Article Text
Abstract
Introduction E-cigarette taxes have been enacted by 30 states through April 2020. E-cigarette tax schemas vary, in contrast to cigarette taxes in the USA that are levied almost exclusively as excise taxes per pack. Some states use excise taxes on liquid and containers, others ad valorem taxes on wholesale prices and others sales taxes. It is therefore difficult to understand the relative magnitudes of these e-cigarette taxes and the overall e-cigarette tax size relative to the cigarette tax size.
Objective To create and publish a database of state and local quarterly e-cigarette taxes from 2010 to 2020, standardised as the rate per millilitre of fluid.
Methods Using Universal Product Code-level e-cigarette sales from the NielsenIQ Retail Scanner Data along with e-cigarette product characteristics collected from internet searches and visits to e-cigarette retailers, we develop a method to standardise e-cigarette taxes as an equivalent average excise tax rate measured per millilitre of fluid.
Results In 2020, the average American resided in a location with $3.08 in cigarette taxes and $0.34 in e-cigarette taxes (assuming 1 pack=0.7 fluid mL).
Conclusions The public availability of this state and local standardised e-cigarette tax data will allow tobacco control researchers to study the relationship between e-cigarette taxes and tobacco and related outcomes more effectively.
- electronic nicotine delivery devices
- taxation
- economics
Statistics from Altmetric.com
Introduction
E-cigarettes were first imported into the USA in 2006. By 2020, 19.6% of high school students used e-cigarettes in the last 30 days, though only one-third used them at least 20 of the last 30 days.1 Adult 30-day e-cigarette use rates are lower at 4.5% in 2019.2 Several studies suggest higher e-cigarette tax rates reduce e-cigarette use.3–5 As of April 2021, 30 states have enacted e-cigarette taxes,6 but unlike cigarette taxes, which are usually levied as a fixed tax amount per pack of cigarettes, e-cigarette taxes are levied as unit-specific excise taxes (ie, fixed tax amount per fluid millilitre (mL) and/or container), ad valorem taxes (ie, per cent tax on the e-cigarette wholesale price) and as sales taxes (ie, per cent tax on the retail price). It is therefore challenging to understand the relative magnitudes of e-cigarette taxes across states with different tax schemas.
We introduce a publicly available dataset of standardised e-cigarette taxes, measured as an average tax rate per mL of fluid, the most common unit for e-cigarette excise taxes (see table 1). We also compare e-cigarette taxes and cigarette taxes over time.
Methods
Standardising e-cigarette taxes requires a variety of assumptions. Our methodology attempts to derive variation from legislated tax changes and minimise the influence of factors such as changes in e-cigarette product offerings and consumer preferences that are otherwise difficult to control for in regression modelling and may introduce bias. We also provide alternative tax measures with relaxed assumptions.
We use NielsenIQ Retail Scanner Data (NRSD) to determine retail prices. The NRSD provides scanner transaction data at the store-by-Universal Product Code (UPC)-by-week level for approximately 30 000–50 000 stores annually, which represents over 20% of national sales in food, drug and mass merchandise stores but only a small share (2%) of convenience store sales. The NRSD categorises specific UPC codes as e-cigarettes beginning in 2013, whereas in earlier years we identify e-cigarette UPC codes grouped with other tobacco products by backdating known e-cigarettes from year 2013. We identified and linked product characteristics from internet searches and visits to e-cigarette retailers for 98.6% of the NRSD e-cigarette sales volume in 2013, or 97.7% from 2011 to 2019.4 7
We convert ad valorem and sales taxes to their excise tax equivalencies for each tax jurisdiction, defined as a state or large population centre (ie, Cook County, Illinois and Montgomery County, Maryland, USA) that levies taxes. There is currently no federal e-cigarette tax.
Ad valorem taxes
We standardise taxes for the 12 jurisdictions using ad valorem taxes by the end of 2020:
(1)
where s indexes a tax jurisdiction on a quarterly basis t.
The variable is the quarterly tax rate reported by the Centers for Disease Control and Prevention (CDC) State System for each state (or Montgomery County, Maryland, USA that also uses an ad valorem tax). As shown in table 1, the ad valorem taxes range from a low of 15% (of wholesale price) in Illinois to a high of 96% in Washington DC. To determine the wholesaler’s selling price (wholesale price per mL), we need to determine the retail price per fluid mL and subtract an estimated retailer markup.
We begin by using the NRSD to calculate the retail price per fluid mL for all e-cigarettes (including disposables, starter kits and replacement cartridges) after omitting zero-nicotine products and products for which fluid or container amounts could not be identified from hand-collected data on UPC product characteristics.4 7 Next, we calculate the sales-weighted retail price per fluid mL for 23 tax jurisdictions not adopting e-cigarette taxes by the end of 2020, which provides a consistent sample of jurisdictions and removes the need to make assumptions regarding pass-through of taxes to prices in jurisdictions that adopt taxes.
We find an average sales-weighted e-cigarette retail price of $4.04 per fluid mL in 2013 for the 23 non-tax jurisdictions. We calculate a wholesale price averaged just for the year 2013 (the first year e-cigarettes are categorised by the NRSD) to reduce the influence of time-varying factors, including the effect that state-level tax changes could have on wholesale prices in other states.
Next, using the $4.04 per fluid mL measure of retail price, we subtract an estimated retailer markup of 35%, which is the markup rate used by a major e-cigarette company for nicotine-containing cartridges based on company purchasing form information reviewed by the authors. This 35% retailer markup yields a wholesale price of $2.63 per fluid mL ($4.04×65%). We alternatively subtract a 20% markup rate used in cigarette research,8 yielding a wholesale price of $3.23 and, hence, a larger ad valorem tax rate. We speculate larger markups were used for e-cigarettes to incentivise store owners to offer limited shelf space for newer e-cigarette products that have less proven profitability than cigarettes.
We examine the extent to which wholesale prices vary across time and geography in two ways. First, we check whether e-cigarette retail prices vary across the 23 non-taxing jurisdictions at a given point in time, which could reflect geographical differences in wholesale prices or retailer markups. We compare retail prices as recorded by the NRSD for the top three best-selling e-cigarette products (by liquid volume) sold across these jurisdictions for the store type that sold the most e-cigarettes (eg, drug stores) in that particular year. For example, in 2011 the most popular product was a Blu e-cigarette and the most popular channel was convenience stores. For this product in these stores, the average sales-weighted price was $8.18 across the 23 non-tax jurisdictions, and the SD was only $0.03, or approximately 0.4% of the retail price mean. Continuing this exercise for the top three products across each of the 9 years provides 27 estimates in total and includes e-cigarettes sold by five different companies. The average SD as a per cent of the price was only 1.6% across these estimates. This provides evidence that large volume e-cigarette companies use national pricing strategies rather than regional pricing strategies and justifies our use of an average wholesale price in our preferred e-cigarette tax measure.
Second, we provide standardised e-cigarette tax rates using a time-varying measure of estimated wholesale prices. Here, instead of using the 2013 wholesale price, we calculate the average wholesale price for each quarter (again in the 23 jurisdictions). The NRSD is not yet available for 2020 as of October 2021, so for this standardised tax, we use Q4 2019 NRSD values and 2020 e-cigarette tax rates to project 2020 standardised tax values.
Other taxes
For two states using sales taxes, we multiply the average retail price per fluid mL of liquid nicotine in 2013 for the 23 jurisdictions by the sales tax rate to determine the tax per fluid mL.
(2)
Similarly, taxes per container are standardised using the tax rate per container and the average containers per fluid mL in 2013 for the 23 jurisdictions.
(3)
The average retail prices per mL and containers per mL for these jurisdictions were $4.05 and $1.02, respectively, in 2013, or $4.49 and $0.99 from 2011 to 2019. For both equations (2) and (3), our time-varying standardised tax uses time-varying retail prices and containers per mL, rather than in 2013.
Results
We calculate population-weighted national average cigarette taxes and e-cigarette taxes and plot them over time in figure 1. We sum the $1.01 federal cigarette tax, state-level cigarette tax provided by the CDC State System and local cigarette taxes provided by the American Non-Smokers Rights Foundation to create total cigarette tax.
To compare cigarette and e-cigarette tax rates, we assume that 0.7 fluid mL, the amount contained in a Juul pod, is equivalent to one pack of cigarettes.9 While this conversion factor is a useful way to compare current taxes, it likely underestimates the e-cigarette tax relative to the cigarette tax in earlier years when salt-based high-nicotine concentration e-liquid products were not yet on market. We find that in the fourth quarter of 2020 the population-weighted average cigarette tax was $3.08 per pack. The average e-cigarette tax for our preferred standardised tax measure was $0.34 per pod in the fourth quarter of 2020, rising from $0.22 in 2018 and $0.26 in 2019.
The online supplemental tables provide actual and standardised e-cigarette tax rates disaggregated for each tax jurisdiction and quarter, through 2020. Online supplemental table 2 lists our preferred e-cigarette tax using the time-invariant estimated tax units assuming a 35% markup; online supplemental table 3 lists e-cigarette taxes using time-invariant estimated tax units assuming a 20% markup and online supplemental table 4 lists e-cigarette taxes using time-varying estimated tax units assuming a 35% markup. Figure 2 shows similarity in the two standardised e-cigarette taxes using a 35% markup except for some divergence occurring in year 2018, potentially due to Juul’s rapid ascendance, before returning to being quite similar in years 2019 and 2020. The decline in e-cigarette taxes from 2018 to 2019 using our time-varying measure, despite legislated tax increases occurring during this period of time, illustrates our rationale for preferring the use of time-invariant units for standardising e-cigarette taxes.
Supplemental material
Discussion
Several studies use earlier versions of the e-cigarette tax data discussed in this study to find that e-cigarette tax rates raise e-cigarette prices,4 10 reduce e-cigarette sales and use, and increase cigarette sales and use.3–5 11 Other researchers are now free to use these data through 2020 in their own studies. We encourage best practices for quasi-experimental research, such as using fixed-effects models for tax jurisdiction and time, and testing parallel trends and other relevant assumptions.12
The Tax Burden on Tobacco (TBOT) publications have long been regarded as a premier source of cigarette price and tax data. The TBOT uses a mail survey of approximately 15 000–16 000 retailers in recent years.13 Our sampling frame compares favourably by using 35 000 stores, and scanner data also provide more accuracy in retail prices than survey data. Our methodology to standardise e-cigarette taxes levied in different ways provides a companion resource to the TBOT for e-cigarette tax data going forward.
In 2020, e-cigarettes were taxed at around 11% of the rate at which cigarettes were taxed (13% assuming a 20% retailer markup). Setting e-cigarette taxes at parity with cigarette taxes is not necessarily optimal; for example, several editorials recommend a strategy of taxing e-cigarettes proportionate to their risk vis-à-vis cigarettes.14–16 As the relative risk of e-cigarettes and cigarettes becomes clearer, the data presented here can support policy decisions about setting e-cigarette tax rates.
What this paper adds
E-cigarette tax schemas vary, making it difficult to understand the relative magnitudes of these e-cigarette taxes and the overall e-cigarette tax size relative to the cigarette tax size.
This study creates a database of state and local quarterly e-cigarette taxes from 2010 to 2020, standardised as the rate per milliliter of fluid.
The public availability of this state and local standardised e-cigarette tax data will allow tobacco control researchers to study the relationship between e-cigarette taxes and tobacco and related outcomes more effectively.
Ethics statements
Patient consent for publication
Ethics approval
This study does not involve human participants.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors CC, EN and NT collected the e-cigarette UPC code data. MFP and SP analysed the Nielsen Retail Scanner Data. All authors were involved in developing the methods and writing/revising the paper.
Funding This study was funded by the National Institute on Drug Abuse of the National Institutes of Health (R01DA045016). MFP and SP are supported in their work on this manuscript by R01DA045016 from the National Institute on Drug Abuse of the National Institutes of Health. Over the past 3 years, authors report funding from the National Institutes of Health (MFP and NT), American Cancer Society (MFP), University of Kentucky Institute for the Study of Free Enterprise (MFP), Robert Wood Johnson Foundation (NT), and for COVID-19-related consulting/expert testimony for the United South and Eastern Tribes and American Civil Liberties Union (CC). NT worked on this project while a faculty member at Bates College.
Disclaimer Sponsors had no role in the design and conduct of the study. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Institutes of Health. Researcher(s) own analyses calculated (or derived) based in part on data from Nielsen Consumer and marketing databases provided through the NielsenIQ Datasets at the Kilts Center for Marketing Data Center at the University of Chicago Booth School of Business.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.