Analysis · Finance
A 150-Cigarette Forecast Gap Was Still Too Uncertain to Call
By AWEI · AI-compiled · Published · Analysis prepared · 1 source · doi.org
A gap of 150 cigarettes per adult per year sounds substantial. Yet in a 2019 BMJ study by Steven J. Hoffman and colleagues, that difference between observed consumption and a modelled baseline could not establish whether a global tobacco treaty had changed consumption. The uncertainty around the baseline mattered more than the apparent gap.
The researchers evaluated the WHO Framework Convention on Tobacco Control, adopted in 2003. Their dataset covered 71 countries from 1970 to 2015, representing over 95% of global cigarette consumption. They measured annual cigarettes consumed per person aged 15 or older. A decline in this measure does not, by itself, establish falling total cigarette volume or market value.
Consumption per adult was already declining before adoption. One quasi-experimental analysis tested whether its annual changes shifted afterward. A second, covering 70 countries, used relationships between consumption and economic, political, industry and human development variables before 2003 to estimate a hypothetical path without the treaty. This was a retrospective comparison, with later explanatory data available, rather than a forecast issued in 2003.
By 2008, observed global consumption exceeded that hypothetical path by 150 cigarettes per adult annually. But exceeding the model’s 95% prediction interval would have required a gap of 364. Neither approach detected a statistically significant acceleration in the global decline. The positive gap therefore did not demonstrate that the treaty increased consumption; the uncertainty also allowed no difference or a reduction.
Those results depended on assumptions about continuing trends or stable relationships between variables. Other changes could conceal an effect or create a misleading association. Coverage was limited in some regions, other tobacco products were excluded, and delayed effects could escape detection. The analysis evaluated the treaty mechanism, not the effectiveness of each policy it recommended.
For assessing a market forecast claim, this offers a concrete check: compare the reported gap with its uncertainty range, then inspect the baseline assumptions and denominator. A precise-looking difference is not automatically evidence of an event’s impact.
Sources used for this article (1)
Direct links to the publisher reports used to prepare this article.
- Source 1
- Impact of the WHO Framework Convention on Tobacco Control on global cigarette consumption: quasi-experimental evaluations using interrupted time series analysis and in-sample forecast event modelling Steven J. Hoffman, Mathieu J. P. Poirier, Susan Rogers Van Katwyk …
