Abstract
Background and Aims:
Historically, alcohol taxation has been one of the best strategies to reduce drinking and related harms in the general population, and prior analyses support that increased alcohol taxes are associated with fewer harms related to drinking during pregnancy. Here we examine how alcohol taxes are related to preterm and low birthweight births.
Methods:
Birth certificate data for 101,191,080 births conceived between January 1993 - March 2019 come from the US National Vital Statistics System. Outcomes were preterm birth (< 37 weeks) and low birthweight (< 2500 grams). Time-varying state-level alcohol tax policies were beverage-specific volume-based, sales, and ad valorem taxes. Final regression models adjusted for individual and state covariates, state trends, and clustered standard errors by state.
Results:
Only sales taxes were consistently related to birth outcomes. Having a beer sales tax was significantly related to lower odds of preterm birth (OR = 0.94, 95% CI: 0.90, 0.98) and lower odds of low birthweight (OR = 0.98, 95% CI: 0.96, 0.99). Similarly, having a wine sales tax was significantly related to lower odds of preterm birth (OR = 0.94, 95% CI: 0.90, 0.98) and lower odds of low birthweight (OR = 0.97, 95% CI: 0.95, 0.99). Finally, having a spirits sales tax was also significantly related to lower odds of preterm birth (OR = 0.95, 95% CI: 0.91, 0.98) and lower odds of low birthweight (OR = 0.97, 95% CI: 0.95, 0.99).
Conclusions:
State-level beverage-specific sales taxes are associated with fewer adverse birth outcomes. This study provides additional evidence that increased alcohol taxes may be relevant for reducing adverse effects related to pregnant people’s alcohol consumption.
Keywords: Alcohol taxes, beer tax, wine tax, spirits tax, birth outcomes, low birthweight, preterm birth
INTRODUCTION
Alcohol use during pregnancy remains prevalent in the U.S.: from 2018–2020, 14% of pregnant individuals reported current drinking and 5% reported binge drinking in the past 30 days (Gosdin, 2022). Drinking alcohol during pregnancy can lead to a range of harms, including fetal alcohol spectrum disorders (FASD), preterm birth, low birthweight, miscarriage, and stillbirth (DeJong et al., 2019). These harms remain prevalent in the U.S., with estimates showing as many as 71 infants born with FASD per every 1,000 live births in some regions (May et al., 2021). Despite widespread implementation of state-level policies regarding alcohol use during pregnancy (Roberts et al., 2017), the prevalence of drinking during pregnancy in the U.S. has not changed since the 1990s (Centers for Disease Control and Prevention, 1997).
General population alcohol policies could be a better strategy for reducing adverse outcomes related to drinking during pregnancy than policies that focus on pregnant people, as pregnancy-specific alcohol policies have shown to be largely ineffective and harmful (Roberts, 2023; Roberts et al., 2019, 2023; Schulte et al., 2024; Subbaraman et al., 2018, 2023). For example, a comprehensive study of U.S. state-level policies regarding alcohol use among pregnant people found that these policies are not associated with better birth outcomes, and that some are actually associated with more low birthweight and preterm births, less prenatal care utilization, and more infant injuries (Roberts et al., 2023; Subbaraman et al., 2018). Recommendations from these studies include examining general population policies, such as government monopolies and policies regulating alcohol availability and pricing, as these may reduce alcohol consumption during pregnancy and related harms (Roberts, 2023; Subbaraman et al., 2018). One recent study found that government monopolies on spirits sales and restrictions on gas station spirits sales were each associated with reduced infant morbidities and injuries (Schulte et al., 2024). Government monopolies on spirits and restrictions on alcohol sales in gas stations are also related to less drinking among women of reproductive age (18–44 years), providing evidence for the mechanisms of how these policies impact downstream birth outcomes (Subbaraman et al., 2023).
Historically, alcohol taxation has been one of the best strategies to reduce drinking and related harms in the general population (Babor, 2010; Babor et al., 2023; Nelson et al., 2013; Wagenaar et al., 2009, 2010). Studies also show that women are particularly sensitive to alcohol prices (Ayyagari et al., 2009; Meier et al., 2010). In the U.S., higher taxes on beer and spirits are associated with reduced alcohol consumption and related problems, especially among African American and Hispanic women (Subbaraman et al., 2020). Prior analyses also support that alcohol taxes are relevant for outcomes related to drinking during pregnancy (Luukkonen et al., 2023; Subbaraman et al., 2024; Zhang, 2010). In Finland, after a 2004 tax cut decreased average off-premises alcohol prices by 33%, the probability of low birthweight (+1.5 percentage points) and preterm birth (+2.0 percentage points) increased among low-income women (Luukkonen et al., 2023). Similarly, small increases in beer, wine, and spirits taxes each were associated with decreases in the incidence of low-birthweight births in the U.S. from 1985–2002 (Zhang, 2010). More recent analyses of state-level taxes in the U.S. found that specific components of alcohol taxes –i.e., wine and spirits sales and volume-based taxes – were related to fewer infant and severe maternal morbidities (Subbaraman et al., 2024). These specific components of alcohol taxes might also relate to better birth outcomes, though no study has examined this comprehensively.
In a meta-analysis of 112 alcohol tax and pricing studies, the authors concluded that they “know of no other preventive intervention to reduce drinking that has the numbers of studies and consistency of effects seen in the literature on alcohol taxes and prices” (Wagenaar et al., 2009). Yet, although numerous studies and review papers consistently report protective effects of alcohol taxation across multiple health outcomes, details regarding the nuances behind alcohol taxes are often left out. Alcohol taxes comprise multiple components (e.g., volume-based, sales, ad valorem), and can vary by beverage type and according to whether consumption is on-or off-premises. Although volume-based or “excise” taxes and sales taxes are most common (9), ad valorem or “value-based” taxes are, sometimes, implemented in the U.S. and elsewhere. One study comparing the beer volume-based tax to a comprehensive tax measure showed that the more comprehensive tax measure predicted binge drinking while the simpler volume-based tax measure did not (Xuan et al., 2015). Incorporating additional tax types (but not beverage types) into the tax measure also explained more variance in binge drinking outcomes (Xuan et al., 2015). However, combining multiple tax measures obscures which specific taxes might be most impactful.
Rationale for current study
Though the number of state-level policies regarding alcohol use during pregnancy has dramatically increased in recent decades, the prevalence of drinking during pregnancy in the U.S. has not changed. Alcohol taxes have been one of the most successful alcohol policies for reducing alcohol use in the general population and might be a better strategy for reducing harms related to drinking during pregnancy than policies focused on pregnant people. Supporting this, results from prior studies in the 1990s in the U.S. and the mid-2000s in Finland suggest that increased alcohol taxes are associated with better birth outcomes. However, given the nuances in alcohol taxes, prior studies are limited in their scope as few studies have examined how different beverage-specific tax types are related to birth outcomes. Attention to birth outcomes is important because these are population-level health indicators related to alcohol use during pregnancy that are routinely tracked across states and over time. Given the evidence that pregnancy-specific policies in the U.S. appear related to more harms than benefits, alternate policy strategies are needed. Understanding the nuances behind alcohol taxes is crucial to identifying which taxes could be most impactful. Thus, here we examine whether beverage-specific volume-based, sales, and ad valorem alcohol taxes are related to preterm and low birthweight births in the U.S. from 1993–2019.
METHODS
Tax data
Data for taxes came from NIAAA’s Alcohol Policy Information System (APIS), the Pacific Institute for Research and Evaluation, the Wine Institute (Blanchette et al., 2020), and original legal research using the online legal research tools Westlaw and HeinOnline. Time-varying state-level alcohol tax policies were beverage-specific volume-based, sales, and ad valorem taxes. Each tax was available for beer, wine, and spirits, for a total of nine tax variables.
In the U.S., volume-based taxes are typically levied per gallon at the wholesale level, though the specific point of collection may vary by state. This tax can be converted to dollars or cents per drink. Volume-based taxes were measured as cents/drink and adjusted to the 2020 Consumer Price Index (CPI). Sales taxes are taxes on goods in general. Not all states have sales taxes, and among those that do, not all apply their sales tax to alcohol. Some states instead have sales taxes specific to each beverage type. Given available data, sales taxes were operationalized here using a Yes/No dichotomous indicator of whether the state has a sales tax for each specific beverage. Finally, ad valorem taxes are levied as a percentage (%) of the beverage’s retail price. Different ad valorem tax rates may apply to on- and off-premises sales; given the low levels of variation over time and across states in off-premise ad valorem taxes, we focus here on on-premise ad valorem taxes. Ad valorem taxes were measured as a percentage of the retail price. Note that APIS does not track wine and spirits ad valorem and volume-based taxes for the 17–18 states with government monopoly control over wine and spirits whose prices are set by the state (the number varies because Washington state privatized its spirits system at the end of 2011). All other tax variables cover the 50 states and D.C.
Birth outcomes data
Birth certificate data for 101,191,080 singleton live births conceived between January 1993 - March 2019 come from the U.S. National Vital Statistics System. These data capture 100% of births in all 50 states and D.C. Outcomes were preterm birth (< 37 weeks) and low birthweight (< 2500 grams). Continuous birthweight and gestational age were obtained from birth certificates and dichotomized for analyses. Outcome data were merged with state/year-level taxes on state of mother’s residence and year of conception. Cases missing outcome data (0.67% for preterm birth, 0.08% for low birthweight) were dropped from analyses.
Covariates
We adjusted for both individual- and state-level covariates. Individual-level covariates were mother’s age, race, marital status, education, and parity, which came from birth certificates. State-level covariates were time-varying state-level cigarette consumption, poverty, unemployment, and per capita alcohol consumption. State-level cigarette consumption came from the U.S. Centers for Disease Control and Prevention (“The Tax Burden on Tobacco, 1970–2019 | Data | Centers for Disease Control and Prevention,” n.d.). State-level poverty and unemployment came from the U.S. Census Bureau (U.S. Census Bureau, 2020). State-level per capita alcohol consumption came from the Alcohol Epidemiologic Data System (National Institute on Alcohol Abuse and Alcoholism, n.d.).
Statistical analyses
We first examined changes in taxes over the study period descriptively and tested changes over time in regression models. We then used logistic regression to test associations between taxes and birth outcomes. Models first included each tax separately, then combined models included the volume-based (cents/drink), dichotomous (Y/N) sales, and ad valorem (%) tax for each specific beverage (beer, wine, spirits) in the same model. Correlations between beverage-specific volume-based taxes range from 0.53 to 0.68; correlations between beverage-specific ad valorem tax rates range from 0.57 to 0.78; and correlations between volume-based taxes and ad valorem tax rates range from −0.14 to −0.10. These correlations indicate that volume-based taxes should not be modeled with other volume-based taxes and that ad valorem taxes should not be modeled with other ad valorem taxes. This also supports the approach of examining volume-based, ad valorem, and sales taxes separately for each specific beverage (Subbaraman et al., 2024).
For each set of models, Model 1 included state-specific linear, quadratic, and cubic state-specific trends; Model 2 further adjusted for mother’s age, race, marital status, education, and parity, and state-level cigarette consumption, poverty, and unemployment; and Model 3 further adjusted for fixed effects for state and year. All models clustered standard errors by state; specifically, we used vce(cluster state) in Stata to specify that observations within state are not independent. Sensitivity analyses further adjusted for state-level per capita alcohol consumption, which reflects both the state’s alcohol environment as well as the strength of other state alcohol policies. We did not include per capita consumption in main models because it may serve as a mediator, thus attenuating the tax-birth outcome association. Given volume-based tax data (in $/gallon) are available for 1972–2019, a second set of sensitivity analyses examined associations between volume-based taxes and outcomes using the longer period (N= 160,027,713).
RESULTS
Tax descriptives
Figure 1 illustrates average U.S. state-level volume-based taxes by beverage for 1993–2019. From 1993–2019, all volume-based taxes decreased significantly (p<.05) on average; specifically, beer volume-based taxes dropped from 4.2 cents/drink to 2.9 cents/drink, wine volume-based taxes dropped from 4.6 cents/drink to 3.4 cents/drink, and spirits volume-based taxes dropped from 7.5 cents/drink to 5.1 cents/drink (Table 1). When examining the longer 1972–2019 period (Figure 2), the recent decrease appears less dramatic than the overall decline since the 1970s, with larger drops in the 1980s and relatively smaller drops in the 1990s and 2000s. For example, beer volume-based taxes decreased by more than 50% from 1972 ($1.07/gallon) to 1982 ($0.50), with an additional decrease to just $0.31/gallon by 2019. Both wine and spirits volume-based taxes eroded similarly from 1972–2019, with similar proportionate decreases.
Figure 1.
Average U.S. state-level volume-based taxes by beverage (cents/drink), 1993–2019
Table 1.
Descriptives of U.S. state-level volume-based, sales, and ad valorem taxesa
| 1993 | 2019 | |||
|---|---|---|---|---|
| Mean (SD) | Range | Mean (SD) | Range | |
| Beer volume-based tax (cents/drink) | 4.2 (3.8) | 0.3–17.5 | 2.9 (2.8) | 0.2–12.0 |
| Wine volume-based tax (cents/drink) | 4.6 (3.1) | 0.8–15.6 | 3.4 (2.3) | 0.8–9.8 |
| Spirits volume-based tax (cents/drink) | 7.5 (2.9) | 3.1–13.5 | 5.1 (2.7) | 1.8–15 |
| Beer ad valorem on-p retail (%) | 1.2 (3.4) | 0–14 | 1.4 (3.9) | 0–15 |
| Wine ad valorem on-p retail (%) | 1.7 (4.2) | 0–15 | 1.9 (4.5) | 0–15 |
| Spirits ad valorem on-p retail (%) | 2.5 (5.2) | 0–16 | 2.3 (4.9) | 0–15 |
| 1993 | 2019 | |
|---|---|---|
| # states with off-premises beer sales tax | 39 | 41 |
| # states with off-premises wine sales tax | 25 | 26 |
| # states with off-premises spirits sales tax | 25 | 25 |
| 1993 | 2019 | |||
|---|---|---|---|---|
| Mean (SD) | Range | Mean (SD) | Range | |
| Beer ad valorem on-p retail (%) | 9.9 (3.0) | 7–14 (n=4) | 9.6 (5.0) | 2.5–15 (n=5) |
| Wine ad valorem on-p retail (%) | 10.9 (3.5) | 7–15 (n=5) | 10.5 (5.0) | 2.5–15 (n=6) |
| Spirits ad valorem on-p retail (%) | 12.0 (3.5) | 7–16 (n=7) | 11.0 (4.8) | 2.5–15 (n=7) |
| 1972 | 1982 | 2019 | ||||
|---|---|---|---|---|---|---|
| Mean (SD) | Range | Mean (SD) | Range | Mean (SD) | Range | |
| Beer volume-based tax ($/gallon) | 1.07 (0.97) | 0.12–4.70 | 0.50 (0.42) | 0.05–2.03 | 0.31 (0.29) | 0.02–1.28 |
| Wine volume-based tax ($/gallon) | 2.89 (2.13) | 0.06–9.17 | 1.35 (1.05) | 0.03–4.64 | 0.88 (0.59) | 0.20–2.50 |
| Spirits volume-based tax ($/gallon) | 15.77 (5.18) | 9.17–27.71 | 7.47 (2.31) | 3.97–12.58 | 4.28 (2.27) | 1.50–12.80 |
In regression models, volume based taxes all decreased significantly during study period; the number of states with beer and wine sales taxes changed signficantly; and ad valorem taxes did not change significantly.
Figure 2.
Average U.S. state-level volume-based taxes by beverage ($/gallon), 1972–2019
Sales taxes also changed over the study period: the number of states with off-premises beer sales taxes increased from 39 to 41, the number of states with off-premises wine sales taxes increased from 25 to 26, and the number of states with off-premises spirits sales taxes did not change. As shown in Table 1, changes in sales taxes occurred in both directions in that some states added sales taxes while some states removed sales taxes for each beverage. Ad valorem taxes remained relatively stable over the study period compared to volume-based and sales taxes (Table 1). Specifically, beer ad valorem taxes increased slightly on average from 1.2% to 1.4% (but decreased slightly from 9.9% to 9.6% in states with non-zero rates); wine increased slightly on average from 1.7% to 1.9% (but decreased slightly from 10.9% to 10.5% in states with non-zero rates); and spirits decreased slightly on average from 2.5% to 2.3% (and decreased from 12.0% to 11.0% in states with non-zero rates).
Associations between taxes and birth outcomes
Table 2 shows results from separate tax models. No tax was significantly related to preterm birth in either Models 1 or 2. In fully adjusted Model 3, having a beer sales tax vs. not (OR = 0.95, 95% CI: 0.92, 0.99) or having a spirits sales tax vs. not (OR = 0.95, 95% CI: 0.92, 0.99) were each significantly related to lower odds of preterm birth. This pattern of findings was generally similar for low birthweight. However, in Model 1, the spirits volume-based tax (OR = 0.99, 95% CI: 0.99, 0.999) and wine ad valorem tax (OR = 1.003, 95% CI: 1.0002, 1.005) were each significantly related to low birthweight. Like for preterm birth, there were no significant associations in Model 2 for low birthweight. Again, like for preterm birth, in fully adjusted Model 3, having a beer sales tax (OR = 0.98, 95% CI: 0.97, 0.999) or having a spirits sales tax (OR = 0.98, 95% CI: 0.96, 0.999) were each significantly related to lower odds of low birthweight.
Table 2.
Odds ratios and 95% confidence intervals from separate regressions of U.S. state-level tax associations with birth outcomes 1993–2019 Vital Statistics (N=101,191,080 live singleton births)a
| Preterm Birth | Low Birthweight | |||||
|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 1 | Model 2 | Model 3 | |
| Tax | Trends | Trends + covariates | Fully adjusted | Trends | Trends + covariates | Fully adjusted |
| Beer volume-based tax (cents/drink) | 1.00 (0.99, 1.00) |
1.00 (0.99, 1.00) |
1.00 (1.00, 1.00) |
1.00 (0.99, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
| Wine volume-based tax (cents/drink) | 1.00 (0.98, 1.01) |
1.00 (0.98, 1.02) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
| Spirits volume-based tax (cents/drink) | 0.99 (0.99, 1.00) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
0.99
(0.99, 0.999)* |
1.00 (0.99, 1.00) |
1.00 (0.99, 1.00) |
| Beer sales tax vs. not | 0.97 (0.91, 1.04) |
0.96 (0.91, 1.02) |
0.95
(0.92, 0.99)** |
1.00 (0.96, 1.04) |
0.99 (0.98, 1.01) |
0.98
(0.97, 0.999)* |
| Wine sales tax vs. not | 0.97 (0.92, 1.02) |
0.96 (0.92, 1.00) |
0.97 (0.93, 1.01) |
1.00 (0.97, 1.03) |
0.99 (0.97, 1.01) |
1.00 (0.97, 1.03) |
| Spirits sales tax vs. not | 0.97 (0.91, 1.04) |
0.96 (0.91, 1.01) |
0.95
(0.92, 0.99)* |
1.00 (0.96, 1.04) |
0.99 (0.96, 1.01) |
0.98
(0.96, 0.999)* |
| Beer ad valorem (%) | 1.00 (1.00, 1.01) |
1.00 (1.00, 1.00) |
1.00 (0.99, 1.01) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
| Wine ad valorem (%) | 1.00 (1.00, 1.01) |
1.00 (1.00, 1.01) |
1.00 (0.99, 1.01) |
1.003
(1.0002, 1.005)* |
1.00 (1.00, 1.00) |
1.00 (0.99, 1.00) |
| Spirits ad valorem (%) | 1.00 (1.00, 1.01) |
1.00 (0.99, 1.01) |
1.01 (1.00, 1.03) |
1.00 (0.99, 1.00) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
Analyses for wine and spirits volume-based and ad valorem taxes are missing data for government monopoly states (N= 75,385,875 for analyses missing data on government monopoly states). Volume-based taxes were measured as cents/drink, sales taxes were measured using a Yes/No dichotomous indicator of whether the state has a specific sales tax for that beverage, and ad valorem taxes were measured as a percentage of the retail price. Volume-based taxes were adjusted to the 2019 Consumer Price Index.
Each tax is included in each model separately.
Model 1 adjusts for state-specific linear, quadratic, and cubic trends; Model 2 further adjusts for mother’s age, race, marital status, education, and parity, and state-level cigarette consumption, poverty, and unemployment; Model 3 further adjusts for fixed effects for state and year.
P<0.05,
P<0.01,
P<0.001
Bold indicates P<0.05
These patterns generally persisted in combined tax models (Table 3). There were again no significant associations between taxes and preterm birth in Models 1 or 2. This was also true for low birthweight. But again, these relationships changed in fully adjusted models (Model 3). Specifically, having a beer sales tax was still significantly related to lower odds of preterm birth (OR = 0.94, 95% CI: 0.90, 0.98) and lower odds of low birthweight (OR = 0.98, 95% CI: 0.96, 0.99) even after adjusting for beer volume-based and beer ad valorem taxes. Similarly, having a spirits sales tax was still significantly related to lower odds of preterm birth (OR = 0.95, 95% CI: 0.91, 0.98) and lower odds of low birthweight (OR = 0.97, 95% CI: 0.95, 0.99) even after adjusting for spirits volume-based and spirits ad valorem taxes. Though not significant in separate tax models, having a wine sales tax was also significantly related to lower odds of preterm birth (OR = 0.94, 95% CI: 0.90, 0.98) and lower odds of low birthweight (OR = 0.97, 95% CI: 0.95, 0.99) in fully adjusted combined tax models (Table 3, Model 3). Results were robust to inclusion of state-level per capita alcohol consumption. Results from sensitivity analyses of volume-based taxes for the years 1972–2019 (not shown) consistently yielded no significant associations between any volume-based tax and outcomes.
Table 3.
Odds ratios and 95% confidence intervals from beverage-specific 1993–2019 Vital Statistics (N=101,191,080 live singleton births)a
| Preterm birth | Low birthweight | |||||
|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 1 | Model 2 | Model 3 | |
| Tax | Adjusted for trends | Trends + covariates | Fully adjusted | Adjusted for trends | Trends + covariates | Fully adjusted |
| Beer | ||||||
| Volume-based (cents/drink) | 1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
| Sales tax (Y vs. N) | 0.97 (0.90, 1.05) |
0.96 (0.91, 1.02) |
0.94
(0.90, 0.98)** |
1.00 (0.96, 1.05) |
0.99 (0.97, 1.02) |
0.98
(0.96, 0.99)** |
| Ad valorem (%) | 1.00 (1.00, 1.01) |
1.00 (1.00, 1.00) |
1.00 (0.99, 1.00) |
1.00 (1.00, 1.01) |
1.00 (1.00, 1.00) |
1.00 (1.00, 1.00) |
| Wine | ||||||
| Volume-based (cents/drink) | 1.00 (0.98, 1.01) |
1.00 (0.98, 1.02) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
| Sales tax (Y vs. N) | 0.97 (0.90, 1.05) |
0.96 (0.91, 1.01) |
0.94
(0.90, 0.98)** |
1.00 (0.96, 1.04) |
0.98 (0.95, 1.01) |
0.97
(0.95, 0.99)* |
| Ad valorem (%) | 1.00 (1.00, 1.01) |
1.00 (1.00, 1.00) |
1.00 (0.99, 1.00) |
1.00 (1.00, 1.01) |
1.00 (0.99, 1.00) |
1.00 (1.00, 1.00) |
| Spirits | ||||||
| Volume-based (cents/drink) | 0.99 (0.99, 1.00) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
0.99 (0.99, 1.00) |
0.99 (0.99, 1.00) |
1.00 (0.99, 1.00) |
| Sales tax (Y vs. N) | 0.96 (0.90, 1.04) |
0.95 (0.90, 1.01) |
0.95
(0.91, 0.98)** |
0.99 (0.95, 1.04) |
0.98 (0.95, 1.01) |
0.97
(0.95, 0.99)** |
| Ad valorem (%) | 1.00 (0.98, 1.01) |
1.00 (0.99, 1.01) |
1.01 (0.99, 1.02) |
1.00 (0.99, 1.00) |
1.00 (0.99, 1.01) |
1.00 (0.99, 1.01) |
Analyses for wine and spirits volume-based and ad valorem taxes are missing data for government monopoly states (N= 75,385,875 for analyses missing data on government monopoly states). Volume-based taxes were measured as cents/drink, sales taxes were measured using a Yes/No dichotomous indicator of whether the state has a specific sales tax for that beverage, and ad valorem taxes were measured as a percentage of the retail price. Volume-based taxes were adjusted to the 2019 Consumer Price Index.
Volume-based, sales tax, and ad valorem taxes are included in models together for each specific beverages.
Model 1 adjusts for state-specific linear, quadratic, and cubic trends; Model 2 further adjusts for mother’s age, race, marital status, education, and parity, and state-level cigarette consumption, poverty, and unemployment; Model 3 further adjusts for fixed effects for state and year.
P<0.05,
P<0.01,
P<0.001
Bold indicates P<0.05
DISCUSSION
We examined how state-level beverage-specific sales taxes are related to adverse birth outcomes using data from more than 100 million U.S. births over a 36-year period. Results show that sales taxes on alcoholic beverages – whether beer, wine, or spirits – are associated with lower odds of preterm and lower odds of low birthweight births. These findings were consistent in both separate and combined tax models as well as in sensitivity analyses. They also corroborate previous analyses showing that wine and spirits sales taxes are particularly protective against severe maternal morbidities (Subbaraman et al., 2024) and add to the evidence base for health benefits of alcohol taxes for infant outcomes more broadly. Findings also support that general population alcohol policies might be an effective strategy for reducing harms related to drinking during pregnancy. We note that while the magnitudes of effects are generally small, they are still meaningful in such a large population.
Implications for alcohol policy
Prior studies on alcohol taxes often combine tax measures, obscuring potential effects of individual tax measures. While others have found that combined tax measures relate to drinking outcomes more consistently than simple volume-based measures (Xuan et al., 2015), examining specific tax variables like we did allows identification of which taxes are most relevant for our outcomes of interest. Here we found significant relationships only between sales taxes and birth outcomes, and no significant associations between either volume-based or ad valorem taxes and outcomes. These findings suggest that policymakers interested in reducing adverse birth outcomes might prioritize implementing sales taxes on alcoholic beverages. Though we found no significant associations between volume-based taxes and birth outcomes here, previously published analyses do show associations between increased wine and spirits volume-based taxes and lower odds of infant morbidities; as these infant morbidities include other consequences of maternal alcohol consumption during pregnancy, including birth defects, increased volume-based taxes might be another policy lever for reducing harms related to drinking during pregnancy.
Notably, volume-based taxes are at an all-time low (Blanchette et al., 2020) with little variation over the study period; thus, the lack of significant associations between volume-based taxes and birth outcomes found here could be due to low values of volume-based taxes overall. Historically, higher volume-based taxes have been linked to lower rates of alcohol consumption and related harms in the general population (Kubik and Moran, 2002; Manning et al., 1995; Wagenaar et al., 2009, 2010). Furthermore, prior analyses show significant protective associations between wine and spirits volume-based taxes and infant morbidities related to maternal alcohol consumption as well as severe maternal morbidities (Subbaraman et al., 2024). As others have noted (Blanchette et al., 2020; Chaloupka et al., 2019), indexing volume-based taxes to inflation could yield more public health benefits, such as reducing adverse outcomes related to drinking during pregnancy.
The current null findings for ad valorem taxes also corroborate previous analyses showing no significant associations between ad valorem taxes and other harms related to drinking during pregnancy, specifically drinking among women of reproductive age and infant and maternal morbidities (Subbaraman et al., 2024). However, previous analyses also show that increased spirits ad valorem taxes are associated with lower rates of alcohol-related treatment admissions among pregnant people, indicating that ad valorem taxes might still influence drinking during pregnancy (Subbaraman et al., 2024). Generally, higher alcohol prices lead to less drinking (Chaloupka et al., 2002; Wagenaar et al., 2009). Thus, any additional taxes that increase alcohol prices should decrease drinking in the general population, which includes pregnant people. Notably, few states implement ad valorem taxes – e.g., only four had an ad valorem beer tax in 1993, which rose to just five in 2019, and only seven had spirits ad valorem taxes throughout the study period. The average values of ad valorem taxes also did not change substantially over time. The low prevalence of ad valorem taxes combined with the lack of variation over time likely affects power to detect associations; future studies might consider focusing on states with ad valorem taxes to examine whether changes in these taxes impact alcohol-related health outcomes.
Strengths and limitations
This is the first study to examine how multiple components of state-level alcohol taxes are related to birth outcomes. Strengths of this study include the examination of multiple tax components over a long period, the use of outcome data that do not rely on self-report, the consistency of results between separate and combined tax models, and the inclusion of all singleton live births across the 50 states and D.C. for the study period. The use of data that are not self-report avoids problems of recall and social desirability biases, while the inclusion of all singleton live births confers an uncommonly high level of generalizability. Results were also robust across various model specifications, further strengthening our conclusions. However, results still come with caveats. First, tax data are not available for government monopoly states whose prices are set by the state. Thus, results from models that exclude government monopoly states might not generalize to those states. Ongoing data collection efforts aim to address this gap. Second, birth certificate data are not collected for research purposes and lack valid data on individual-level cigarette and alcohol use, both of which are related to outcomes. Our adjustment for state-level cigarette and alcohol consumption is a proxy for individual-level measures. Third, in sensitivity analyses using volume-based data for 1972–2019, birth certificate data are limited to a 50% sample for 1972–1984 and measure race and ethnicity inconsistently until 1989; these issues might limit power and inference. Fourth, only five states added or removed wine and/or spirits sales during the study period. Thus, estimates rely more on cross-state variation than on variation over time and should be interpreted as associations rather than causal effects. Fifth, we did not adjust for other state-level alcohol policies; however, prior analyses of state-level alcohol taxes show that taxes are not correlated with policies affecting alcohol availability (Subbaraman et al., 2020), suggesting that confounding by other related policies is likely not an issue. Furthermore, our inclusion of state-level per capita consumption in sensitivity analyses alleviates concerns related to potential confounding by other state-level policies. Sixth, regarding alternate methods, Autoregressive and Integrated Moving Average (ARIMA) models might increase accuracy and reduce bias of results given potential autocorrelation in birth outcomes. However, ARIMA models generally require a longer time series (e.g., at least 50 timepoints), which were not available here. Finally, these results could be strengthened by inclusion of other consequences of prenatal alcohol exposure. While consistent data on fetal alcohol spectrum disorders are not available across states and over time in the U.S., future studies in other geographic contexts that have consistently measured fetal alcohol spectrum disorder data should examine these outcomes.
Conclusion
In analyses of more than 100 million births from 1993–2019, we find that state-level beer, wine, and spirits taxes are related to lower odds of preterm and low birthweight births. This study provides additional evidence that general population alcohol policies, including increased alcohol taxes, may be relevant for reducing adverse effects related to pregnant people’s alcohol consumption.
PUBLIC HEALTH SIGNFICANCE.
Policies related to alcohol use during pregnancy in the U.S. are related to more harms than benefits. Broader alcohol policies that affect the general public – and not just pregnant people – might be better for reducing harms related to drinking during pregnancy. Increased alcohol taxes are linked to better birth outcomes and are one example of broader alcohol policies that may reduce harms from alcohol use during pregnancy.
ACKNOWLEDGEMENTS
This work was supported by NIAAA R01 AA023267 from the US National Institute on Alcohol Abuse and Alcoholism. The authors gratefully acknowledge Jason Blanchette for sharing volume-based tax data.
Footnotes
DECLARATION OF INTEREST
None to declare.
DATA AVAILABILITY
Data from U.S. Vital Statistics are publicly available from the National Center For Health Statistics (https://www.cdc.gov/nchs/data_access/vitalstatsonline.htm). Data on ad valorem and sales taxes will be shared on reasonable request to the corresponding author. The authors do not have permission to share data on volume-based taxes.
REFERENCES
- Ayyagari P, Deb P, Fletcher J, Gallo W, & Sindelar J (2009). Sin Taxes: Do Heterogeneous Responses Undercut Their Value? (No. w15124) (p. w15124). Cambridge, MA: National Bureau of Economic Research. doi: 10.3386/w15124 [DOI] [Google Scholar]
- Babor TF (2010). Public health science and the global strategy on alcohol. Bulletin of the World Health Organization, 88(9), 643. doi: 10.2471/BLT.10.081729. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Babor TF, Casswell S, Graham K, Huckle T, Livingston M, Österberg E, et al. (2023). Alcohol: No Ordinary Commodity: Research and public policy (Third Edition, New to this Edition: Third Edition, New to this Edition:). Oxford, New York: Oxford University Press. [Google Scholar]
- Blanchette JG, Ross CS, and Naimi TS (2020). The Rise and Fall of Alcohol Excise Taxes in U.S. States, 1933–2018. Journal of Studies on Alcohol and Drugs, 81(3), 331–338. Alcohol Research Documentation, Inc. doi: 10.15288/jsad.2020.81.331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention (CDC). (1997, April 25). Alcohol consumption among pregnant and childbearing-aged women—United States, 1991 and 1995. MMWR. Morbidity and Mortality Weekly Report, 46(16), 346–350. [PubMed] [Google Scholar]
- Chaloupka FJ, Grossman M, & Saffer H (2002). The effects of price on alcohol consumption and alcohol-related problems. Alcohol Research & Health: The Journal of the National Institute on Alcohol Abuse and Alcoholism, 26(1), 22–34. [PMC free article] [PubMed] [Google Scholar]
- Chaloupka FJ, Powell LM, & Warner KE (2019). The Use of Excise Taxes to Reduce Tobacco, Alcohol, and Sugary Beverage Consumption. Annual Review of Public Health, 40(1), 187–201. doi: 10.1146/annurev-publhealth-040218-043816. [DOI] [PubMed] [Google Scholar]
- Dejong K, Olyaei A, & Lo JO (2019). Alcohol Use in Pregnancy. Clinical Obstetrics and Gynecology, 62(1), 142–155. doi: 10.1097/GRF.0000000000000414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gosdin LK, Deputy NP, Kim SY, Dang EP, & Denny CH (2022). Alcohol Consumption and Binge Drinking During Pregnancy Among Adults Aged 18–49 Years - United States, 2018–2020. MMWR. Morbidity and Mortality Weekly Report, 71(1), 10–13. doi: 10.15585/mmwr.mm7101a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kubik JD, & Moran JR (2002). Can Policy Changes Be Treated as Natural Experiments? Evidence from State Excise Taxes. SSRN Electronic Journal. doi: 10.2139/ssrn.1808888 [DOI] [Google Scholar]
- Luukkonen J, Junna L, Remes H, & Martikainen P (2023). The association of lowered alcohol prices with birth outcomes and abortions: A population-based natural experiment. Addiction (Abingdon, England), 118(5), 836–844. doi: 10.1111/add.16119. [DOI] [PubMed] [Google Scholar]
- Manning WG, Blumberg L, & Moulton LH (1995). The demand for alcohol: the differential response to price. Journal of Health Economics, 14(2), 123–148. doi: 10.1016/0167-6296(94)00042-3. [DOI] [PubMed] [Google Scholar]
- May PA, Hasken JM, Hooper SR, Hedrick DM, Jackson-Newsom J, Mullis CE, et al. (2021). Estimating the community prevalence, child traits, and maternal risk factors of fetal alcohol spectrum disorders (FASD) from a random sample of school children. Drug and Alcohol Dependence, 227, 108918. doi: 10.1016/j.drugalcdep.2021.108918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meier PS, Purshouse R, & Brennan A (2010). Policy options for alcohol price regulation: the importance of modelling population heterogeneity. Addiction (Abingdon, England), 105(3), 383–393. doi: 10.1111/j.1360-0443.2009.02721.x. [DOI] [PubMed] [Google Scholar]
- National Institute on Alcohol Abuse and Alcoholism. (n.d.). Alcohol Epidemiologic Data System (AEDS). Retrieved January 5, 2026, from https://www.niaaa.nih.gov/alcohol-epidemiologic-data-system-aeds-re-competition
- Nelson TF, Xuan Z, Babor TF, Brewer RD, Chaloupka FJ, Gruenewald PJ, et al. (2013). Efficacy and the strength of evidence of U.S. alcohol control policies. American Journal of Preventive Medicine, 45(1), 19–28. doi: 10.1016/j.amepre.2013.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Roberts SCM (2023). Commentary on Luukkonen et al. : Policy possibilities for reducing public health harms related to pregnant people’s alcohol consumption. Addiction (Abingdon, England), 118(5), 845–846. doi: 10.1111/add.16164. [DOI] [PubMed] [Google Scholar]
- Roberts SCM, Mericle AA, Subbaraman MS, Thomas S, Treffers RD, Delucchi KL, & Kerr WC (2019). State Policies Targeting Alcohol Use during Pregnancy and Alcohol Use among Pregnant Women 1985–2016: Evidence from the Behavioral Risk Factor Surveillance System. Women’s Health Issues: Official Publication of the Jacobs Institute of Women’s Health, 29(3), 213–221. doi: 10.1016/j.whi.2019.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Roberts SCM, Schulte A, Zaugg C, Leslie DL, Corr TE, & Liu G (2023). Association of Pregnancy-Specific Alcohol Policies With Infant Morbidities and Maltreatment. JAMA Network Open, 6(8), e2327138. doi: 10.1001/jamanetworkopen.2023.27138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Roberts SCM, Thomas S, Treffers R, & Drabble L (2017). Forty Years of State Alcohol and Pregnancy Policies in the USA: Best Practices for Public Health or Efforts to Restrict Women’s Reproductive Rights? Alcohol and Alcoholism (Oxford, Oxfordshire), 52(6), 715–721. doi: 10.1093/alcalc/agx047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schulte A, Liu G, Subbaraman MS, Kerr WC, Leslie D, & Roberts SCM (2024). Relationships Between Alcohol Policies and Infant Morbidities and Injuries. American Journal of Preventive Medicine, 66(6), 980–988. doi: 10.1016/j.amepre.2024.01.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Subbaraman MS, Mulia N, Kerr WC, Patterson D, Karriker-Jaffe KJ, & Greenfield TK (2020). Relationships between US state alcohol policies and alcohol outcomes: differences by gender and race/ethnicity. Addiction (Abingdon, England), 115(7), 1285–1294. doi: 10.1111/add.14937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Subbaraman MS, Schulte A, Berglas NF, Kerr WC, Thomas S, Treffers R, et al. (2024). Associations between alcohol taxes and varied health outcomes among women of reproductive age and infants. Alcohol and Alcoholism (Oxford, Oxfordshire), 59(3), agae015. doi: 10.1093/alcalc/agae015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Subbaraman MS, Sesline K, Kerr WC, & Roberts SCM (2023). Associations between state-level general population alcohol policies and drinking outcomes among women of reproductive age: Results from 1984 to 2020 National Alcohol Surveys. Alcoholism, Clinical and Experimental Research, 47(9), 1773–1782. doi: 10.1111/acer.15156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Subbaraman MS, Thomas S, Treffers R, Delucchi K, Kerr WC, Martinez P, & Roberts SCM (2018). Associations Between State-Level Policies Regarding Alcohol Use Among Pregnant Women, Adverse Birth Outcomes, and Prenatal Care Utilization: Results from 1972 to 2013 Vital Statistics. Alcoholism, Clinical and Experimental Research, 42(8), 1511–1517. doi: 10.1111/acer.13804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- The Tax Burden on Tobacco. 1970–2019 | Data | Centers for Disease Control and Prevention. (n.d.). Retrieved January 7, 2026, from https://data.cdc.gov/Policy/The-Tax-Burden-on-Tobacco-1970-2019/7nwe-3aj9/about_data
- U.S. Census Bureau. (2020). American Community Survey 2015–2019 5-Year Data Release. Retrieved August 17, 2023, from https://www.census.gov/newsroom/press-kits/2020/acs-5-year.html
- Wagenaar AC, Salois MJ, and Komro KA (2009). Effects of beverage alcohol price and tax levels on drinking: a meta-analysis of 1003 estimates from 112 studies. Addiction (Abingdon, England), 104(2), 179–190. doi: 10.1111/j.1360-0443.2008.02438.x [DOI] [PubMed] [Google Scholar]
- Wagenaar AC, Tobler AL, & Komro KA (2010). Effects of alcohol tax and price policies on morbidity and mortality: a systematic review. American Journal of Public Health, 100(11), 2270–2278. doi: 10.2105/AJPH.2009.186007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xuan Z, Chaloupka FJ, Blanchette JG, Nguyen TH, Heeren TC, Nelson TF, & Naimi TS (2015). The relationship between alcohol taxes and binge drinking: evaluating new tax measures incorporating multiple tax and beverage types. Addiction (Abingdon, England), 110(3), 441–450. doi: 10.1111/add.12818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang N (2010). Alcohol taxes and birth outcomes. International Journal of Environmental Research and Public Health, 7(5), 1901–1912. doi: 10.3390/ijerph7051901. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data from U.S. Vital Statistics are publicly available from the National Center For Health Statistics (https://www.cdc.gov/nchs/data_access/vitalstatsonline.htm). Data on ad valorem and sales taxes will be shared on reasonable request to the corresponding author. The authors do not have permission to share data on volume-based taxes.