Maple syrup production in the United States: growth decomposition, variability and policy implications

Maple syrup is a natural sweetener that has been produced predominantly in Eastern Canada and the Northeastern United States for centuries (Massachusetts Maple Producers Association (MMPA), 2025; Wang et al., 2025a). It is produced through boiling and concentrating the sap harvested from select varieties of maple trees that grow in this region. As a result of its long history and unique characteristics, maple syrup is considered as an iconic symbol of this region's agriculture, natural landscape, culture and community heritage (Bal, 2023; Stiefel, 2020; Wang et al., 2025a, b). From its beginning as a subsistence sweetener, maple syrup has grown into as a globally recognized natural food product with significant economic, environmental and cultural contributions (Murphy et al., 2012; Wang et al., 2025a, b). For example, in Canada, the maple syrup industry provides about 12,600 full-time jobs, contributes 1.11 billion Canadian dollars (C$) to the Canadian gross domestic product (GDP) and generates $C235 million in tax revenue annually in recent years (Quebec Maple Syrup Producers (QMSP), 2026). Although the annual production value of maple syrup in the United States is between only $160 and $210 million in recent years, the maple syrup industry provides significant income and employment for many rural communities through direct, indirect, and induced contributions (Becot et al., 2015; Cannella et al., 2022; Magry et al., 2024; Sheppard et al., 2020). For example, Becot et al. (2015) estimated the total economic contributions of the maple syrup sector in Vermont, the largest maple syrup producer in the United States, to be $330 million in 2013, with the direct, indirect and inducted contributions at $222 million, $61 million and $47 million, respectively.

As shown in Figure 1, total US maple syrup production fluctuated around 1.3 million gallons from 1992 to 2000 but has increased steadily since 2001, reached a record high of 5.99 million gallons in 2022 and then dropped slightly to 5.77 million gallon in 2025. Figure 2 shows the total annual number of taps and the yield per tap in the United States from 2001 to 2025. Their growth trends indicate that the increase in the number of taps and yield per tap were both responsible for the increase in total maple syrup production presented in Figure 1. This figure also suggests that the growth in the number of taps was slightly faster than that in the yield per tap and the volatility of the yield per tap was significantly larger than that of the number of taps over the period.

Because agricultural production growth can be fueled by multiple factors such as increasing land and other inputs, improving the yield of inputs through technical progress or better resource management and implementing institutional or policy changes, decomposing the growth in production into the contributions of these factors is a crucial analytical tool for understanding the drivers of the production growth, guiding policy development and evaluation, and managing resources more effectively (Hulten et al., 2019; OECD, 2001). For example, decomposition analysis allows analysts and policymakers to assess the contributions of input increases and yield improvement as well the returns from public investment in research and development over a specific period (Cao et al., 2025; Huber et al., 2014). It can also help policymakers predict and evaluate the potential impacts of alternative public investment and policy options on production growth. For example, Blomqvist et al. (2020) reported that rising global yields contributed 89% to the growth in global crop production from 1961 to 2015, while cropland expansion only contributed 11% to the growth. The finding demonstrated that yield improvement through technical progress and innovation, not cropland expansion, was the primary driver of meeting the increasing global food demand.

Although several recent studies have discussed the potential contributions of technical progress and other factors to the growth in US maple syrup production (Farrell and Chabot, 2012; Thapa et al., 2024; Wang et al., 2025b), there are limited quantitative studies on the contributions of yield, number of taps and other factors to the growth in maple syrup production and the variability of production, yield and the number of taps using robust statistical methods. For example, while Abah et al. (2026) estimated the variabilities of US maple syrup production, yield and number of taps using the coefficients of variation (CVs), their results likely overestimated the actual variabilities due to that the CV of a variable significantly overestimates the variable's variation when the variable shows a positive or negative trend over time (Cuddy and Della Valle, 1978).

The major purpose of this study is to decompose the growth in US maple syrup production from 2001 to 2025 into contributions of yield improvement, increase in the number of taps and their interaction through regression analysis; to assess the variability in production, yield and the number of taps using the Cuddy Della Valle Index (CDVI); and to derive economic and policy implications for the maple syrup industry to achieve more sustainable growth.

Data used in this study include US national data from 2001 to 2025 with 25 observations and pooled time-series (2001–2025) and cross-sectional (seven major production states) data with a total of 175 observations. While the national dataset includes the yearly maple syrup production (Q in 1,000 gallons), yield per tap (Y in gallon per tap) and the number of taps (N in 1,000) from 2001 to 2025, the pooled dataset includes data for the same variables over the same period by seven states (Vermont, Maine, New York, New Hampshire, Pennsylvania, Michigan and Wisconsin). The summary statistics of the two datasets are reported in Table 1.

Although USDA-NASS started to publish annual maple syrup production data in five states in 1984 and the national production data in 1992, data on the number of taps and the yield per tap is only available starting in 2001. In its annual reports, USDA-NASS has frequently updated the data in subsequent years, as each annual report usually includes data for the current year and two or three previous years. The data used in this study was compiled from both the USDA-NASS annual reports and its Quick Stats database online. For the values with inconsistency across reports or sources, great efforts were made to crosscheck the data and select the most reliable values.

To decompose the growth in maple syrup production (Q) into contributions of the yield improvement (Y) and changes in the number of taps (N) using the time-series data through regression analysis, a log–log regression model is selected due to the multiplicative relationship between the three variables (Hulten et al., 2019; OECD, 2001). This method is suitable for maple syrup because production (Q) is calculated as the product of Y and N:

Taking the natural logarithm of both sides to transform the multiplicative relationship into a linear model relationship and then differentiating the entire equation with respect to time (t) yield the following equation:

When the chain rule of differentiation is applied to equation (2), the equation is transformed to the following relationship:

Note that, when the change in time (dt) is very small or when t is a continuous variable, the term on the left-hand side of equation (3) is the instantaneous or continuous average growth rate of production and the two terms on the right-hand side are the continuous average growth rates of Y and N, respectively. This equation yields a useful relationship: the continuous average growth rate of production is equal to the sum of the continuous average growth rates of yield and the number of taps.

Empirically, the continuous average growth rates of Y, N and Q can be estimated through the following regression models:

where Y_t, N_t and $Qt$ are, respectively, the yield, number of taps, and production at time point t, $α0,a,r0,r,$ $b0,and b$ are the regression coefficients to be estimated, and $ϵt,ηt and mt$ are the error terms. Note that a, r and b are the estimated continuous average growth rates of Y, N and Q, respectively. The conclusion about the three continuous average growth rates from Equation (3) can be presented by $a + r=b.$

When the annual data over multiple years is used to estimate the three models, a, $r and b$ are the estimated continuous average annual growth rates (CAAGRs) and they are mathematically different from the discrete average annual growth rates (DAAGRs). The DAAGRs are more meaningful for interpretation in agricultural analysis and can be calculated directly from the corresponding CAAGRs (i.e. a, r and b) according to Aimran (2014) and Andersen (2019):

where GR_Y, GR_N and GR_Q denote the DAAGRs of Y, N and Q, respectively. It is straightforward to show that the CAAGR and DAAGR for each variable are approaching the same value when the change in t approaches zero or when a, $r and b$ are very small. In this study of the production growth of maple syrup using annual data, the DAAGRs are clearly more meaningful and applicable because each data period is one year. In the next section, both the estimated CAAGRs (a, $r and b)$ and DAAGRs (GR_Y, GR_N and GR_Q) for Y, N and Q will be reported and compared.

Mathematically, the following relation can be derived for the three DAAGRs (Parmar and Devi, 2021):

While this equation indicates that the DAAGR of Q is equal to the sum of the corresponding DAAGRs of Y and N and their interaction term, the interaction term is close to zero when both GR_Y and GR_N are very small. Equation (8) also suggests that the percentage contributions of yield improvement, change in the number of taps and their interaction to the growth in production over a specific period can be calculated from the percentage shares of GR_Y, GR_N and (GR_Y * GR_N) in GR_Q:

While both DAAGRs and CAAGRs are meaningful in agricultural applications, they serve different analytical purposes, with the DAAGRs often being more practical for reporting and comparing results over specific periods. In this study, DAAGRs are selected due to their practicality, data availability, comparability and interpretability (Altinay, 2004; Brick et al., 2016). Because agricultural production often involves processes that have natural and distinct time intervals, such as planting seasons, harvest cycles or breeding seasons for livestock, agricultural data are typically collected at specific intervals (e.g. daily, weekly, monthly, and annually) rather than continuously at every instant, making DAAGRs a logical and practical choice for analysis and reporting. The DAAGR is also a standard metric used by institutions such as the USDA's Economic Research Service to report and compare productivity and output growth over different business cycles or longer time spans. Furthermore, the resulting rate represents the actual change over a defined period, which is intuitive for policymakers, farmers, and researchers to understand and use for planning (e.g. “output grew at an average annual rate of 0.61%”).

On the other hand, CAAGRs are often used to model natural processes, such as underlying biological processes that are always in motion. CAAGRs also have been used in some research or economic modeling where continuous models can offer a more detailed representation of instantaneous changes and underlying dynamics, helping to bridge the gap between theoretical models and real-world phenomena. While theoretically useful, the continuous rate can be less practical for real-world applications because it is difficult to measure growth at every instant, and its interpretation can be less intuitive for non-specialists compared to a simple “percent per year.”

While both growth and variability or instability are important in assessing agricultural development and growth, this study assesses the variability in maple syrup production (Q), yield per tap (Y) and the number of taps (N) using CDVI (Cuddy and Della Valle, 1978):

where CV is the standard coefficient of variation (i.e. standard deviation divided by the mean), and R² is the adjusted coefficient of determination from a simple trend regression of the time-series data. CDVI is used in this study due to the significant upward trends in maple syrup production, yield and number of taps. CDVI that de-trends the data can provide a more accurate measure of the variability than the corresponding CV.

The three regression models, presented in equations (4), (5) and (6) in the previous section, are estimated using the national data and the pooled time-series and cross-sectional data from 2001 to 2025. The estimation results using the two datasets are reported in Table 2 and Table 3, respectively. Note that, in Table 3, the state effects or the variations across the seven states are controlled through six dummy variables with Maine as the reference state.

The results reported in the two tables suggest four major conclusions. First, as indicated by the values of R-squared, adjusted R-squared and F-statistic, the estimated models for production (Q) and number of taps (N) fit the data very well and explained more than 97% of the variation in Q and more than 92% of the variation in N. The major reasons for the lower explanatory power of the two estimated yield models, 74.3% for the national model and 60.0% for the state data, are the weather and other factors not included in the model. While the sap yield each day and the number of days in each sugaring season are highly affected by many weather and climate factors, such as daily high and low temperatures and rain and snow precipitation (Wang et al., 2025b), such factors are not included in the yield models due to the lack of data availability.

Second, while the estimated CAAGRs of N, Y and Q are directly the estimated coefficients of the time variable in the three regression models, the DAAGRs of N, Y and Q are calculated according to the formulas in Equation (7). The results reported in Table 2 and Table 3 empirically confirm that the CAAGR of Q is equal to the sum of the CAAGRs of N and Y (i.e. 4.33% + 2.89% = 7.22% in Tables 2 and 3.73% + 2.83% = 6.56% in Table 3). The numerical results also confirm the relationship in Equation (8) that the sum of the DAAGRs of the taps and yield plus their interaction term is equal to the DAAGR of production (i.e. 4.43% + 2.93% + 4.43%*2.98% = 7.49% in Tables 2 and 3.8% + 2.87% + 3.8%*2.87% = 6.78% in Table 3).

Third, while the estimated DAAGR in maple syrup production from 2001 to 2025 was 7.49% from the national data reported in Table 2, the contributions of taps, yield and their interaction were 59.14%, 39.12% and 1.74%, respectively. According to the estimation results from the pooled data reported in Table 3, the estimated DAAGR of maple syrup production is 6.78% based on the state-level data, and the contributions of taps, yield and their interaction were 56.05% 42.33%, and 1.62%, respectively. Note that these contributions are calculated using the formulas presented in Equations (9), (10) and (11) in the previous section.

Fourth, regarding the state effects reported in Table 3, while Vermont and New York experienced significantly higher growth rates in the number of taps than the five other major production states, Vermont also achieved a higher growth rate in the yield per tap. Although Wisconsin and New York achieved higher growth rates in the yield per tap and total maple syrup production, respectively, the differences are not statistically significant.

The estimation results of CDVI using the national and state data, reported in Table 4 and Figure 3, suggest three major findings. First, the estimated CDVIs are significantly lower than the corresponding CVs, especially for the number of taps (N) and total production (Q), due to their significant upper trends over the study period. These results also suggest that the CV values significantly overestimated the variability levels of the three variables over the period. Specifically, the estimated CVs for N, Y and Q were 408%, 104% and 223% higher than their corresponding CDVIs, respectively. Second, the instability in production is higher than that in both yield (Y) and the number of taps (N), as both the instabilities in Y and N contribute to the variability in Q. Third, there are significant variations in the variability level for all three variables across the seven major production states, but there is no significant correlation between the variability level and the rank of the states by their production levels. Note that, in Figure 3, the seven states are in order from the largest producer (VT) to the smallest producer (NH).

Like many other empirical studies based on secondary data, the quantitative analysis presented in this paper has three major limitations. First, while maple syrup production is highly affected by many weather and climate factors during the sugaring season each year, such factors are not included in our analysis due to the lack of reliable data. Second, in our analysis using state-level data, dummy variables were used to capture the differences across the states due to the lack of data on the factors that may explain the differences in the number of taps and yield per tap across the major production states. Third, in contrast to other field and tree crops, the maple syrup sector includes many small and hobby producers and their inclusion in the USDA-NASS data likely affected the average yield per tap. While the average yield of hobby producers is significantly lower than the large commercial producers (Cannella et al., 2022), the aggregate yield used in this study likely underestimated the average yield of maple syrup as an industry.

Our decomposition analysis using national and state-level data indicates that for the steady growth in US maple syrup production from 2001 to 2025, 56.05–59.14% of the growth was from an increased number of taps, 39.12–42.33% from improved yield and 1.62–1.74% from their interaction. These decomposition results are significantly different from many other farm products with more than 95% of their production growth from yield improvement alone. For example, Zulauf et al. (2024a, 2024b) reported that 100% of the production growth in US field crops (i.e. feed grains, food grains and oilseeds) from 2000 to 2002 to 2020–2022 was due to yield improvement, because the increase in farmland over the period was almost zero. Global comparisons presented in the same study also suggested that agricultural growth in more developed regions, like Europe and the United States, was driven by yield improvement, while the growth in less-developed regions such as South America and Sub-Saharan Africa was more dependent on land expansion (Zulauf et al., 2024a, b). The finding about agricultural growth in less-developed regions was also supported by Parmar and Devi (2021) who decomposed soybean production in Gujarat, India. They found that the increase in soybean production in the region from 2010 to 2011 to 2019–2020 was 66.94% from increased acreage, 20.31% from improved yield, and 12.75% from their interaction.

While yield improvement has dominated the growth of major agricultural crops in the United States, it has contributed only about 40% of the growth in maple syrup production since 2001. Also, land costs and associated constraints, which are major challenges for many agricultural crops, have not yet presented the same challenges for maple syrup production. Consideration of trends in competing timber products and forested land use in the Northeastern regions in the United States provides valuable insights into the factors influencing expanded forest acreage utilization and the prospects for continued expansion. As reported by Farrell and Chabot (2012), only 0.4% of viable maple trees in the United States were being tapped. In the state of Vermont, the largest producer in the country, only 2.94% of viable maple trees were being tapped. However, a large proportion of untapped forest acres are likely under long-term timber management plans that are typical of land use tax abatement programs, meaning a conversion from timber to non-timber maple management is necessary. This highlights unique policy implications for maple syrup as a non-timber forest product that is significantly different from field crop products. Continued forest use conversion to maple sap and syrup may magnify the lack of updated forest management guidelines for professional foresters that challenge the regulatory oversight of forest parcels (Isselhardt et al., 2022).

Public investments in the maple syrup sector must distinguish between yield enhancement and tapping more trees, depending on the regional audience of producers and the prevailing forest land use and competition for forest access. While national statistics have shown significant differences in production variability and yield levels across US states (USDA, 2026), research has proven that modern powered vacuum sap collection systems are capable of increasing annual sap yield incrementally as the vacuum level in the tubing system increases (Wilmot et al., 2007; van den Berg et al., 2016) and consistently outperform gravity (non-vacuum) bucket or tubing collection (Abah et al., 2026).

Because natural maple sap flow is highly influenced by climatic factors (Thapa et al., 2024), commercial maple syrup producers with large financial investments in sap harvest technology or managing modern technology more intensively have consistently demonstrated higher yields than traditional producers. For example, a multi-year financial and management benchmarking program showed that maple enterprises with higher capital investment levels or more intensively managed vacuum sap collection systems were able to achieve higher overall yields and more consistent year-to-year performance (Cannella et al., 2017; Cannella and Lindgren, 2023).

Programs can facilitate adoption of modern sap collection practices that are proven to moderate variability and downside risk to total production and yield rather than encourage tap expansion. In regions most likely to continue the expansion of total taps, updated policies and governance overseeing the conversion of forest management from timber to non-timber forest products needs to become a priority. Also, policy analysis can assess the economic returns and environmental impacts of public investment in improving the yield per tap versus the investment in increasing the number of taps. These analyses must consider stakeholder input related to changes in timber harvest, recreation, and other forest utilization activities prevalent across maple production states.

Policy and development initiatives for maple development or regulation must also accommodate a wide range of business typologies based on scale or management practices. For example, the national aggregate data for average yields used in this study is notably lower than typical modern maple enterprises, which project a break-even target yield of 0.50 gallons per tap. The USDA dataset and the overall maple community include many producers not utilizing modern technology to its fullest potential. Policies and programs can redress deficiencies or enhance new opportunities across the entire industry, but new programs must consider the management characteristics and enterprise scale of different business typologies to achieve program objectives. For example, energy efficient equipment support programs and their anticipated benefit outcomes vary widely when designed for small, medium or larger enterprises.

Our empirical assessment of variability of maple syrup production, yield and number of taps, based on CDVI, suggests that all three variables, especially yield and production, experienced significant instability over the study period. Also, the estimated CDVIs for the seven major production states indicate that a state's scale of production does not significantly affect its variability in yield, number of taps or production. For example, Vermont produced about 50% of the maple syrup in the United States, but its large production does not make its CDVIs lower than those in other states. Instead, the variability in maple yield per tap is mainly caused by weather and climate factors (Wang et al., 2025a). Therefore, there is a growing need for information on how weather and climate factors affect sap production and tree health and on technical innovations for more climate-resilient technologies. Overall domestic production variability presents concerns to the national commercial distribution agreements needed to market the ever-increasing crop. Any disruption to domestic supply will complicate pre-existing business contracts and prevailing price agreements which can have consequences for long-term market growth. Another strategy for stabilizing producer prices and mitigating supply risks is to implement a maple production quota and reserve system like the one implemented in Quebec, Canada, in 2004, though it may lead to lower production growth and prices in the long-run (Wang et al., 2025a).

The maple syrup industry, especially the small and medium-sized producers, needs more support in research, education and outreach. Many maple syrup producers would benefit from information and programs that help them adopt new technologies, management strategies, reduce costs, enhance forest management, access new markets and participate in government programs (Caughron et al., 2021; Hershberger et al., 2024; Velardi et al., 2023). Our findings indicate that the maple syrup sector is subject to risk and uncertainty, but there are limited options, such as insurance programs, for maple syrup producers to mitigate risk and uncertainty. For example, although various federal programs, including the Whole Farm Revenue Protection (WFRP) and Noninsured Crop Disaster Assistance Program (NAP) for weather-related losses, are available to maple syrup producers, many small producers need assistance in understanding and participating in such programs.

This empirical study decomposed the growth in US maple syrup production from 2001 to 2025 into contributions of the number of taps, yield per tap and their interaction, assessed the variability of production, yield and the number of taps, and discussed potential economic and policy implications for more sustainable growth of the industry. Specifically, using the most recently available data at the national level and in seven major production states from 2001 to 2025, regression analysis was conducted to decompose the growth, and CDVIs were estimated to assess the variability in production, yield and taps. Results based on the national data suggest that, with an estimated DAAGR of 7.49% in production, the contributions from increased taps, yield improvement and their interaction were 59.14%, 39.12% and 1.74%, respectively. Using state-level data, the three corresponding contributions were 56.05%, 42.33% and 1.62%, respectively, with a slightly lower DAAGR of 6.78% in production. While the estimated CDVIs for national production, taps and yield were 0.157, 0.062 and 0.114, respectively, the estimated values for the seven states showed significant variations across the states.

This study suggests two major economic and policy recommendations. First, while US agricultural and environmental policy and investment have primarily focused on yield improvement rather than increasing acreage, this study suggests that these policy and investment focuses should be modified for the maple syrup sector. Specifically, policy and programs should distinguish between increasing the yield per tap and expanding the number of trees tapped, depending on factors such as forest endowment, utilization, access and the most pressing risks in each region. Second, sustainable growth of the industry will benefit from targeted support for the development and adoption of climate resilient technologies and management strategies that seek to reduce the variability of yield as an alternative to tap expansion. For example, in certain locations with low elevation gradient (level ground), adoption of remote (non-grid) power generation like portable generators can support powered vacuum sap collection systems that will moderate seasonal variability of yield and reduce business risks. For maple forests with high elevation gradient but limited power access, conversion of 5/16-inch sap collection tubing to the innovative 3/16-inch diameter tubing can produce an enhanced “natural vacuum” without mechanized pumps that will promote sap flow during marginal weather time periods when sap flow may otherwise stop.