Abstract

Phosphorus (P) is an essential plant nutrient, but it can pollute rivers and lakes. State laws banning P in lawn fertilizer can reduce pollutant loads from urban areas. A random effects panel probit model using nationwide data from various sources was estimated to analyze factors affecting passage of these laws. To test hypotheses regarding alternative theories, we incorporated variables relating to public interest, private interest, ideology, and diffusion literatures and found support for all of them. States with higher percentages of water area were more likely to pass P-free laws, as were those with higher percentages of employment in water-related industries. Surprisingly, states with more fertilizer companies were more likely to pass P-free laws, but the effect was quadratic. We found an S-shaped diffusion curve as a function of time. No laws have been passed since 2013, when Scotts dropped P from their Turf Builder fertilizers.

Concerns about water pollution have increased in the United States, due in part to excessive use of fertilizer by farmers and households (National Science and Technology Council [NSTC], 2016). The main water quality problem caused by excessive nutrients is harmful algal blooms (HABs). These can damage not only the environment and a state’s public health but also its economy. Specifically, HABs could negatively influence the drinking water as well as the tourism and fisheries industries. The economic impacts from the affected sectors have resulted in millions of dollars of losses in the United States (Environmental Protection Agency [EPA], 2015; Lopez, Jewett, Dortch, Walton, & Hudnell, 2008; NSTC, 2016). For example, the tourism industry in Ohio near Grand Lake St. Marys experienced revenue reductions of about $35–$45 million due to algal blooms in 2009 and 2010 (Davenport & Drake, 2011). Mistiaen, Strand, & Lipton (2003) estimated reduced revenue of crab harvests in Patuxent River, MD, due to hypoxia (i.e., when water has low dissolved oxygen, between 1% and 30% of its potential). Also, Kashian & Kasper (2010) estimated property values near Wisconsin lakes that frequently experienced algal blooms; they found that properties near these lakes had lower prices, between $159 and $414 less, than properties near cleaner lakes.

The most problematic nutrient in fertilizer relative to algal blooms is phosphorus (P). Excessive use of P fertilizer is threatening water quality because it is generally the limiting nutrient in a freshwater system (Scavia, Allan, Arend, Bartell, Beletsky, Bosch, Brandt, Briland, Dalog˘lu, DePinto, Dolan, Evans, Farmer, Goto, Han, Höök, Knight, Ludsin, Mason, Michalak, Richards, Roberts, Rucinski, Rutherford, Schwab, Sesterhenn, Zhang, & Zhou, 2014; Schindler, Hecky, Findlay, Stainton, Parker, Paterson, Beaty, Lyng, & Kasian, 2008). P can come from both agricultural and urban landscapes, and the relative importance of each source will vary with location. P can also come from nonfertilizer sources such as leaf litter and pet waste Hobbie, Finlay, Benjamin, Nidzgorski, Millet, & Baker, 2017). The Chesapeake Bay Program [CBP] (2010) estimated that urban/suburban runoff was the source of about one-third of P in the bay, and the only source that was increasing. Moreover, much of the P applied to lawns is unnecessary. Established lawns usually do not need P because N, not P, is the limiting nutrient in the soil (Bigelow, Tudor, & Nemitz, 2012). Many studies find that reducing phosphorus inputs can lead to better water quality (Hobbie et al., 2017; Jeppesen, SØndergaard, Jensen, Havens, Anneville, Carvalho, Coveney, Deneke, Dokulil, Foy, Gerdeaux, Hampton, Hilt, Kangur, Köhler, Lammens, Lauridsen, Manca, Miracle, Moss, Nõges, Persson, Phillips, Portielje, Romo, Schelske, Straile, Tatrai, Willén, & Winder, 2005; Kara, Heimerl, Killpack, van de Bogert, Yoshida, & Carpenter, 2012; King, Balogh, Agrawal, Tritabaugh, Ryan, 2012; Schindler et al., 2008). Fertilizers without P are available in stores; however, voluntary adoption of P-free fertilizers is currently quite low (Brehm, Pasko, & Eisenhauer, 2013; McCann and Shin, 2018).

Reducing P applications on established lawns can improve water quality without negative effects on lawn health.1 Eleven states have passed a law that bans the use of P fertilizer on lawns in urban areas: Illinois, Maine, Maryland, Michigan, Minnesota, New Jersey, New York, Vermont, Virginia, Washington, and Wisconsin (Miller, 2012). In most cases, exceptions are made for new lawns and if soil test results indicate a P deficiency. There is evidence that these laws have been effective.2 Vlach, Barten, Johnson, & Zachay (2010) found that P-free fertilizer helped to reduce the total phosphorus (TP) level from urban residential areas in Minnesota by 12–15 percent. Lehman, Bell, Doubek, & McDonald (2011) found that a lawn fertilizer ordinance near Ann Arbor, MI, was effective, reducing TP by 11–23 percent. In 2013, a leading lawn fertilizer company, Scotts Miracle-Gro, announced that it was removing P from its Turf Builder lawn fertilizer product in 2013, and many companies have followed suit. Since then, no further P-free laws have been passed. However, the Pennsylvania Senate recently passed a bill banning P in lawn fertilizer, and a vote in the House is scheduled for October 2018 (General Assembly of Pennsylvania, 2018; Reitman, 2018).

Controlling the pollution caused by nutrients, such as P, is important and some states are trying to address the issue. To our knowledge, there is no research analyzing state regulation of phosphorus in lawn fertilizer. The objective of this study was to identify the factors affecting the passage of phosphorus-free lawn fertilizer laws in the United States to help predict how future regulatory decisions relating to nutrients will likely proceed.

The political economy of environmental policy has been studied for various topics, such as the Clean Water Act (Helland, 1998), the Acid Rain Program (Perino and Talavera, 2014), pesticide regulation (Cropper, Evans, Berardi, Ducla-Soares, & Portney, 1992), the Endangered Species Act (Ando, 1999), renewable energy (Lyon and Yin, 2010), and hydropower licensing (Kosnik, 2010). These studies have found that political factors affect the enforcement of environmental standards. Three theories of regulatory influence have been examined: public interest, regulatory capture, and ideology. However, some studies focused on only one theory, using the relevant variables for that theory. In our study, we included variables related to all three political economic categories in the analyses to identify which, if any, have affected the adoption of P-free laws. We also examined individual states’ behavior, rather than that of one legislative agency, like the EPA. In addition to political economy factors, we considered the effect of diffusion over time (Geroski, 2000; Griliches, 1957; Rode & Weber, 2016; Rogers, 2003), which can influence policy decision making (Walsh, Bird, & Heintzelman, 2015).

We found that the influences on P regulatory decisions were related to public interest, private interest, ideology, and diffusion factors. States having more water area were more likely to require P-free fertilizer, showing the effect of potential pollution problems. States with higher employment in water-related industries were more likely to pass P-free laws, but the effect was nonlinear. Somewhat surprisingly, states with more fertilizer firms were more likely to pass P-free laws. States with U.S. senators having stronger pro-environmental ideology were more likely to pass a law. Also, we found an S-shaped diffusion curve as a function of time.

Background on phosphorus

Before reviewing the literature to generate specific hypotheses, we provide information on the characteristics of phosphorus. P is an essential nutrient for plants, as it has a key role in energy metabolism. Insufficient P results in smaller leaf size, lower photosynthesis rate (Zambrosi, Ribeiro, Marchiori, Cantarella, & Landell, 2015), and slower growth (Pagliari, Kaiser, & Rosen, 2018). However, excessive use of P can cause water pollution. Rapid growth of algae is caused when P levels in lakes and rivers increase, which is called eutrophication. As the abundant algae die and decompose, the oxygen in the water is depleted. The low oxygen levels can then cause the death of fish and other aquatic life.

In nature, P exists as phosphate (PO43-) because it is rapidly combined with oxygen in the air. As shown in Figure 1, there are three types of P in soils, soluble P (immediately available), active P (readily available), and fixed P (low availability; Syers, Johnston, & Curtin, 2008; Pagliari et al., 2018). Active P and fixed P compounds exist in solid form because their solubility is low, but plants can take up soluble P, which is 0.1% of total P in soil (Zou, Binkley, & Doxtader, 1992). As soluble P concentration decreases due to plant uptake, some phosphate in active P is released to the soil as soluble P. Excessive application of P may actually increase fixed P by reducing the release of active P to soluble P (Pagliari et al., 2018).

Phosphorus fertilizer contains soluble phosphorus. After fertilizer is applied, some portion of phosphate is taken up by the plant, but some reacts with elements in the soil to form nonsoluble P. Many soils have high P levels because of accumulation due to applications in excess of plant needs. Therefore, soil erosion and runoff to nearby lakes or rivers is a large contributor of P in urban areas. Also, detecting P change takes some time due to its relatively stable characteristics. Lehman et al. (2011) mentioned that detecting the effects of P reduction varies depending on the type: total phosphorus (within 2 years; including soluble, active, and fixed P), soluble P (2–3 years), and active P (about 8 years).

FIGURE 1

Phosphorus cycle in soil. Adapted from Syers, Johnston, & Curtin (2008).

FIGURE 1

Phosphorus cycle in soil. Adapted from Syers, Johnston, & Curtin (2008).

Close modal

The effects of reducing P inputs on water quality have been studied by many researchers (Jeppesen et al., 2005; Kara et al., 2012; King et al., 2012; Schindler et al., 2008). For example, King et al. (2012) found that reducing the fertilizer application rate on turf grass by more than 75% and using organic P is effective at reducing soluble P and TP. Given the potential reductions in P loadings that could result from decreased P application, policies to reduce negative externalities may be necessary. Therefore, the literature review examines alternative theories relating to environmental regulation.

Literature review

Previous studies have demonstrated the importance of political factors in regulation. Even though most of the studies in economics have focused on regulatory agencies and enforcement, political characteristics also affect the adoption of environmental policy by legislative bodies (Walsh et al., 2015).

Three alternative political economy theories of regulation have been studied. The first is called the “public interest theory,” in which regulations are to solve market failures and reflect public interests (Joskow & Noll, 1981). The second is “private interest or capture theory,” in which regulations are structured to maximize regulators’ or industries’ benefits (Stigler, 1971). The third is “theory of ideology,” in which ideology drives political and regulatory decisions (Kalt & Zupan, 1984).

Public interest theory

The public interest theory argues that regulation results from the need to protect the public from negative externalities, such as pollution (Joskow & Noll, 1981). There is no incentive for firms to reduce negative external costs to society without government regulation. The theory considers the government as a social welfare maximizer through regulation.

In the literature, public interest variables are shown to have a significant effect. Environmental quality is used in various studies and has been shown to increase the likelihood of regulation. For example, Perino & Talavera (2014) found that states with more severe acid rain issues tend to have stronger regulatory standards in the Acid Rain Program, and similar results were found by Matisoff & Edwards (2014) for energy policy and Daley & Garand (2005) for the federal Superfund program.

Water pollution can cause a reduction in some industries’ size or profitability and have other negative effects, such as job loss, reducing a state’s welfare. However, investigation of the effect of the percentage of employment in the affected industry or unemployment rate on regulation shows mixed results. States with higher employment in a water-related industry tend to be less likely to have strict regulations (Helland, 1998; Joskow & Schmalensee, 1998), but there are also studies that show the opposite relationship to the percentage of employment in the related industries. For example, Walsh et al. (2015) studied local fracking bans and found a negative and significant relationship between employment in arts and tourism and adoption of the regulation.

Lastly, natural resource endowment could be a potential factor affecting water pollution policy. For example, Walsh et al. (2015) used two natural resource variables (open water and wetlands) to identify the relationship with local fracking regulation, but they found mixed results, a positive relationship with open water but a negative relationship with wetlands.

Private interest theory

Stigler (1971) argued that regulation comes from the industry’s demand and is designed primarily for the industry’s benefit. Becker (1983) extended the capture theory, showing that regulations can be enabled by competition among pressure groups, not by a single group.

In the literature, interest group variables have shown an effect on regulation. The variables and relationship vary by study. For example, Perino & Talavera (2014) found that if the mining industry is being negatively affected by acid rain regulation, the stringency of regulation is likely to decrease. However, there are also studies that find a negative but nonsignificant relationship between the number of manufacturers as an interest group and the stringency of enforcement under the Clean Water Act (Helland, 1998).

Theory of ideology

After the public and the private interest theories of regulation, researchers started to be interested in the ideological behavior of politicians. There are two types of ideology, citizen and legislator, such as that of senators. Researchers found that senators’ ideology is the key determinant in predicting their voting patterns (Kalt and Zupan, 1984; Levitt, 1996). Also, Poole & Rosenthal (1984) found that senators from the same state but different parties (Democrats versus Republicans) have very different behavior. For example, Perino and Talavera (2014) showed that if there is a greater share of Democrats in a state’s legislature, it tends to have more stringent environmental regulation under the Acid Rain Program.

Furthermore, environmental ideology is important for the type of regulation. In the literature, the environmental ideology of legislators is shown to have a significant effect on listing of endangered species (Ando, 1999) and coal strip-mining regulations (Kalt & Zupan, 1984). For instance, Ando (1999) found that pro-environmental ideology has a positive effect on listing of endangered species.

Diffusion effect

The literature on adoption of new technologies and environmental practices by individuals shows that diffusion of a practice occurs over time (Rogers, 2003). Economists generally view diffusion as cumulative adoption. The diffusion process has three phases. Initially, there are only a small number of adopters, but the adoption rate takes off after the introductory phase. Later, the diffusion slows because there are fewer potential adopters than in the other phases and the innovation may not be appropriate for the whole population. The diffusion effect has been shown to be relevant for adoption of agricultural practices. For example, Fuglie & Kascak (2001) found S-shaped curves for the cumulative adoption of conservation tillage, integrated pest management and soil nutrient management, and similar results were found for hybrid corn by Griliches (1957) and Ryan & Gross (1943). Diffusion may also occur with states’ adoption of regulations. Grossback, Nicholson-Crotty, & Peterson (2004) argued that as more states adopt particular laws, it provides more information and thus less uncertainty. Municipal or county-level ordinances may also facilitate adoption of laws at the state level by providing information on their effectiveness and acceptability. For example, in Minnesota, a P-free ordinance in Plymouth preceded the state-wide ban by three years (Lehman et al., 2011).

Other variables

Beyond the political economic variables, average household income can affect the adoption of laws. Walsh et al. (2015) found that the more affluent a state is, the more likely it is to have a law banning fracking, and similar results were found by Lester, Franke, Bowman, & Kramer (1983) for hazardous waste. However, there are other studies that found a nonsignificant or opposite relationship (Helland, 1998; Krause, 2011). The information from the literature will be used to develop the hypotheses presented in the conceptual framework section.

Conceptual model

Based on the public interest theory literature (Daley & Garand, 2005; Helland, 1998; Perino & Talavera, 2014), poor environmental quality would increase P regulation because it could negatively impact public health and environmental amenities. States with more water area may also increase the stringency of regulation, as found for fracking (Walsh et al., 2015). Thus, states with poor water quality and more water area are expected to be more likely to adopt P-free-fertilizer laws.

In addition to environmental benefits for citizens, P-free fertilizer could also be beneficial for the state’s economy. If water quality in a state is getting worse, then the profitability of industries dependent on water, such as fishing and tourism, may be reduced. Pollution could have a negative effect on the economy of the state, due to job losses and lower profits. The fishing industry would be directly impacted by poor water quality, such as shellfish bed closures, while the tourism industry would indirectly experience negative effects on lodging and restaurants near the impacted water bodies. Therefore, states with a higher share of employment in water-related industries are expected to be more likely to adopt P-free laws.

Private interest theory literature (Helland, 1998; Perino & Talavera, 2014) shows that the presence of a regulated interest group would decrease regulation. The expected interest group for P-free laws is lawn fertilizer companies that produce fertilizer including P. Costs of lawn fertilizer companies may increase due to P-free laws because they may need to change their production facilities, or people may learn about the change and perceive fertilizer use in general as bad. If there are many lawn fertilizer producers in a state, the legislatures may have a greater hurdle to overcome in adopting P-free laws. On the other hand, the price of phosphorus doubled between 2005 and 2013 (USDA-ERS, 2013). In this case, lawn fertilizer firms may even support P-free laws to reduce input costs. Additionally, smaller groups tend to be more well-organized than larger ones due to free riders (Olson, 1965; Pecorino, 2015). We include the number of fertilizer firms to capture the effect of industry group pressure, but the direction of the impact is unclear.

Based on the ideology literature (Ando, 1999; Kalt & Zupan, 1984), legislators who have a more pro-environmental ideology would increase the likelihood of passing environmental regulations. States with more pro-environmental senators are thus expected to be more likely to adopt P-free laws.

Diffusion refers to the spread of adoption over time (Rogers, 2003). Dissemination of information about P-free laws and their effectiveness would be important in the policy decision-making process. As explained earlier, due to the nature of P, environmental impacts of P-free laws are not immediate. States may also want to see the economic and political impacts of such laws in other states before passing P-free laws. Therefore, diffusion of P-free laws may take several years.

Lastly, states with higher median income are more likely to adopt P-free laws because of demand from constituents for better environmental quality (Panayotou, 2016). Demand theory implies that demand for a normal good increases with income, all else equal.

Data

For this research, we used yearly state data that included all 50 states from 2005 to 2013 (the first such law was passed by Minnesota in 2002 and became effective in 2004). The dependent variable was whether a state had passed the law and the year (1 = law passed that year and all subsequent years, and 0 = otherwise).

The independent variables are grouped into four categories: public interest, private interest, environmental ideology, and states’ characteristics. For public interest, environmental quality is the maximum level of total phosphorus (TP)3, lagged by one year due to P’s physical characteristics. We used monitoring data from a freshwater body such as a lake or river. The TP data are from the Water Quality Portal (a cooperative effort of the United States Geological Survey, the Environmental Protection Agency, and the National Water Quality Monitoring Council). Since not all states had water quality data for all years,4 this reduced the number of observations. Additionally, the water area5 data were gathered in 2010, 2012, and 2013. We used average values for each state of these three years because all states had data for these years and there was no large variation in water area by state (all states’ standard deviations were less than 0.02).

The employment in water resources-based industries is proxied by the share of employment in agriculture, fishing, and natural resources. Using employment in just fishing would be ideal, since it is most directly linked to water quality, but only aggregated data with agriculture, fishing, and natural resources were available for all states and years. The share of employment in accommodation and food is used as a proxy for the tourism industry. The squares of the employment variables were also included due to the possible existence of nonlinearity. The employment information and water area are from the U.S. Census.

The private interest variable is the number of lawn fertilizer producing companies in the state, representing an interest group. The information on existing fertilizer companies in the United States was gathered from the Websites of major lawn fertilizer retailers, such as Lowe’s, Home Depot, Walmart, and Amazon. For those companies, we then checked the labels and their Websites to find out where they produced their fertilizers. We excluded companies if they did not have any lawn fertilizer products containing P in 20176 (e.g., only nitrogen fertilizers). We used a quadratic form for the number of fertilizer firms to capture the possibility of nonlinearity.

For environmental ideology, the national environmental scorecard from the League of Conservation Voters (LCV) is used.7 This measure for each U.S. senator is the index calculated by dividing the number of pro-environment votes8 by the total number of votes. The value ranges from 0 to 100, with 0 being most anti-environment and 100 being most pro-environment. Senators do not make state regulations, but our assumption is that senators are likely to reflect the overall views of the population, more so than representatives, and senators have been used in other studies.

To capture the diffusion effect over time, we used time dummies with the first year in our dataset, 2005, as the base. Lastly, as a control variable for states’ characteristics, we used median household income from the U.S. Census from 2005 to 2013.9

Summary statistics for the variables are shown in Table 1. On average, TP was 10.89 mg/L. California, Minnesota, and Ohio had levels that exceeded 40 mg/L TP. Employment in agriculture, forestry, and fishing was 1.92% on average; South Dakota, North Dakota, and Montana had more than 5% of employment in these industries. On average, tourism-related industries accounted for 7.10% of employment; and Nevada and Hawaii had more than 10%. Percentage of water area overall was 7.52%, and Michigan, Hawaii, and Rhode Island had more than 30%. Environmental ideology was 52.08 and Maryland, Rhode Island, Vermont, and California had LCV values (pro-environment) of more than 95. For the number of lawn fertilizer companies in a state, California had the most, with seven. Median household income was $49.64 thousand; and Maryland, New Jersey, Connecticut, and Alaska had a median income of more than $65 thousand on the average between 2005 and 2013.

TABLE 1

Summary statistics (n = 427, 50 states for years 2005–2013).

VariableMeanStd. dev.MinMax
P-free law adoption (dependent variable)  0.11 0.32 
Lagged max total phosphorus (TP) a 10.89 45.41 0.01 678.00 
Percentage of water area (average of 2010, 2012, 2013) b 7.52 9.17 0.24 41.54 
Employment in agriculture, forestry, and fishing (%) 1.92 1.57 0.24 7.49 
Employment in accommodation and food services (%) 7.10 1.62 4.88 17.15 
Number of fertilizer companies 0.55 1.32 0.00 7.00 
League of Conservation Voters Score (LCV Senate) 52.08 34.73 0.00 100.00 
Median household income (thousand dollars) 49.64 8.30 32.94 72.48 
VariableMeanStd. dev.MinMax
P-free law adoption (dependent variable)  0.11 0.32 
Lagged max total phosphorus (TP) a 10.89 45.41 0.01 678.00 
Percentage of water area (average of 2010, 2012, 2013) b 7.52 9.17 0.24 41.54 
Employment in agriculture, forestry, and fishing (%) 1.92 1.57 0.24 7.49 
Employment in accommodation and food services (%) 7.10 1.62 4.88 17.15 
Number of fertilizer companies 0.55 1.32 0.00 7.00 
League of Conservation Voters Score (LCV Senate) 52.08 34.73 0.00 100.00 
Median household income (thousand dollars) 49.64 8.30 32.94 72.48 
a

Some states’ data were missing for some years. See note 4 for details.

b

The water area data were gathered in 2010, 2012, and 2013 and averaged.

Methods

Probit, logit, and hazard models have been used to identify the adoption of a policy as a function of factors that could influence decision making. Sapat (2004) pointed out that the hazard model may not be appropriate if a policy is not homogeneous across states. Phosphorus-free fertilizer legislation has some variation by states. For example, Maryland, New Jersey, and Vermont, which adopted it in 2011, have an additional limitation on nitrogen and there are some distinctions regarding exemptions and point-of-sale restrictions (Table 2).

TABLE 2

P-free laws: passed and effective years and exemptions (all states listed had passed P legislation by 2017).

TABLE 2

P-free laws: passed and effective years and exemptions (all states listed had passed P legislation by 2017).

Close modal

Results from probit and logit models are similar because the cumulative probability functions used in the models are almost identical. In this paper, we used a random effects panel probit model to estimate the likelihood of policy adoption, due to the dichotomous nature of the dependent variable and use of panel data. If a state adopts the policy, or has previously adopted the policy, then the independent variable is 1; otherwise (does not adopt) it is 0. Thus, the equation will be

yit=xitβi+vi+εit(1)...

 

For i = 1, . . . , n states, and t = t1, . . . , tT, xit = (x1t, . . . , xkt) is a transposed vector including independent variables, and β = (β1, . . . , βk)' is a vector of coefficients. The vi are individual (state) specific time-invariant random effects to control for unobserved heterogeneity10 and are assumed to be independently and identically distributed (i.i.d.), N(o, σv2). εit are the error terms with the assumption of i.i.d. and a normal distribution, N(o, σε2=1). The probability can be denoted as pi = P(Y = 1|X). The random effects panel probit regression is thus

P(yit=1|xit)=φ(xitβ+vi),(2)...

 

where ϕ(·) is the cumulative normal distribution.

Results and discussion

We examined correlation coefficients for the explanatory variables as well as the variance inflation factors (VIF) to check for multicollinearity in the regression. The correlation coefficients were smaller than 0.48 in absolute value for all variable pairs, except the three squared variables. There was no evidence of multicollinearity when the VIF were examined, because all variables had VIF values less than 3,11 except squared terms.12 The chi-squared value indicated that the model was significant (p < 0.001). Also, the pseudo-R2 value for the random effects probit regression was 0.71 and ρ13 was nonzero, which means that the panel estimator is different from the pooled estimator14 (Table 3). The Akaike information criterion (AIC)15 was 95.70, which was a lower value than when government ideology was used (104.4, data not shown).

Random effects probit regression results and average marginal effects for the final model are shown in Table 3. The model measures the conditional probability changes in adoption of P-free laws by states, given the independent variables.

For the public interest variables, more water area had a positive effect on P-free laws (p < 0.01), consistent with the research by Walsh et al. (2015) that states with more natural resources were more likely to pass a fracking law. The magnitude of the effect can be seen in the marginal effects column. As the percentage of water area increased by 1 percent, the likelihood of passing legislation went up by 0.4 percentage points.

TABLE 3

Results of random effects probit model of P-free law adoption.

VariablesCoefficientsMarginal effects
Lagged max TP −0.026 −0.000 
 (0.044) (0.000) 
Water area (%) 0.363*** 0.004* 
 (0.129) (0.002) 
Employment in agriculture, forestry, and fishing (%) 14.596** 0.040* 
 (5.801) (0.023) 
Square of (agriculture, forestry, and fishing) −3.137*  
 (1.631)  
Employment in accommodation and food services (%) 92.357** −0.001 
 (37.608) (0.040) 
Square of (accommodation and food services) −6.867**  
 (2.722)  
Number of fertilizer companies 14.992*** 0.109** 
 (3.963) (0.049) 
Square of (number of fertilizer companies) −3.995***  
 (1.302)  
League of Conservation Voters (LCV Senate) 0.142*** 0.001** 
 (0.040) (0.001) 
Median household income (thousand dollars) 0.242* 0.002 
 (0.128) (0.002) 
2006 (base year of 2005) −0.556 −0.005 
 (3.495) (0.034) 
2007 1.634 0.016 
 (4.355) (0.045) 
2008 −0.519 −0.005 
 (3.913) (0.038) 
2009 3.425 0.034 
 (4.025) (0.041) 
2010 13.018*** 0.128** 
 (4.229) (0.061) 
2011 16.416*** 0.161** 
 (4.850) (0.074) 
2012 17.645*** 0.173** 
 (4.930) (0.076) 
2013 16.947*** 0.167** 
 (5.038) (0.077) 
Constant −369.790***  
 (130.598)  
ρ 0.985  
Observations 427 427 
Pseudo-R0.715  
AIC 95.70  
VariablesCoefficientsMarginal effects
Lagged max TP −0.026 −0.000 
 (0.044) (0.000) 
Water area (%) 0.363*** 0.004* 
 (0.129) (0.002) 
Employment in agriculture, forestry, and fishing (%) 14.596** 0.040* 
 (5.801) (0.023) 
Square of (agriculture, forestry, and fishing) −3.137*  
 (1.631)  
Employment in accommodation and food services (%) 92.357** −0.001 
 (37.608) (0.040) 
Square of (accommodation and food services) −6.867**  
 (2.722)  
Number of fertilizer companies 14.992*** 0.109** 
 (3.963) (0.049) 
Square of (number of fertilizer companies) −3.995***  
 (1.302)  
League of Conservation Voters (LCV Senate) 0.142*** 0.001** 
 (0.040) (0.001) 
Median household income (thousand dollars) 0.242* 0.002 
 (0.128) (0.002) 
2006 (base year of 2005) −0.556 −0.005 
 (3.495) (0.034) 
2007 1.634 0.016 
 (4.355) (0.045) 
2008 −0.519 −0.005 
 (3.913) (0.038) 
2009 3.425 0.034 
 (4.025) (0.041) 
2010 13.018*** 0.128** 
 (4.229) (0.061) 
2011 16.416*** 0.161** 
 (4.850) (0.074) 
2012 17.645*** 0.173** 
 (4.930) (0.076) 
2013 16.947*** 0.167** 
 (5.038) (0.077) 
Constant −369.790***  
 (130.598)  
ρ 0.985  
Observations 427 427 
Pseudo-R0.715  
AIC 95.70  

Note: Standard errors in parentheses.

*** p < 0.01, ** p < 0.05, * p < 0.1.

Employment in both categories of water-related industries (natural resources and tourism) had a quadratic relationship with adoption of the regulation. The sign on the quadratic or squared terms means that the effects of employment in these industries are increasing, but diminishing as the share of employment increases, which is not what we would expect if the motivation for P-free laws were primarily for the public interest. This surprising result may be explained by Becker (1983), who argued that political effectiveness comes from controlling free riding, but that there will be diseconomies of scale as the size of interest groups expands, so controlling free riding becomes more difficult. Also, Peltzman (1976) pointed out that if an interest group is large and diffuse, it may not be as effective. Based on the marginal effects16 column, employment in agriculture, forestry, and fishing had a fairly large effect on adoption. The value of 0.04 means that if the industry size increased by 1 percentage point, the state was about 4 percentage points more likely to adopt the law. It should also be noted that agricultural industries might oppose fertilizer regulations even if employment in agriculture were small, and even if the fertilizer regulations were directed at homeowners.

The regression results showed that the number of fertilizer companies also had a quadratic relationship with adoption of P-free laws, but the effect was positive, rather than negative, as expected. It may be that fertilizer companies also benefited from these regulations. The price of super-phosphate had doubled between 2005 and 2013 (USDA-ERS, 2013), so companies could reduce costs without impacting fertilizer performance for established lawns. Thus, lawn fertilizer firms may actually support P-free laws and produce P-free lawn fertilizer products, but, as with employment, the effect diminishes with larger numbers of companies.17 Some companies that sell products across state lines may also support more uniform regulations. Figure 2 (plotted using marginal effects) shows that the likelihood of adoption was increasing with fertilizer firms in a state but the effect begins to decrease after two firms. The cost of P also explains the finding of Stahlman and McCann (2012) that while few states had implemented laws requiring phytase in nonruminant rations, feed companies had begun to use it in the early 2000s to reduce dicalcium phosphate costs. Examining the marginal effects column, the number of fertilizer companies also had a large effect on adoption of P-free laws: one additional firm resulted in a state being about 11 percentage points more likely to adopt the law.

The environmental ideology variable was positively correlated with the adoption of P-free laws (p < 0.01), consistent with Ando (1999) and Kalt and Zupan (1984), who found that states with more senators who had a pro-environmental ideology were more likely to pass environmental regulations. However, the marginal effect was small. The sign on median income was positive, as expected.

FIGURE 2

Marginal effects with respect to the number of fertilizer companies. The solid line is the average marginal effect and the shaded band is the 95% confidence interval computed by the delta method.

FIGURE 2

Marginal effects with respect to the number of fertilizer companies. The solid line is the average marginal effect and the shaded band is the 95% confidence interval computed by the delta method.

Close modal
FIGURE 3

Diffusion curve. The solid line is the average marginal effect and the shaded band is the 95% confidence interval computed by the delta method.

FIGURE 3

Diffusion curve. The solid line is the average marginal effect and the shaded band is the 95% confidence interval computed by the delta method.

Close modal

Dummy variables for time effects on the probability of passing a law became statistically significant in 2010 compared with the base of 2005, the first year of the dataset. Diffusion, the level of adoption with respect to time, showed the expected S-shaped curve (Figure 3, plotted using marginal effects). The results were consistent with the research by Rogers (2003), Griliches (1957), and Ryan and Gross (1943), who found the S-shaped curve in technology adoption, and showed that the S-shaped curve was also relevant for policy adoption.

It is notable that no further states have adopted P-free laws since 2013 (Table 2). A potential explanation is the switch in political power from Democrats to Republicans in state legislatures after 2011, since Republicans may be less likely to support environmental protection policies, as shown by many studies (Agnone, 2007; Perino and Talavera, 2014; Walsh et al., 2015). For example, 15 states switched from Democratic to Republican control of the state legislature between 2005 and 2013. Of these states, Ohio, Arizona, Oklahoma, and Virginia had higher TP levels than the average of the data set. Ohio has had major problems with algal blooms in Lake Erie, including one requiring a drinking water ban in Toledo in 2014. While Ohio still has not approved a P-free lawn fertilizer law, a bill to reduce nutrients from agriculture was passed in 2015 by the Republican legislature, which may reflect the fact that the fundamental source of P is agriculture rather than households (Ohio Environmental Protection Agency [Ohio EPA], 2010). As indicated earlier, there is also a bill in Pennsylvania to ban P in lawn fertilizers.

Therefore, another explanation for the lack of further P-free laws might be that many P-free lawn fertilizers became widely available in markets after Scotts Miracle-Gro announced the removal of P from their Turf Builder lawn fertilizers in May of 2013.18 Scotts’ action may be explained as preemptive self-regulation; they may have changed their formulation to preempt further P-free laws. Maxwell, Lyon, & Hackett (2000) see self-regulation as a strategic response to the threat of regulation, especially when the cost of self-regulation is low, which is the case for P-free fertilizers. Given that Scotts markets its fertilizers nationwide, they would want to limit variations in regulations across states. It is also notable that Scotts is based in Ohio, a state with major problems related to harmful algal blooms. Preemptive self-regulation and the cost of P may thus help explain the surprising results regarding the number of fertilizer companies.

Conclusions

Reducing P loading into freshwater resources is important for alleviating water pollution. Runoff from urban areas is one of the main sources of P pollution. Worldwide, half of the population lives within 3 km of a freshwater body (Kummu, de Moel, Ward, & Varis, 2011). In this P regulatory decision study, all hypothesized factors, public interest, private interest, ideology, and diffusion, were influential in the passage of P-free laws.

P-free lawn fertilizer regulations have been passed in 11 states. As the rate of states passing these regulations accelerated, one of the leading lawn fertilizer companies, Scotts, initiated removal of P from their Turf Builder lawn fertilizer products, and P-free lawn fertilizer products have subsequently become widely available from various companies. Since 2013, no more states have implemented P-free laws, although Pennsylvania is now considering this. Even though the companies’ decisions may be based on their own private interest, their actions may positively influence water quality even in states that have not implemented a P-free law by making it easier for consumers to purchase these fertilizers. As companies change their formulations for some states, P-free fertilizers may become the default for consumers in other states. In other words, P-free laws may be a win–win policy instrument because they not only improve water quality with little or no impact on established lawns, but also enable firms to reduce their costs. For employment in water-related industries, as well as the number of fertilizer companies, increases in those variables increased adoption, but this effect was quadratic; it diminished as group size increased. This may be due to the free rider problem, as groups become larger and more diffuse.

While it is possible that households may switch to composted manure, which is high in P, thus counteracting the goals of the bans, McCann and Shin (2018) point out that commercial P-free fertilizers are more convenient and that compost is not viewed by users as a fertilizer.

Our findings suggest some future research. A rigorous case study of Scotts could examine whether their actions were primarily motivated by preemptive self-regulation, input cost savings, or public relations. The behavior of lawn management companies in states that are not regulated could examine whether they are also reducing or eliminating P to save money. Consumers’ awareness and use of P-free fertilizers in those states would also be of interest. As Hobbie et al. (2017) pointed out, when P-free laws are in effect, other urban sources such as leaf litter and/or pet waste become relatively more important and will need to be examined. This is important because pet waste also is a source of disease-causing organisms. It is also the case that many municipalities have implemented bans, such as those in Florida (Hochmuth, Nell, Bryan Unruh, Trenholm, & Sartain, 2012), and an examination of adoption of bans by municipalities and counties would be an interesting extension of this research. They may either facilitate statewide bans or substitute for them. Bans could be analyzed in the context of behavioral economics, given the relatively low adoption rates and difficulty in changing consumer behavior, as well as the seemingly win–win nature of the technology.

For states that have not implemented bans on P in home fertilizers, educational campaigns could be a substitute. Hochmuth et al. (2012) cite studies showing that education regarding BMPs resulted in delisting of some water bodies in Florida, although the threat of a ban, which was seriously examined in 2007, may have been helpful. Training of landscaping company personnel may be effective. According to Eisenhauer (2010, as cited in Aveni, Berger, Champion, Felton, Goatley, Keeling, Law, & Schwartz, 2013), 70 percent of professional landscapers and turf managers were willing to reduce lawn fertilizer application rates after training.

Notes

1.

Hochmuth, Nell, Bryan Unruh, Trenholm, & Sartain (2012) caution that poor turf condition can increase both runoff and leaching of nutrients.

2.

Impacts of reduced fertilizer applications are a function of a number of factors, besides the quantity applied. Interested readers are encouraged to examine the review by Hochmuth et al. (2012).

3.

TP was the maximum value from a single monitoring point over the whole state to capture states that had a higher probability of experiencing eutrophication due to high phosphorus concentrations.

4.

The following states and years were missing values for the water quality variable: Delaware (2005, 2006, 2008, 2009, 2012, 2013), Hawaii (2005, 2006, 2008, 2009, 2010, 2013), New Hampshire (2008, 2009, 2010, 2011, 2013), Rhode Island (2006, 2007, 2008, 2009), and Utah (2008, 2009).

5.

We used perennial water data that included both inland and territorial water area (up to 13.8 mi); only inland water area would be better for this study, but only aggregated data were available for 2012 and 2013.

6.

We used 2017 due to limited information about previous product history. In some areas, local stores, such as those associated with farm supply companies, may sell locally produced fertilizers, but this is a small portion of sales (Bob Broz, personal communication).

7.

We also used a more general variable from Berry, Ringquist, Fording, & Hanson (1998) as a proxy of political ideology, the weighted index of political affiliation adjusted by voting records from Americans for Democratic Action and Americans for Constitutional Action, and then weighted by state legislature (50%: house 25%, senate 25%) and governor (50%), but that was not significant, and environmental voting record is a better variable for environmental policy.

8.

Legislation that was related to water, air, wildlife, climate change, etc.

9.

In an alternative model, we also tried GDP per capita.

10.

Heterogeneity can be a problem if there are considerable differences in the environment, economy, and political ideology. If these factors are state-specific, and similar or constant over time, it could cause autocorrelation in the error term. The random effects probit model can control for this.

11.

Kutner, Nachtsheim, Neter, & Li (2005) suggested that VIF values in excess of 10 be taken as an indication of multicollinearity.

12.

We tried to include an education variable, but it and median household income had very large VIF values (greater than 100) and there was a strong correlation between income and education (0.82).

13.

ρ is the proportion of the total variance contributed by the panel-level variance component, where ρ = σv2σv2 + 1.

14.

The data for the pooled model are not shown, but are available upon request.

15.

AIC is used to compare models, and a lower value of the AIC means a better fit with a given dataset.

16.

The quadratic relationship in the marginal effects column in Table 3 is as follows: P(yit=1|x,z)=φ(α+βx+γx2+zπ), where x has a quadratic relationship, z contains other variables, and φ(⋅) is the standard normal probability density function. Taking the derivative, P(yit = 1|xz)x = ϕ (αβx + γx2 + zπ)(βx + 2γx) in probit results (Table 3). Therefore, the marginal effect of the squared term also diminishes as x increases.

17.

As a robustness check, we ran models deleting the states with the largest number of fertilizer companies, California as well as Ohio and Texas, but the results for this variable and all others were unchanged.

References

Agnone, J. (2007). Amplifying public opinion: The policy impact of the U . S . environmental movement. Social Forces, 85(4), 1593–1620. https://www.jstor.org/stable/4495000. Accessed August 12, 2018.
Ando, A. W. (1999). Waiting to be protected under the Endangered Species Act: The political economy of regulatory delay. Journal of Law and Economics, 42(1), 29–60. https://doi.org/10.1086/467417.
Aveni, M., Berger, K., Champion, J., Felton, G., Goatley, M., Keeling, W., Law, N., & Schwartz, S. (2013). Recommendations of the expert panel to define removal rates for urban stormwater retrofit projects. ChesapeakeBay Program. https://www.chesapeakebay.net/channel_files/18983/attachment_a1–final_unm_expert_panel_report_12062012.pdf. Accessed September 20, 2018.
Becker, G. S. (1983). A theory of competition among pressure groups for political influence. Quarterly Journal of Economics, 125(3), 371–400. https://doi.org/10.2307/1886017.
Berry, W. D., Ringquist, E. J., Fording, R. C., & Hanson, R. L. (1998). Measuring citizen and government ideology in the American states, 1960–93. American Journal of Political Science, 42(1), 327–348. https://www.jstor.org/stable/2991759. Accessed November 1, 2016.
Bigelow, C. A., Tudor, W. T., & Nemitz, J. R. 2012. Turfgrass management: Facts about phosphorus and lawns. Purdue Extension. https://www.extension.purdue.edu/extmedia/AY/AY-334-W.pdf. Accessed August 4, 2018.
Brehm, J. M., Pasko, D. K., & Eisenhauer, B. W. 2013. Identifying key factors in homeowner’s adoption of water quality best management practices. Environmental Management, 52(1), 113–22. https://doi.org/10.1007/s00267-013-0056-2.
Chesapeake Bay Program. (2010). Bay barometer: A health and restoration assessment of the Chesapeake Bay and watershed in 2009. Annapolis, MD.
Cropper, M. L., Evans, W. N., Berardi, S. J., Ducla-Soares, M. M., & Portney, P. R. (1992). The determinants of pesticide regulation: A statistical analysis of EPA decision making. Journal of Political Economy, 100(1), 175–97. https://www.journals.uchicago.edu/doi/abs/10.1086/261812. Accessed June 14, 2017.
Daley, D. M., & Garand, J. C. (2005). Horizontal diffusion, vertical diffusion, and internal pressure in state environmental policymaking, 1989–1998. American Politics Research, 33(5), 615–44. https://doi.org/10.1177/1532673X04273416.
Davenport, T., & Drake, W. (2011). PA commentary: Grand Lake St. Marys, Ohio – The case for source water protection: Nutrients and algae blooms. Lakeline, 31(3), 41–46. doi:10.1016/j.jfca.2014.12.012. http://dx.doi.org/10.1016/j.jfca.2014.12.012.
Environmental Protection Agency. (2015). A compilation of cost data associated with the impacts and control of nutrient pollution. (May) 1–110. http://www2.epa.gov/sites/production/files/2015-04/documents/nutrient-economics-report-2015.pdf. Accessed October 21, 2016.
Fuglie, K. O., & Kascak, C. A. (2001). Adoption and diffusion of natural-resource conserving agricultural technology. Review of Agricultural Economics, 23(2), 386–403.
General Assembly of Pennsylvania. (2018). Senate bill No. 792.
Geroski, P. (2000). Models of technology diffusion. Research Policy, 29(4–5), 603–25. https://doi.org/10.1016/S0048-7333(99)00092-X.
Griliches, Z. (1957). Hybrid corn: An exploration in the economics of technological change. Econometrica, 25(4), 501–22. https://www.jstor.org/stable/1905380. Accessed October 13, 2017.
Grossback, L. J., Nicholson-Crotty, S., & Peterson, D. A. M. (2004). Ideology and learning in policy diffusion. American Politics Research, 32(5), 521–45. http://journals.sagepub.com/doi/10.1177/1532673X04263801. Accessed May 21, 2017.
Helland, E. (1998). The revealed preferences of state EPAs: Stringency, enforcement, and substitution. Journal of Environmental Economics and Management, 35(3), 242–61. https://www.jstor.org/stable/1905380. Accessed July 7, 2017.
Hobbie, S. E., Finlay, J. C., Janke, B. D., Nidzgorski, D. A., Millet, D. B., & Baker, L. A. (2017). Contrasting nitrogen and phosphorus budgets in urban watersheds and implications for managing urban water pollution. Proceedings of the National Academy of Sciences, 114(16), 4177–82. doi:10.1073/pnas.1618536114.
Hochmuth, G., Nell, T., Bryan Unruh, J., Trenholm, L., & Sartain, J. (2012). Potential unintended consequences associated with urban fertilizer bans in Florida – A scientific review. HortTechnology, 22(5), 600–616. http://horttech.ashspublications.org/content/22/5/600.full.pdf. Accessed August 3, 2018.
Jeppesen, E., SØndergaard, M., Jensen, J. P., Havens, K. E., Anneville, O., Carvalho, L., Coveney, M. F., Deneke, R., Dokulil, M. T., Foy, B., Gerdeaux, D., Hampton, S. E.,Hilt, S., Kangur, K., Köhler, J., Lammens, E. H. H. R., Lauridsen, T. L., Manca, M.,Miracle, M. R., Moss, B., Nõges, P., Persson, G., Phillips, G., Portielje, R., Romo, S., Schelske, C. L., Straile, D., Tatrai, I., Willén, E., & Winder, M. (2005). Lake responses to reduced nutrient loading – An analysis of contemporary long-term data from 35 case studies. Freshwater Biology, 50(10), 1747–71. https://doi.org/10.1111/j.1365-2427.2005.01415.x.
Joskow, P.L., & Noll, R. G. (1981). Regulation in theory and practice: An overview. In G. Fromm (Ed.), Studies in Public Regulation (pp. 1–78). Cambridge, MA: MIT Press. http://www.nber.org/chapters/c11429. Accessed October 20, 2016.
Joskow, P. L., & Schmalensee, R. (1998). The political economy of market-based environmental policy: The U.S. acid rain program. Journal of Law and Economics, 41, 37–83. https://doi.org/10.1086/467384.
Kalt, J. P., & Zupan, M. A. 1984. Capture and ideology in the economic theory of politics. American Economic Review, 74(3), 279–300. https://www.jstor.org/stable/1804008. Accessed May 20, 2017.
Kara, E. L., Heimerl, C., Killpack, T., van de Bogert, M. C., Yoshida, H., & Carpenter, S. R. (2012). Assessing a decade of phosphorus management in the Lake Mendota, Wisconsin watershed and scenarios for enhanced phosphorus management. Aquatic Sciences, 74(2), 241–53. https://doi.org/10.1007/s00027-011-0215-6.
Kashian, R., & Kasper, J. (2010). Tainter Lake and Lake Menomin – The impact of diminishing water quality on value. Department of Economics, University of Wisconsin. https://www.edpwi.com/lakes.pdf. Accessed July 17, 2017.
King, K. W., Balogh, J. C., Agrawal, S. G., Tritabaugh, C. J., & Ryan, J. A. (2012). Phosphorus concentration and loading reductions following changes in fertilizer application and formulation on managed turf. Journal of Environmental Monitoring: JEM, 14(11), 2929–38. http://www.ncbi.nlm.nih.gov/pubmed/23026887. Accessed December 4, 2016.
Kosnik, L. (2010). Balancing environmental protection and energy production in the Federal hydropower licensing process. Land Economics, 86(3), 444–66. http://doi.org/10.3368/le.86.3.444.
Krause, R. M. (2011). Policy innovation, intergovernmental relations, and the adoption of climate protection initiatives by U.S. cities. Journal of Urban Affairs, 33(1), 45–60. https://doi.org/10.1111/j.1467-9906.2010.00510.x.
Kummu, M., de Moel, H., Ward, P. J., & Varis, O. (2011). How close do we live to water? A global analysis of population distance to freshwater bodies. PLoS ONE, 6(6). https://doi.org/10.1371/journal.pone.0020578.
Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied linear statistical models, 5th ed. New York: McGraw–Hill/Irwin.
Lehman, J. T., Bell, D. W., Doubek, J. P., & McDonald, K. E. (2011). Reduced additions to river phosphorus for three years following implementation of a lawn fertilizer ordinance. Lake and Reservoir Management, 27(4), 390–97. http://www.informaworld.com/10.1080/07438140903117217. Accessed October 6, 2016.
Lester, J. P., Franke, J. L., Bowman, A. O., & Kramer, K. W. (1983). Hazardous wastes, politics, and public policy: A comparative state analysis. Western Political Quarterly, 1(3), 295–97. https://doi.org/10.1177/106591298303600207.
Levitt, B. S. D. (1996). How do senators vote? Disentangling the role of voter preferences, party affiliation, and senator ideology. American Economic Review, 86(3), 425–41. https://www.jstor.org/stable/2118205. Accessed July 6, 2017.
Lopez, C. B., Jewett, E. B., Dortch, Q., Walton, B. T., & Hudnell, H. K. (2008). Scientific assessment of freshwater harmful algal blooms. Interagency Working Group on Harmful Algal Blooms, Hypoxia, and Human Health of the Joint Subcommittee on Ocean Science and Technology. Washington, DC (December) 65.
Lyon, T. P., & Yin, H. (2010). Why do states adopt renewable portfolio standards? An empirical investigation. Energy Journal, 31(3), 133–58. https://www.jstor.org/stable/41323297. Accessed December 4, 2017.
Matisoff, D. C., & Edwards, J. (2014). Kindred spirits or intergovernmental competition? The innovation and diffusion of energy policies in the American states (1990–2008). Environmental Politics, 23(5), 795–817. http://dx.doi.org/10.1080/09644016.2014.923639.
Maxwell, J. W., Lyon, T. P., & Hackett, S. C. (2000). Self-regulation and social welfare: The political economy of corporate environmentalism. Journal of Law and Economics, 43(2), 583–618. http://www.journals.uchicago.edu/doi/10.1086/467466. Accessed December 22, 2017.
McCann, L., & Shin, D. W. (2018). Household use of composted manure and phosphorus-free fertilizers: Feeling good versus doing good. Journal of Environmental Protection, 9, 140–57. https://doi.org/10.4236/jep.2018.92011.
Miller, K. (2012). State laws banning phosphorus fertilizer use. 2012-R-0076 (pp. 1–4). Connecticut General Assembly, Office of Legislative Research, Hartford, CT.
Mistiaen, J. A., Strand, I. E., & Lipton, D. (2003). Effects of environmental stress on blue crab (Callinectes sapidus) harvests in Chesapeake Bay tributaries. Estuaries, 26(2A), 316–22. https://doi.org/10.1007/BF02695970
National Science and Technology Council (NSTC). (2016). Harmful algal blooms and hypoxia comprehensive research plan and action strategy: An interagency report (February).
Ohio Environmental Protection Agency. (2010). Ohio Lake Erie Phosphorus Task Force final report.
Olson, M. (1965). The logic of collective action: Public goods and the theory of groups. Cambridge, MA: Harvard University Press.
Pagliari, P. H., Kaiser, D. E., & Rosen, C. J. (2018). Understanding phosphorus in Minnesota soils. University of Minnesota Extension. https://extension.umn.edu/phosphorus-and-potassium/understanding-phosphorus-minnesota-soils#the-phosphorus-cycle-631361. Accessed September 20, 2018.
Panayotou, T. (2016). Economic growth and the environment. In N. Haenn, R. Wilk, and A. Harnish (Eds.), The environment in anthropology (pp. 140–48). New York: NYU Press. https://www.unece.org/fileadmin/DAM/ead/sem/sem2003/papers/panayotou.pdf. Accessed September 20, 2018.
Pecorino, P. (2015). Olson’s Logic of collective action at fifty. Public Choice, 162(3–4), 243–62. https://doi.org/10.1007/s11127-014-0186-y.
Peltzman, S. (1976). Toward a more general theory of regulation. Journal of Law and Economics, 19(2), 211–40. https://doi.org/10.1086/466865.
Perino, G., & Talavera, O. (2014). The benefits of spatially differentiated regulation: The response to acid rain by U.S. states prior to the acid rain program. American Journal of Agricultural Economics, 96(1), 108–23. https://doi.org/10.1093/ajae/aat084.
Poole, K. T., & Rosenthal, H. (1984). The polarization of American politics. Journal of Politics, 46(4), 1061–79. https://doi.org/10.2307/2131242.
Reitman, J. (2018). Pa. legislature mulls limits on phosphorus in fertilizer. Turfnet.com. https://www.turfnet.com/news.html/_/pa-legislature-mulls-limits-on-phosphorus-in-fertilizer-r1051?forcePrint=1. Accessed May 18, 2018.
Rode, J., & Weber, A. (2016). Does localized imitation drive technology adoption? A case study on rooftop photovoltaic systems in Germany. Journal of Environmental Economics and Management, 78, 38–48. http://dx.doi.org/10.1016/j.jeem.2016.02.001.
Rogers, E. M. (2003). Diffusion of innovations, 5th ed. New York: Free Press.
Ryan, B., & Gross, N. (1943). The diffusion of hybrid seed corn in two Iowa communities. Rural Sociology, 8(1), 15–54.
Sapat, A. (2004). Devolution and state innovation: The adoption of environmental policy innovations by administrative agencies. Public Administration Review, 64(2), 141–51. https://doi.org/10.1111/j.1540-6210.2004.00356.x.
Scavia, D., Allan, J. D., Arend, K. K., Bartell, S., Beletsky, D., Bosch, N. S., Brandt, S. B., Briland, R. D., Dalog˘lu, I., DePinto, J. V., Dolan, D. M., Evans, M. A., Farmer, T. M., Goto, D., Han, H., Höök, T. O., Knight, R., Ludsin, S .A., Mason, D., Michalak, A. M., Richards, R. P., Roberts, J. J., Rucinski, D. K., Rutherford, E., Schwab, D. J., Sesterhenn, T. M., Zhang, H., & Zhou, Y. (2014). Assessing and addressing the re-eutrophication of Lake Erie: Central basin hypoxia. Journal of Great Lakes Research, 40(2), 226–246. https://doi.org/10.1016/j.jglr.2014.02.004.
Schindler, D. W., Hecky, R. E., Findlay, D. L., Stainton, M. P., Parker, B. R., Paterson, M. J., Beaty, K. G., Lyng, M., & Kasian, S. E. M. (2008). Eutrophication of lakes cannot be controlled by reducing nitrogen input: Results of a 37-year whole-ecosystem experiment. Proceedings of the National Academy of Sciences, 105(32), 11,254–58. https://doi.org/10.1073/pnas.0805108105.
Stahlman, M., & McCann, L. M. J. (2012). Technology characteristics, choice architecture, and farmer knowledge: The case of phytase. Agriculture and Human Values, 29(3), 371–79. https://doi.org/10.1007/s10460-011-9346-6.
Stigler, G. J. (1971). The theory of economic regulation. Bell Journal of Economics and Management Science, 2(1), 3–21. https://www.jstor.org/stable/3003160. Accessed July 7, 2017.
Syers, J. K., Johnston, A. E., & Curtin, D. (2008). Efficiency of soil and fertilizer phosphorus use. Food and Agriculture Organization. http://www.fao.org/3/a-a1595e.pdf. Accessed August 8, 2017.
USDA-ERS. (2013). USDA ERS – Fertilizer use and price. http://www.ers.usda.gov/data-products/fertilizer-use-and-price.aspx#.UXALY8q53Lo. Accessed September 20, 2018.
Vlach, B., Barten, J., Johnson, J., & Zachay, M. (2010). Assessment of source reduction due to phosphorus-free fertilizers. University of Minnesota. http://stormwaterbook.safl.umn.edu/case-studies/case-study-9-assessment-source-reduction-due-phosphorus-free-fertilizers. Accessed July 19, 2017.
Walsh, P. J., Bird, S., & Heintzelman, M. D. (2015). Understanding local regulation of fracking: A spatial econometric approach. Agricultural and Resource Economics Review, 44(2), 138–63.
Zambrosi, F. C. B., Ribeiro, R. V., Marchiori, P. E. R., Cantarella, H., & Landell, M. G. A. (2015). Sugarcane performance under phosphorus deficiency: Physiological responses and genotypic variation. Plant Soil, (386), 273–83.
Zou, X., Binkley, D., & Doxtader, K. G. (1992). A new method for estimating gross phosphorus mineralization and immobilization rates in soils. Plant and Soil, 147(2), 243–50. https://doi.org/10.1007/BF00029076.