18 June 2021

Considerations when using sales volume data to estimate tobacco product substitution effects

Below is the transcript of the video:

Hello, my name is Ondrej Koumal, and I work as Director Regulatory Communications at Philip Morris International. It is my pleasure to be here with you today at this year's Global Forum on Nicotine. In this short video I'd like to explore some of the considerations we need to take into account when using sales volume data to estimate tobacco product substitution effects. 

One of the most important questions when it comes to potentially less harmful tobacco and nicotine products is their impact on cigarette consumption. Are they used to replace cigarettes or do people use them in addition to smoking?

Disco- News - Conference - Article-2

Click here to watch the other presentations

Population surveys are the most appropriate tool to examine this question. However, researchers may also analyze tobacco product sales volumes to gain information about any substitution effects. While this may sound lazy, it’s anything but. 

First, we need to obtain sales volume data. For example, it can come from a market research agency that tracks sales from retailers to consumers in a subset of points of sales, which is then extrapolated to the larger geography. 

Or, from fiscal authorities as manufacturers pay excise taxes based on the volume of tobacco products that they sell. 

Or, from other third-party provider, which collects data on the sales from tobacco companies and distributors to the trade channel. 

This type of data is typically readily available for traditional tobacco products, such as cigarettes. 

But cigarettes are not the only tobacco product available on the market today. We also have roll-your-own tobacco, pipe tobacco, cigars of different shapes and sizes. In some countries, oral tobacco products are very popular. Finally, we also have novel tobacco products such as e-cigarettes or heated tobacco products. And most recently, nicotine pouches.  

To be able to estimate the evolution of the marketplace, we ideally need sales volume data for all these product categories over time. But here’s a problem. The system that we described earlier and that exists for cigarettes, often does not exist for other tobacco and nicotine product categories. Therefore, the estimates for tobacco products other than cigarettes are often much less precise as they incorporate various extrapolations and assumptions. 

And there’s more. Not all products are available in a stick format. Therefore, if we want to sum up the total market or examine substitution effects, we have to convert all these different tobacco product formats into a common equivalent – for example, a cigarette stick. As there is no universally accepted formula, every manufacturer uses a different one to translate roll-your-own tobacco, pipe tobacco, or snus, or e-cigarette, or a heated tobacco product into a cigarette equivalent. Unsurprisingly, none of these conversion formulas is perfect as it can be very difficult to quantify smoking using, well … not smoking, or smoking a combustible product that is not a factory manufactured cigarette.  

But the complexity does not end here. Even if we collected the sales volumes of all the product categories using reliable methods, and we can now look at the trends over time, we may still be unable to fully understand the substitution effects. 

Sales volumes can go up or down as consumers increase or decrease their consumption with the underlying total number of consumers remaining unchanged. 

What are the underlying demographic indicators? Is the population increasing, decreasing, or stable? What if our conversion rate does not reflect actual consumer behavior? What is the proportion of illicit trade? In some European countries, it is estimated that around one fifth of the domestic cigarette consumption is supplied from illicit sources, and therefore not captured in legal sales volume estimates. Changes in illicit trade can have a significant effect on legal volumes. 

Then there is tourism and legal cross-border sales. Cross-border sales in some countries represent a significant proportion of legal domestic sales. 

All of these factors impact sales volumes over time and may make precise quantification of substitution effects challenging, in particular for a product that has been on the market for a short period of time. 

Yet, despite its limitations, sales volume data can provide valuable insights that can contribute to the evaluation of the early population level impact of new tobacco products on public health. 

To find out more, read our case study by clicking on the hyperlink shown on the screen.  I hope you enjoyed my presentation and thank you very much for joining me today. 

Click here to watch the other presentations.

Related Articles