1Synthetic control methods estimate what would have happened to a single treated unit, like a country, by weighting together untreated units.

2In 1994, the newly elected government in South Africa began aggressively raising cigarette excise taxes. By 2004, the real excise tax had climbed by 249%, and the average retail price by 110%. Per capita consumption fell, and the natural conclusion is that the tax policy caused the decline.

3But consumption was already falling before 1994. The economy was contracting, public health awareness was building, and modest regulatory steps were underway. A simple before-and-after comparison would credit the tax for declines that were happening anyway. The question every policy evaluator faces in moments like this is, what would have happened without the intervention?

4We can't observe that directly, and no single country is a clean comparison. Every other low- or middle-income country differs from South Africa on demographics, economy, baseline consumption, and regulatory environment, and matching and regression adjustment aren't built for a setting with one treated unit. Synthetic control methods are. We construct a counterfactual by weighting together untreated units so the weighted average closely tracks the treated unit's trajectory before the intervention. If it tracks well before, it provides a credible estimate of what would have happened after.

5The procedure runs in five steps. First, identify the treated unit and the intervention date. Second, assemble a donor pool of untreated units that are plausible building blocks. Third, choose matching predictors, the variables believed to drive the outcome. Fourth, solve for weights that minimize the gap between the treated unit and the synthetic control in the pre-period. The weights are non-negative and sum to one, so the synthetic control is an interpolation among real units, not an extrapolation beyond them. Fifth, project it forward, and the gap after the intervention is the estimated effect.

6One team built a synthetic South Africa from a donor pool of 24 countries that hadn't implemented large-scale tobacco control during the study period. They matched on the real price of cigarettes, real GDP per capita, per capita alcohol consumption, the proportion of adults in the population, and lagged cigarette consumption. The optimization gave positive weights to five countries: Brazil at 47.6%, Argentina at 27.6%, Chile at 14.6%, Tunisia at 9.4%, and Romania at 0.7%. The rest received zero.

7From 1980 to 1993, this weighted combination tracked South Africa's cigarette consumption almost exactly. The average annual difference was about one cigarette per capita, against a baseline of roughly 1,000. After the tax increases took effect, the two series diverged. By 1995, the first full post-policy year, South Africa's consumption was 38 cigarettes per capita below the synthetic counterfactual, about 4%. By 2004, the gap had widened to 290 cigarettes, 36% below what consumption would have been without the tax increases.

8A synthetic control study lives or dies on two questions. Did the pre-intervention fit actually match? And could the estimated effect be driven by something other than the intervention? The first is visual. We should be able to see the two series tracking closely before the intervention, and if they diverge in the pre-period, the gap after it isn't a clean treatment effect estimate.

9The second question comes from sensitivity checks, and the clearest is leave-one-out. We re-estimate the synthetic control after dropping each donor unit in turn. If the result depended on a single country, say Brazil, which carried nearly half the weight, dropping it should make the estimated effect collapse. It didn't. The pattern of divergence held across all five leave-one-out specifications, and synthetic South Africa's trajectory was similar under an alternative donor pool.

10A third common check is the in-space placebo. We apply the same procedure to each donor country, treating it as if it were the treated unit. If the actual treated country's estimated effect is unusually large compared to the placebo effects, that's evidence the result isn't just noise. This permutation logic is how synthetic control studies typically argue for statistical significance, because the method doesn't yield a confidence interval the way regression does.

11Synthetic control fits a specific gap in the quasi-experimental toolkit. We use it when one unit, usually a country, region, or state, experienced an intervention, and we have a credible donor pool of comparable untreated units along with several years of pre-intervention data. National or sub-national policies are the natural target: tobacco taxes, abortion restrictions, and large-scale funding shifts.

12One study used a variant of the approach to estimate the mortality impact of U.S. global health investments. It aggregated 16 recipient countries into a single treated unit and constructed the synthetic control from 19 nonrecipient countries. The conservative estimate was 1.0 to 1.3 million deaths averted among women of reproductive age between 2009 and 2019.

13The method's deepest limitation is shared by every counterfactual estimator. If something else shifted the treated unit's trajectory at the same time as the intervention, the estimated effect captures both. The South Africa team handled this by ending their analysis in 2004, before broader international tobacco-control efforts began affecting many of the donor countries and contaminating the comparison.

14Three additional constraints matter. Pre-intervention fit has to be good, and without it, the method isn't reliable. The donor pool has to be rich enough that a combination of donor units can plausibly approximate the treated unit, and if the treated unit is too unusual relative to anything available, the method can't produce a good match. And inference is permutation-based, not classical, so be cautious of synthetic control studies that report standard p-values without explaining how they arrived at them.