). Because California was not randomly assigned, this is an observational design.

The datasets

Switch datasets with the tabs. Each shows the full variable dictionary plus a sortable statistics table with mini distributions and data coverage.

Expand all datasets expand to search (Ctrl/⌘+F) or print across all datasets

Known limitations & caveats

• Single treated unit. Only 12 of 1,209 observations are treated (California, 1989–2000). With one treated unit, placebo (permutation) inference is the only valid procedure — the jackknife is undefined and the bootstrap is unreliable; statistical power is inherently limited.
• Outcome-only panel. The file carries no income, price, or demographic covariates — only cigarette packs per capita. This is deliberate: synthetic control and SDID see exactly the same information, so any difference in their answers comes from the estimator, not from different predictors.
• Strongly balanced. Every state is observed in every year 1970–2000 with no gaps; this balance is required by all three estimators in the post.
• Observational, not experimental. California was not randomly assigned Proposition 99; identification assumes a stable comparison group, no large contemporaneous shock unique to California in 1989, and no cross-state spillovers (e.g., border cigarette purchases contaminating donor states).
• Distributed dataset. Values are byte-faithful to the canonical prop99_example.dta shipped with the sdid package; this data dictionary adds only metadata (variable + value labels).