Chapter 6. An assessment of the effectiveness, efficiency and feasibility of nitrogen policy instruments
This chapter evaluates, generically, the pros and cons of different policy instruments for nitrogen management, and their combinations, with respect to the criteria of effectiveness, cost-efficiency and feasibility. It provides examples of evaluation of effectiveness, cost-efficiency and feasibility for a number of instruments implemented in Australia, France, Japan, Sweden and the United States.
6.1 Key findings
Drummond et al., 2016 assesses the advantages and disadvantages, in the generic sense, of the seven categories of instruments defined in Chapter 5. and their groupings. The evaluation leads to the following conclusions.1
First, instruments should be ‘impact’-based, and applied as close to the point of emission as possible to maximise effectiveness and cost-efficiency. However, this is often not technically or administratively feasible (on some mobile or non-point pollution sources, for example). Similarly, upstream application of a pricing instrument, for example, may not be feasible for some pollutants, such as nitrogen oxides (NOx) and nitrous oxide (N2O) from combustion (the magnitude of which is dependent on the combustion technology, rather than a function simply of process inputs).
Second, pricing and direct environmental regulation are often most effective in reducing pollution, when subject to credible monitoring and enforcement. Whilst pricing instruments are likely to be the most cost-efficient policy option to achieve a given level of abatement (and direct regulation often the least), they are relatively ‘blunt’ instruments, and may create pollution ‘hotspots’ across space, time and sectors (depending on the scope of the instrument). ‘Tax aversion’, or the relatively low public and political acceptability of pricing instruments (in particular, but also other ‘polluter pays’ instruments)2 often leads to the introduction of exemptions, discounts and other measures (e.g. permit grandfathering), to secure support, reducing cost-efficiency. Polluter pays instruments may also lead to the Pollution Haven Effect,3 although the evidence for the existence and strength of this phenomenon is mixed.
Third, ‘beneficiary pays’ instruments)4 (e.g. public financial support), along with voluntary and informational instruments are often more politically feasible than ‘polluter pays’ instruments (e.g. pricing mechanisms), due to the lack of the direct cost to firms and the circumvention of issues contributing to tax aversion - although they are likely to be less effective (and in the case of financial support, in particular, less cost-efficient).
Fourth, combining ‘polluter pays’ with ‘beneficiary pays’ instruments may be more politically feasible to introduce to achieve a given level of environmental effectiveness than either employed individually. A key example is through the use of a ‘feebate’ instrument.5 The risk of inducing a Pollution Haven Effect is also reduced against the use of an equivalent ‘polluter-pays’ instrument introduced in isolation.
Fifth, combining pricing or public financial support instruments with direct environmental regulation may be mutually beneficial for various reasons. The application of the direct regulation as a ‘secondary’ instrument to support a ‘primary’ pricing or public financial support instrument reduces vulnerability to market distortions (such as split-incentives and environmentally harmful subsidies), but also the potential for pollution hotspots. Targeted secondary direct regulation on a sub-set of participants to the primary pricing instrument reduces the administrative complexity required if attempting to tackle hotspots through complex compliance rules (e.g. differentiated tax rates, access rules, permit multipliers, etc.) within the primary instrument itself. The cost-efficiency (and potentially effectiveness) of such combination is thus increased against the use of the primary instrument alone. The application of the pricing or public financial support instruments as secondary instruments to a primary direct regulation may increase cost-efficiency for a given level of abatement against the use of a direct regulation alone. Public and political acceptability may be increased in both instances.
Sixth, voluntary and information instruments may require the lowest administrative capacity (and produce the least transaction costs), although transaction costs and political acceptability for the introduction of an effective voluntary negotiated agreement may be prohibitive. Voluntary instruments (in particular) may also experience the greatest risk of regulatory capture.6 However, the use of information instruments is likely to increase the effectiveness, cost-efficiency or feasibility (or a combination of these assessment criteria) when combined with all other instrument categories.
A summary of results is provided in Table 6.1. Each instrument category displays very different characteristics against each of the three assessment criteria, with further differences induced depending on: (i) the specific design of the instrument concerned; (ii) the type of pollution the instrument is applied to; (iii) the source of the pollution in question (e.g. stationary, mobile or non-point); and, (iv) the institutional context within which it operates. Trade-offs between each of the three criteria are therefore unavoidable in a second-best world, characterised by market failures, uncertainties and practical constraints.
As such, in order to maximise effectiveness and cost-efficiency within ‘feasible’ constraints, an instrument mix is often required (OECD, 2007). Table 6.2 summarises the key pros and cons of each instrument pairing against the three assessment criteria. It is assumed that each instrument in the combination presented is present at the same point in time (although not necessarily introduced at the same time).
6.2 Case studies of policy instruments
Such 'generic' assessment of each instrument category (or groupings thereof) against each of the three assessment criteria based on key design options may be broadly applicable to any type of pollutant, and applied to tackle the same pollutant, produced or released via the same pathway(s). In the case of nitrogen pollution, the analysis may be broadly applied to the release of nitrogen into the atmosphere from the combustion of fossil fuels, or into water bodies from the application of fertiliser, for example. Case studies of where such instruments or their combinations have been applied to reduce anthropogenic disturbance of the nitrogen cycle are presented below. However, discussion of the complex interaction between different pathways of nitrogen pollution, and appropriate policy mixes that may be required to most appropriately tackle this, is beyond the scope of this Chapter.
6.2.1 The Swedish refund emission payment for nitrogen oxides (NOx): a combination of environmentally related tax and public financial support (PFS)
Since the 1980s, soil acidification has been a significant political issue in Sweden, a country, due to its underlying geology, particularly susceptible to acid deposition via acid rain, damaging both aquatic and terrestrial ecosystems. In order to partially address this, in 1992 the Swedish government introduced a tax on measured NOx emissions from stationary large combustion plants producing at least 50 GWh of useful energy annually (impacting around 200 plants), with exemptions for activities for which costs would have been excessive, such as the cement and lime industry, mining, refineries, blast furnaces, the glass and insulation material industry, wood board production and the production of biofuels (OECD, 2013). The tax was part of a wider strategy to reduce Swedish NOx emissions by 30% between 1980 and 1995 (Höglund-Isaksson and Sterner, 2009). The tax base was later increased with a threshold for entry reduced twice; the tax is currently applicable to plants with an annual useful output of at least 25 GWh (around 400 plants) (OECD, 2013). The tax rate was initially set at SEK 40 (USD 5.5)/kg NOx emitted, an exceptionally high rate compared to other NOx pricing systems in the OECD area (e.g. 200 times the rate levied in France) (Ecotec, 2001). In 2008, the rate was increased to SEK 50 (USD 6.4)/kg NOx. All revenue from the tax is recycled back to participants7 in proportion to their production of useful energy, meaning that firms with low (high) NOx- intensity per unit of energy produced become net beneficiaries (losers) (OECD, 2013). As such, the instrument is often referred to in the literature as a ‘Refund Emission Payment’.
Instrument effectiveness. NOx emissions from obligated installations fell by over 50% between 1980 and 1997 (coupled with rapidly increasing energy output)8 (OECD, 2013), with the 1992 NOx tax likely responsible for at least two-thirds of this (with other local regulation also having some influence), indicating high effectiveness (Ecotec, 2001). However, more recent studies suggest that total emissions from obligated entities have decreased only marginally between 1992 and 2013, with total emissions from the net ‘losers’ decreasing by around 30%, largely offset by increased total emissions from net ‘winners’. However, emission intensity (NOx/KWh) have more than halved in the same timeframe, driven particularly by the net ‘winners’ (Naturvårdsverket, 2014).
Instrument cost-efficiency. An environmental tax on NOx applied downstream requires continuous emission monitoring, which is infeasible for mobile sources, and expensive for small installations (OECD, 2013), with the result that the instrument has only been applied to relatively large stationary combustion installations. Whilst this reduces overall efficiency, the market-based pricing approach provides static cost-efficiency across participating polluters with highly heterogeneous marginal abatement costs (OECD, 2013). However, basing the Refund Emission Payment on the proportion of energy produced is cost-inefficient from the perspective of NOx abatement, and instead implies a subsidy to energy generation, incentivising excessive production and investment in some facilities (e.g. waste incineration plants). Dynamic cost-efficiency is also relatively high, with the continued incentive to abate delivered by the initial taxation, and to some degree, the Refund Emission Payment. Innovation as a result of the tax has been rapid, with the production and diffusion of new technologies, alongside the adoption of existing best-practice technologies, found to be ‘very important’ in delivering the NOx abatement achieved (Höglund-Isaksson and Sterner, 2009; Sterner and Turnheim, 2008). The Refund Emission Payment may have inhibited further diffusion of new innovations, as an individual firm is incentivised to retain the innovation to reduce NOx emissions as far as possible relative to other regulated installations, in order to obtain a higher rebate level (Höglund-Isaksson and Sterner, 2009). However, recent analysis questions these findings, and argues that inhibited innovation and diffusion of technology may have occurred in the absence of a refund mechanism (Bonilla et al., 2014, SOU, 2017). In 2011, total revenues amounted to SEK 794 million (USD 101 million) (OECD, 2013).
Instrument feasibility. The tax had broad public and political support when proposed and introduced. The impact of NOx on the environment was visible and widely understood, and the tax was proposed by an all-party parliamentary commission that also included representatives from all relevant ministries and civil society. The high tax rate introduced was made possible by these factors, but also by the innovative recycling mechanism, along with the exemption of industries with potential competitiveness issues (OECD, 2013).
6.2.2 Selected instrument combinations of relevance to nitrogen pollution
Environmentally related taxes and charges and tradable permit systems (TPS) (combination cost-efficiency)
Both taxes and charges and TPS may lead to the concentration of pollutants geographically, sectorally and temporally (i.e. the creation of a pollution ‘hotspot’)9 (Drummond et al., 2015b). Whilst this does not matter for pollutants such a greenhouse gases (GHGs), for some pollutants, the specific conditions of their release may have substantial impact on their marginal damages. For example, a unit of NOx in the atmosphere will have substantially higher marginal costs (via impacts on human health) in a dense, highly populated urban area than in a remote, rural location. The application of an additional tax or charge to firms participating in a tradable permit system, and which are located in an air/watershed where the emission of a given pollutant creates marginal damages higher than for participants in other air/watersheds, may tackle this issue.
Environmentally related taxes and charges and direct environmental regulation (DER) (combination effectiveness)
A substantial share of environmentally related taxes and charges are used in combination with at least one regulatory instrument (OECD, 2006). DER is often able to overcome certain market failures, such as principal-agent problems, information and other market failures (including environmentally harmful subsidies) that inhibit the effectiveness of environmentally related taxes and charges (and other market-based mechanisms). Conversely, such taxes and charges are able to reduce the ‘rebound’ effect10 that direct regulation may induce. Such interactions generally hold true whether each instrument directly targets the same group, or whether the instruments are applied to different actors but aim to tackle the same pollution source (e.g. upstream pricing of fossil fuels with downstream energy efficiency requirements). As such, if well designed, implementing such instruments in combination may have positive, mutually re-enforcing effects on environmental effectiveness, cost-efficiency and feasibility.
Direct regulation combined with a pricing instrument may also reduce the occurrence of pollution hotspots. For example, whilst pricing instruments influence total pollution, direct regulation may influence the specific characteristics (e.g. location and timing) of its release. For example, an ‘impact-based’11 regulation may set localised ambient quality standards, whilst an ‘outcome-based’12 approach may ban or require the use of certain technologies or specified activities, to prevent or reduce location- or time-specific environmental harm that a broad pricing instrument would not have incentivised against.13 Such instrument combinations have been widely and successfully deployed for this purpose (OECD, 2006; Bennear and Stavins, 2007). However, such a combination may potentially induce a Pollution Haven Effect without sufficient compensatory measures.
Environmentally related taxes and charges and public financial support (PFS) (combination effectiveness)
From a theoretical standpoint, the marginal incentives for pollution abatement provided by a tax or charge instrument (embodying the ‘polluter pays’ principle) and a PFS instrument (embodying the ‘beneficiary pays’ principle) are identical, if the tax/charge and support rates are set at the same level, and the externality is targeted directly (i.e. impact-based) (Baumol and Oates, 1988). However, even a well-designed PFS instrument would likely be less environmentally effective than an equivalent tax or charge. This is because whilst a ‘polluter pays’ instrument such as a tax acts to increase average total costs to the firm, PFS instruments act to decrease them. This allows firms (e.g. older, less efficient and more polluting firms) to maintain operation that may have otherwise been forced out of the market by a taxation instrument. Additionally, PFS does not allow the market to communicate the true cost of the product to the consumer, maintaining sub-optimal demand (too high, in the case of a negative externality) (Kolstad, 2011). Such issues maintain pollution levels above that expected from an equivalent taxation instrument.
As such, it would appear nonsensical to deploy both instruments to reduce a given pollutant from a given source, when a single instrument (particularly a tax or charge) could achieve the same objective. However, combining such instruments in a well-designed approach may increase the likelihood of achieving the desired abatement that feasibility constraints prevent a single instrument from achieving.
A common approach is through the use of ‘feebate’ instruments (or ‘bonus-malus’), which levy a tax or charge (‘fee’) on activities or products below a given environmental threshold performance or limit, and financial support to activities or products that exceed this performance level (‘rebate’). Figure 6.1 illustrates the design of a feebate instrument. In this example, the threshold (or ‘pivot point’) above which a ‘fee’ is due and below which a ‘rebate’ is given is 30 units of pollution. The Y-axis illustrates absolute values, meaning the gradient of the line indicates the marginal abatement costs and benefits (i.e. the rate of the tax and rebate due). Also illustrated in Figure 6.1 is the design of a similar tax/charge-only approach. As is clear, the gradient of the two slopes is the same, indicating equivalent marginal incentives, with the reasoning behind the differences in average total costs clearly demonstrated.
Source: Drummond et al. (2015b).
Whilst feebates may be applied to a wide range of pollutants from different sources (using a variety of specific designs), they are most commonly investigated for and applied to passenger car registration duties (against carbon dioxide (CO2) intensities). In this function (and more broadly), the evidence suggests well-designed examples are generally successful in improving environmental performance (Greene et al., 2005; Johnson, 2006; de Haan et al., 2009). However, this depends on the specific design of the instrument, including the pivot point and the levels of fees and rebates set (D’Haultfoeuille et al., 2014).14
Environmentally related taxes and charges and payments for ecosystem services (PES) (combination effectiveness)
As a PES instrument may be considered a specific category of PFS instrument, combining environmental taxes or a charge with a PES instrument produces dynamics similar to those produced by a broader tax/charge and PFS combination, discussed in the section above. This applies both to ‘government-funded’15 and ‘user funded’16 PES approaches (Drummond et al., 2015b).
However, some key specifics may be highlighted. Although relatively few rigorous ex post analyses of PES instruments have been undertaken (Drummond et al., 2015a), those that are available indicate a generally positive, but limited effectiveness (Pattanayak et al., 2010). Despite this, combining a tax or charge (or tradable permit system) with a PES instrument is likely to be less environmentally effective than a combination with other PFS approaches (particularly impact-based), both in achieving a targeted level of abatement, and in tackling pollution hotspots, in which, as with a tax/charge and PFS combination, the tax/charge is primary, with the PFS instrument secondary.
The key reason for this is the difficulty inherent in defining and directly targeting the ecosystem service in question. This often requires a detailed understanding of causal pathways (recognising spatial extent and distribution) that are yet either not fully understood, or impractical to monitor (Tomich et al., 2004). As such, proxies or indicators are commonly used (i.e. an ‘outcome-based’ approach). This links to the second issue; the difficulty in accurately defining the baseline to ensure ‘additionality’ of supported actions, and providing a suitable level of payment to encourage uptake (e.g. set at least at the differential between the private economic value gained from the presence of the ecosystem service (e.g. forest conservation), and the most profitable use of land to the owner (e.g. conversion to pasture). In other words, the opportunity cost of maintaining the ecosystem service in question (Drummond et al., 2015a)). As would be expected, preventing excessive support is equally as challenging, with consequences for cost-efficiency.
Environmentally related taxes and charges (combination cost-efficiency)
Whilst this Chapter discusses combinations of instrument between different instrument categories, it is often possible to effectively combine instruments from within the same category, but of different design. For example, in many OECD countries wastewater charges are applied to households, with a pollution tax levied further downstream on releases to water bodies from wastewater treatment plants (WWTPs). In a ‘first-best’ world, a cost-efficient approach would be to apply an impact-based tax at the point of wastewater release from the WWTPs, with the cost of the tax or abatement measures installed (with marginal abatement costs at or lower than the tax level) passed upstream to users (e.g. households) based on their individual pollutant contribution (plus other associated operating costs). These users would then also hold an incentive to reduce the level of wastewater (and concentration of pollutants in the wastewater) to the economically optimal level. However, determining the individual pollution contribution of users would in most cases by administratively and cost prohibitive, preventing the effective transmission of this price signal. Instead, a charge is levied on users often based on wastewater volume rather than pollutant load (and actually often based on metered water use rather than wastewater discharge, with the assumption of parity), with the objective of reducing overall wastewater discharge. An impact based-pollution tax, where present, then incentivises the WWTP to treat the remaining discharge to the economically optimal level (based on the level of the tax), before release into a water body.
Direct environmental regulation (DER) and public financial support (PFS) (combination cost-efficiency)
In January 2018 a political agreement was made concerning a new and more ‘targeted regulation’ of the agricultural emissions of nitrogen in Denmark waters.17 One key element of the new DER is to specifically target efforts according to geographic differences. This contrasts with the previous DER scheme, where all farmers where required to reduce their emissions by the same amount, regardless of geographical variation in the nitrogen balance. The new DER involves dividing up the country into approximately 3 000 distinct areas and setting up a nitrogen emissions reduction target for each one in order to protect groundwater and surface water. Under the regulation, farmers are free to choose the most cost-effective tool for their locality.
The new DER is combined with PFS for agricultural production that reduces the emissions of nitrogen (e.g. catch crops, which are sown after harvesting in order to take up soil nitrogen). Despite a more targeted approach, such subsidy based regulation does not create a long-term incentive to relocate the production with the highest emissions to areas in Denmark where the emissions will damage the aquatic environment the least (Danish Economic Councils, 2018). This disadvantage could be avoided if the DER instead incurred geographically differentiated taxes on those agricultural activities which emit nitrogen to the aquatic environment (ibid).
Such a tax on agricultural activities that release nitrogen into the aquatic environment could relatively easily be extended to other types of agricultural emissions, such as ammonia (NH3) or N2O (Danish Economic Councils, 2018). Nitrogen taxes, as opposed to taxes on GHG emissions, will need to vary geographically; this should not increase the complexity of the regulation from the farmer's perspective since there would still be only one tax for each type of activity (ibid). In the long run such extended taxation would provide a clearer and more direct common regulation of different environmental effects from agricultural production (ibid).
6.2.3 The Greater Miami Watershed Trading Programme: an example of tradable permit system (TPS)
Around 40% of surface waters in the Great Miami River watershed of around 10 000 km2 in Southwest Ohio, United States, were consistently failing to meet regulatory water quality standards. In response, the regional water management agency, the Miami Conservancy District (MCD), in 2004 introduced the Greater Miami Watershed Trading Programme (GMWTP) as a pilot scheme for a cost-effective approach to improving water quality, in anticipation of regulatory measures to set nutrient release limits on point-source WWTPs.
The GMWTP takes a baseline-and-credit approach, and encourages farmers to adopt voluntary ‘best-practice’ measures to generate Emission Reduction Credits (ERCs), that may be traded to WWTPs to allow for future regulatory compliance. Agricultural activities cover around 70% of land in the watershed, are the primary contributor to the excess nutrient release (Newburn and Woodward, 2012), and according to Kieser and Associates (2004), are able to achieve nutrient reductions at a cost around thirty times lower than WWTPs. Any agricultural operation within the watershed is eligible to implement measures to generate credits. A farmer, in co-operation with the local Soil and Water Conservation District (SWCD) authority, proposes a bid suggesting practice changes, the value requested to implement the changes, and calculates the number of credits these actions generate (1 ERC = 1 lb (0.45 kg) avoided nitrogen or phosphorus release), based on annual savings against a projected baseline of nutrient release, multiplied by the number of years the proposed actions are expected to generate savings. Phosphorus and nitrogen credits are fully fungible. Credits are then purchased from farmers (via SWCDs) by the MCD (which acts as a clearinghouse) by ‘reverse’ auction,18 which may then be bought by WWTPs (although only credits generated by activities upstream from the purchasing WWTP may be acquired) (Newburn and Woodward, 2012; Shortle, 2012).
Instrument effectiveness. By mid-2014, 397 agricultural projects were contracted, generating over 1.14 million credits worth over USD 1.6 million, and producing an estimated 572 metric ton reduction in nutrient discharges to surface waters in the watershed against the counterfactual (MCD, 2014). As binding legislation on point sources has yet to be introduced, these transactions are driven by trading ratios that favour action in advance of expected legislative requirements.19
Instrument cost-efficiency. Trades completed have thus far produced a cost of USD 1.48/lb of nutrient abated. This is significantly less than the estimated USD 4.72/lb abated estimated for measures implemented directly by WWTPs (Kieser and Associates, 2004), suggesting significantly higher cost-efficiency than a regulatory approach applied to point sources only. However, whilst the use of a reverse auction should induce competition between suppliers and reveal reserve prices, it is likely that over successive auction rounds strategic bidding came into play, with SWCDs and farmers learning how to achieve the highest price whilst still having their bid accepted. This was likely facilitated by the lack of variation in the maximum bid price set by the MCD over successive rounds (USD 2/lb), reducing static cost-efficiency. The incentive for innovation has been low, as the credit calculation methodology relies on data only available for relatively common nutrient management practices (Newburn and Woodward, 2012).
Instrument feasibility. The cost-effective nature of the GMWTP compared to other regulatory options for achieving anticipated regulatory nutrient release limits makes the GMWTP popular amongst WWTPs. During formulation of the GMWTP, over a hundred meetings took place with a wide range of stakeholders, leading to wide acceptance and support for the instrument. The use of the SWCDs as agents also actively helped advertise and encourage the participation of individual farmers (alongside subsequent annual monitoring, the administrative cost of which is factored in to bids). This reduces both transaction costs and administrative burden. These factors combine to make the GMWTP a popular and well-functioning instrument (Newburn and Woodward, 2012).
6.2.4 Japan’s automobile ‘nitrogen oxides (NOx) law’: an example of direct environmental regulation (DER)
Despite various regulations placed on stationary sources of local air pollutants in Japan, levels of NOx pollution continued to rise throughout the 1980s, as a result of increasing vehicle traffic. In response, the 1992 ‘Law Concerning Special Measures for Total Emission Reduction of Nitrogen Oxides from Automobiles in Specified Areas’ (NOx Law) was enacted, which required the prefectures of the large urban areas of Tokyo and Osaka to establish and implement local plans to reduce NOx emissions from vehicles, with the overarching aim of achieving national ambient nitrogen dioxide (NO2) concentration standards by 2000 in these prefectures (a 27% decrease from 1990 levels). The law includes, in particular, ‘vehicle type regulation’, which prohibits both the use and registration of vehicles that fail to meet specified emission standards in the obligated prefectures (with specific ‘retirement’ dates differentiated by vehicle type). However, a 10% increase in vehicle mileage coupled with underperformance of other management measures (including preferential taxation, low-interest loans and investment in alternative fuel stations to promote the use of low-emission vehicles) produced a reduction in NOx emissions of just 3% (OECD, 2002). In 2001, as a countermeasure, particulate matter was added as a target substance, and Nagoya was added as an additional obligated prefecture. These amendments were codified as a revised ‘Law Concerning Special Measures for Total Emission Reduction of NOx and particulate matter (PM) from Automobiles in Specified Areas’ (NOx and PM Law).
Instrument effectiveness. NOx emissions decreased around 20% between 2000 and 2009 in the areas subject to the revised law, with reductions experienced at twice the rate experienced outside the obligated prefectures (Hasunuma et al., 2014). Iwata et al. (2014) finds that the vehicle type regulation, by banning the registration of high-polluting vehicles, and forcing the replacement of those already in circulation, was the key driver behind this result. The prevalence of asthma in obligated areas decreased significantly at -0.073%/year, with the induced NOx (and PM) decrease significantly correlated to this trend (Hasunuma et al., 2014).
Instrument cost-efficiency. Whilst there are no ex post cost estimates of the ‘NOx Law’, ex ante estimates suggest a total cost of up to JPY 521 billion (around USD 5 billion) (Iwata and Arimura, 2008; Iwata et al., 2014). The total social benefits of the regulation, however (including those of reduced health impacts, and spillover impacts to neighbouring, non-obligated regions), are almost certainly likely to have exceeded this cost (by at least double, according to lower-bound estimates by Arimura and Iwata, 2006). However, there is significant variation between marginal abatement costs among vehicle type (with standard passenger cars experiencing a marginal abatement cost more than double that of a small bus, for example), suggesting a significant potential for both reduced compliance cost and increased net social benefit by changing the schedule of retirement enforcement (Arimura and Iwata, 2006).
Instrument feasibility. Rapid economic growth in Japan from the mid-20th Century produced significant growth in air pollution, and consequential health impacts on the population (particularly in large urban areas). Public pressure, along with an increasing number of successful lawsuits filed by organisations representing the interests of victims of pollution-related health damage, led to a succession of anti-pollution laws from the 1960s onwards. The awareness of pollution-related health issues in the population is likely to have provided support for the NOx law, along with differentiated retirement dates and the presence of parallel instruments (such as graded vehicle taxation rates and subsidies for low-emission vehicles), although the literature is lacking on the extent to which public support was present at its introduction, and the contribution of these factors. The instrument is enforced through the Japanese Vehicle Inspection Programme; non-compliant vehicles cannot undergo this mandatory inspection if they are not in compliance with the NOx law (Colls and Tiwary, 2010).
6.2.5 Pennsylvania’s Resource Enhancement and Protection Programme: an example of public financial support (PFS)
Established in 2008, the Resource Enhancement and Protection Programme (REAP) provides tax credits to farmers, business and individuals in return for implementing pre-approved ‘Best Management Practices’ (BMPs) in agricultural operations in Pennsylvania that enhance farm production and protect natural resources (exceeding legal requirements). Applicants may receive between 50% and 75% of private implementation costs as Pennsylvania state tax credits (up to USD 150 000, depending on the BMP implemented, with publicly-funded costs ineligible), issued upon project completion. Non-agricultural businesses and individuals (subject to taxation by the Commonwealth of Pennsylvania)20 are able to fund BMP measures and receive the associated tax credits, which are valid for fifteen years, and available for transfer to other tax payers. Credits are repayable if the practice is not continued for the pre-agreed lifespan.
Instrument effectiveness. Between 2008 and the end of 2011, over 950 agricultural operations had received tax credits for introducing at least one BMP measure, reducing estimated emissions of nitrogen to water bodies by over 5 700 metric tons (along with a reduction of phosphorus and sediment by around 430 tonnes each) (PSCC, 2011).
Instrument cost-efficiency. Over this period, REAP participants claimed tax credits worth around USD 39 million (PSCC, 2011), allowing them to leverage USD 5-8 million in private investment per year. Such figures would suggest the initiative has been highly cost effective in inducing pollution abatement in agricultural operations in Pennsylvania. However, as tax credits are issued based on the introduction of standard, pre-approved practices and products (an outcome-based approach), the impact on pollution will vary from site to site and over time, reducing static cost-efficiency. Additionally, it is not clear to what extent such investments were previously prevented by misaligned incentives, or by issues such as information failures, with further consequences for cost-efficiency. The eligibility of only pre-approved BMPs also reduces the incentive to innovate, preventing dynamic cost-efficiency.
Instrument feasibility. By allowing any tax paying entity to fund (or ‘sponsor’) improvements, potential issues of access to capital are tackled. An annual cap on the number and value of credits available is set each year, allowing the ability to learn from previous years.
6.2.6 The Agriculture and Environment Programme for Vittel area: an example of payment for ecosystem services (PES)
The bottled water brand ‘Vittel’ sources and bottles its water from an aquifer in the village of the same name, in northeast France. To maintain its status as ‘natural mineral water’ the nitrate concentration must remain stable, not exceed 15mg/l, not contain pesticides, and not be treated at any stage in the process. In the early 1980s, the intensification of agriculture in the local catchment threatened to violate these conditions, and by extension, the continuation of the Vittel operation. Various options were considered to mitigate this threat, including purchasing the agricultural land causing the problem and using legal action to force the agricultural operations to alter their practices. However, legal and practical issues rendered such options infeasible. As such, in 1989, a system to incentivise farmers to voluntarily alter their practices was decided upon. Extensive research was undertaken to determine the relationship between farming practices and the nitrate rate in the aquifer, the practices available to minimise excessive nitrogen release, and the incentives necessary to encourage farmers to adopt these practices (beyond legal requirements) as a proxy for the provision of water with low nitrate concentrations (an outcome-based approach). Based on this, an incentive package was created. Farmers must agree to eliminate corn crops, ban pesticides, limit the use of artificial fertiliser, compost all animal waste, limit livestock intensity and ensure high standards for buildings (Depres et al., 2005). In exchange, Vittel would provide up to EUR 150 000 per farm to cover the cost of new equipment and building modernisation, cover the cost of labour for the application of compost in fields, provide free technical advice and assistance, and an average support of EUR 200/ha/year for five years to ensure income during practice transition (Perrot-Maitre, 2006).
Instrument effectiveness. The Agriculture and Environment Programme for Vittel area (AGREV) has been widely recognised as a success. By 2004, all farms in the area had joined the scheme, protecting 92% of the at-risk area, with the nitrogen levels in the aquifer remaining stable. The causal links between existing agricultural practices and nitrogen pollution in the local catchment were well researched and understood, in order to determine appropriate measures to promote. However, it cannot be known to what extent the instrument has reduced nitrogen concentrations in the aquifer below those that would have resulted in the absence of the scheme. Payments are not conditional upon reduced nitrogen pollution (impact), but on the purchase and implementation of given technologies and practices (outcome), reducing both static and dynamic cost-efficiency (as the impact on nitrogen release will likely differ according to variables including location and time).
Instrument cost-efficiency. It is estimated that Vittel spent EUR 24.25 million over the first seven years of the AGREV programme - a rate of EUR 980/ha, and EUR 1.52 per m3 of bottled water produced. The cost effectiveness of the scheme in the long run is shown by the continued profitability of the Vittel brand (Perrot-Maitre, 2006).
Instrument feasibility. A critical component of success of the scheme was the effort made to understand the choices faced and made by each operation within the catchment individually (including non-cost issues such as inheritance laws), and long-term, open dialogue to forge trust and mutual understanding. The final package received by farmers was negotiated on an individual basis to address their particular needs and concerns. Although this increased feasibility, it likely increased transaction costs significantly (Perrot-Maitre, 2006).
6.2.7 Australia’s ‘FERTCARE’: an example of information measure
Agriculture is central to many environmental debates in Australia (Drew, 2007), with the extensive use of manufactured fertiliser in Australia of significant concern, particularly surrounding consequential eutrophication in inland and coastal waterways (of which a decline in water quality of the latter threatens the health of the Great Barrier Reef to the northeast of the country). In 2004 the FERTCARE programme (the ‘Programme’, hereafter), a joint initiative between the Australian Fertiliser Services Association (AFSA), Fertiliser Australia, and the Departments of Environment and Heritage (DEH) and Agriculture, Fisheries and Forestry (DAFF), was launched as the centrepiece instrument for tackling these issues. Farmer advisory services, sellers of fertiliser spreading machines and fertiliser manufacturers seeking to minimise the environmental damage from manufactured fertiliser application can undertake FERTCARE training. Advisors can take a step beyond FERTCARE training and be recognised as FERTCARE Accredited Advisors (FAA). FAA advice on soil and fertiliser management should be based on high quality soil and/or plant testing methodology and on laboratories applying good practice and accepted science in Australia. FAA advisors are subject to a biennial audit. Sellers of fertiliser spreaders can apply for FERTCARE "Accu-Spread" certification. To display the FERTCARE "organisation" logo, fertiliser manufacturers must comply with the objectives of the fertiliser industry in terms of staff training. FERTCARE has been constantly updated and won a Business and Higher Education Round Table (B-HERT) award in 2012.
Instrument effectiveness. The Programme aims to train all eligible people. As of 2016, twelve years after the start of the programme, only 76% of the eligible staff in respondent companies had successfully completed FERTCARE training (Fertiliser Australia, 2016). Whilst this might suggest the Programme has been ineffective in achieving its objectives, it is likely that staff turnover and failure for some persons to successfully complete the training mean that 100% training is infeasible (ibid). Eligible staff turnover continues to be an issue facing fertiliser businesses. Additionally, in 2016, only 74 spreading machines were Accu-Spread certified, which is probably low in percentage although there is no reliable data on the number of spreading machines in Australia. In 2016, Australia had 256 FAA advisors and 11 FERTCARE "organisation" businesses. The ‘impact’ (reduced nitrogen pollution) as a result of the Programme has not been evaluated and is not directly monitored; meaning the environmental effectiveness of the instrument cannot be directly assessed. However, the ‘outcome’ (induced practice changes on farms and other sites) may be used as a proxy to some extent. Cummins (2016) find in its evaluation of the FERTCARE Carbon Farming Extension Project (FCFEP), which ran from 2013 to 2016, that FAA advisors were giving better advice to farmers based on their improved knowledge of emissions reductions, carbon storage and government policy. Overall, 59% of survey participants indicated that FERTCARE nitrogen use efficiency training materials were very relevant to them as advisors, and that figure rises to 69% for those who have completed the training compared to 43% of those who have not. Therefore, whilst the Programme appears to be relatively successful in achieving its direct objectives of widespread training and the provision of high-quality advice to farmers, the overall environmental impact is unclear.
Instrument cost-efficiency. In 2005, fertiliser sales in Australia totalled nearly AUD 2.5 billion (USD 1.9 billion). Drew (2007) estimated the cost to the fertiliser industry of achieving the targets discussed above is likely to be around AUD 4 million (USD 3 million). As such, the Programme represents a relatively low, but still significant cost to the industry.
Instrument feasibility. The potential threat of future regulatory action may have been a factor in achieving acceptance among the fertiliser industry. However Drew (2007) identifies other factors contributing to its feasibility, including:
-
Clear statements of the issues of fertiliser overuse by reputable parties
-
Identifications of positive implications for the industry from the Programme
-
Funding support from the Australian Government for initial establishment
-
The commitment to 100% compliance encouraged all participants to invest in the scheme
-
The ‘all encompassing’ approach to training to suit all levels of role complexity
-
The involvement of other stakeholders, particularly from the public sector
-
The use of an external qualifications framework to manage accreditation and record keeping.
6.2.8 Chesapeake 2000 Programme: an example of voluntary scheme
The Chesapeake Bay is a significant source of economic activity (including fishing, tourism and shipping) for the Mid-Atlantic United States, the development of which over the past several decades has caused degradation (eutrophication) of the Bay’s waters, impairing further economic growth. In response, in 2000 the states of Maryland, Virginia, Pennsylvania, the District of Colombia and the federal government, building on an existing partnership, created the Chesapeake 2000 Programme – a voluntary agreement containing over 100 overarching goals, including removal of the Bay from the US Environmental Protection Agency (USEPA)’s ‘impaired waters’ list, and more specific actions, such as protecting and restoring 114 000 acres (46 000 hectares) of submerged aquatic vegetation (vegetation that grows to the surface of, but does not emerge from, shallow water), support the restoration of key rivers and tributaries, and developing and promoting wastewater treatment options to be (mostly) achieved by 2010 (Cramer, 2014; USEPA and USDA, 2006).
Instrument effectiveness. Cramer (2014) finds that whilst nitrogen loads in the Bay were around 13% lower between 2000 and 2010 than between 1990 and 2000, statistical analysis determines that the trend is not significant (a result that also stands for phosphorus loads). However, when controlled for hurricane and tropical storm events, the programme does appear to have been significant in reducing nutrient loads, with a total reduction of 40 million tonnes of nitrogen between 1990 and 2010 (and 1 million tonnes of phosphorus). However, Chesapeake Bay remained on the list of ‘impaired waters’, and by the end of 2010, became subject to TMDL regulations by the USEPA (see case study on the Chesapeake Bay Watershed in Chapter 3. ).
Instrument cost-efficiency. Ex ante estimates suggested a total projected cost of USD 18.7 billion to achieve all the goals of Chesapeake 2000, compared to an 1989 estimate of the value of the Bay at USD 678 billion (in 1987 dollars) (CBC, 2003). The value of committed resources by the termination of the Programme in 2010 is not clear, although even if it were, given the multitude of overlapping objectives, the specific cost-efficiency of achieved nitrogen pollution abatement through the Programme would be difficult to disaggregate and determine.
Instrument feasibility. Although the Programme was backed up by a credible threat of regulation (evidenced by the eventual imposition of TMDL regulations), other factors are likely to have combined to prevent more significant reduction in nutrient loads. For example, whilst some goals were highly specific in their objective, around half were unquantifiable, producing ambiguity. Additionally, provisions to prevent free-riding by each of the signatories were also lacking, along with positive incentives (e.g. visible participant benefits) to achieve the objectives of the agreement (Cramer, 2014).
References
Arimura, T.H. and K. Iwata (2006), “Inefficiency in Command-and –Control Approach Toward Adoption of Low Emission Vehicles: Japanese Experience of Air Pollution from Diesel Trucks”, Paper presented at 3rd World Congress of Environmental and Resource Economists, Kyoto, Japan.
Baumol, W.J. and W.E. Oates (1988), The Theory of Environmental Policy, 2nd Ed., Cambridge University Press.
Bennear, L.S. and R.N. Stavins (2007) “Second-Best Theory and the Use of Multiple Policy Instruments”, Environmental Resource Economics, 37.
Berkhout, P.H.G. et al. (2000), “Defining the Rebound Effect”, Energy Policy, 28(6-7).
Bonilla, J. et al. (2014), “Diffusion of NOx Abatement Technologies in Sweden”, Working Papers in Economics N°585, Department of Economics, University of Gothenburg.
Chesapeake Bay Commission (CBC) (2003), The Cost of a Clean Bay: Assessing Funding Needs Throughout the Watershed, Chesapeake Bay Commission, Annapolis, United States.
Colls, J. and A. Tiwary (2010), Air Pollution: Measurement, Modeling and Mitigation, 3rd Ed., Oxford, Routledge.
Cramer, S. (2014), “An Examination of Levels of Phosphorus and Nitrogen in the Chesapeake Bay Before and After the Implementation of the Chesapeake 2000 Program”, The Public Purpose, 12.
Cummins, T. (2016), “FERTCARE Carbon Farming Extension Project Second Survey – Advisor Knowledge and Attitudes”, Tim Cummins and Associates.
Danish Economic Councils (2018), “Economy and Environment, 2018”, Summary and Recommendations, De Økonomiske Råd.
D’Haultfoeuille, X. et al. (2014), “The Environmental Effect of Green Taxation: The Case of the French Bonus/Malus”, The Economic Journal, 124(578).
De Haan, P. et al. (2009), “How Much Do Incentives Affect Car Purchase? Agent-Based Microsimulation of Consumer Choice of New Cars – Part II: Forecasting Effects of Feebates Based on Energy-Efficiency”, Energy Policy, 37(3).
Depres, C. et al. (2005), On Coasean Bargaining with Transaction Costs: The Case of Vittel, CESAER Working Paper 2005/3.
Drew, N. (2007), “Fertcare – Putting Best Practice into Stewardship”, in: Fertiliser Best Management Practices: General Principles, Strategy for their Adoption and Voluntary Initiatives vs Regulations, International Fertiliser Industry Association, Paris.
Drummond, P. et al. (2016), “Policy Instruments and Combinations to Manage the Unwanted Release of Nitrogen into Ecosystems – Effectiveness, Efficiency and Feasibility”, paper presented to the Working Party on Biodiversity, Water and Ecosystems at its meeting on 11-12 May 2016, ENV/EPOC/WPBWE(2016)5.
Drummond, P. et al. (2015a), “Policy Instruments to Manage the Unwanted Release of Nitrogen into Ecosystems – Effectiveness, Cost-Efficiency and Feasibility”, paper presented to the Working Party on Biodiversity, Water and Ecosystems at its meeting on 19-20 February 2015, ENV/EPOC/WPBWE(2015)8.
Drummond, P. et al. (2015b), “Policy Instrument Combinations to Manage the Unwanted Release of Nitrogen into Ecosystems – Effectiveness, Cost-Efficiency and Feasibility”, paper presented to the Working Party on Biodiversity, Water and Ecosystems at its meeting on 21-22 October 2015, ENV/EPOC/WPBWE(2015)14.
Ecotec (2001), “Study on the Economic and Environmental Implications of the Use of Environmental Taxes and Charges in the European Union and its Member States”, Ecotec, Brussels.
Fertiliser Australia (2016), “Sustainability and Stewardship 2016”, www.fertilizer.org.au/Portals/0/Documents/Reports/Sustainability%20and%20Stewardship%202016.pdf?ver=2018-01-22-151251-473.
Greene, D.L. et al. (2005), “Feebates, Rebates and Gas-Guzzler Taxes: A Study of Incentives for Increased Fuel Economy”, Energy Policy, 33(6).
Hasunuma, H. et al. (2014), “Decline of Ambient Air Pollution Levels due to Measures to Control Automobile Emissions and Effects on the Prevalence of Respiratory and Allergic Disorders among Children in Japan”, Environmental Research, 131.
Hatfield-Dodds S. (2006), “The Catchment Care Principle: A New Equity Principle for Environmental Policy, with Advantages for Efficiency and Adaptive Governance”, Ecological Economics, 56(3).
Höglund-Isaksson, L. and T. Sterner (2009), Innovation Effects of the Swedish NOx Charge, OECD Global Forum on Eco-Innovation, 4-5 November 2009, OECD, Paris.
Iwata, K. and T.H. Arimura (2008), “Economic Analysis of a Japanese Air Pollution Regulation: An Optimal Retirement Problem under Vehicle Type Regulation in the NOx-Particulate Matter Law”, Discussion Paper 08-15, Resources for the Future, Washington D.C.
Iwata, K. et al. (2014), The Effectiveness of Vehicle Emission Control Policies: Evidence from Japanese Experience, Working Paper E-77, Tokyo Centre for Economic Research, Tokyo.
Johnson, K.C. (2006), “Feebates: An Effective Regulatory Instrument for Cost-Constrained Environmental Policy”, Energy Policy, 34(18).
Kieser and Associates (2004), “Preliminary Economic Analysis of Water Quality Trading Opportunities in the Great Miami River Watershed, Ohio”, Kieser and Associates, Kalamazoo, United States.
Kolstad, C.D. (2011), Intermediate Environmental Economics, 2nd ed., Oxford, Oxford University Press.
Miami Conservancy District (MCD) (2014), Water Quality Credit Trading Program: A Common Sense Approach to Reducing Nutrients, newserver.miamiconservancy.org/water/documents/WQCTPfactsheet2014FINAL_000.pdf.
Naturvårdsverket (2014), Ändring av kväveoxidavgiften för ökad styreffekt: Redovisning av ett regeringsuppdrag, Rapport 7747, Naturvårdsverket, Stockholm.
Newburn, D.A. and R.T. Woodward (2012), “An Ex Post Evaluation of Ohio’s Great Miami Water Quality Trading Program”, Journal of the American Water Resources Association, 48(1).
OECD (2013), "The Swedish Tax on Nitrogen Oxide Emissions: Lessons in Environmental Policy Reform", OECD Environment Policy Papers, No. 2, OECD Publishing, Paris, doi.org/10.1787/5k3tpspfqgzt-en.
OECD (2006), The Political Economy of Environmentally Related Taxes, OECD Publishing, Paris, doi.org/10.1787/9789264025530-en.
OECD (2002), OECD Environmental Performance Reviews: Japan 2002, OECD Environmental Performance Reviews, OECD Publishing, Paris, doi.org/10.1787/9789264175334-en.
Pattanayak, S.K. et al. (2010), “Show Me the Money: Do Payments Supply Environmental Services in Developing Countries?”, Review of Environmental Economics and Policy, 4(2).
Perrot-Maitre, D. (2006), The Vittel Payments for Ecosystem Services: A ‘Perfect’ PES Case?, International Institute for Environment and Development, London.
Pennsylvania State Conservation Commission (PSCC) (2011), PA REAP Resource Enhancement and Protection FY 2009-10 and FY 2010-11 Annual Report, Pennsylvania Department of Agriculture, United States.
Shortle, J. (2012), “Water Quality Trading in Agriculture”, paper prepared for the OECD Joint Working Party on Agriculture and the Environment, COM/TAD/CA/ENV/EPOC(2010)19/FINAL, OECD, Paris.
SOU (2017), “Brännheta skatter! Bör avfallsförbränning och utsläpp av kväveoxider från energiproduktion beskattas?”, Betänkande av Förbränningsskatteutredningen, Statens Offentliga Utredningar, SOU 2017:83, Stockholm.
Sterner, T. and B. Turnheim (2008), “Innovation and Diffusion of Environmental Technology: Industrial NOx abatement in Sweden under Refunded Emission Payments”, Discussion Paper, Resources for the Future, Washington D.C.
Tomich, T.P. et al (2004), “Environmental Services and Land Use Change in Southeast Asia: From Recognition to Regulation or Reward?”, Agriculture Ecosystems and Environment, 104.
USEPA and USDA (2006), “Evaluation Report: Saving the Chesapeake Bay Watershed Requires Better Coordination of Environmental and Agricultural Resources”, US Environmental Protection Agency and US Department of Agriculture, Washington D.C
Notes
← 1. It must be noted that as individual instruments (under a given definition) and their combinations may hold significantly varied specific characteristics and operate under very different contextual conditions, the assessment performed by Drummond et al., 2016, 2015a and 2015 b can be viewed as indicative only.
← 2. The ‘polluter pays’ principle is defined as ‘the principle according to which the polluter should bear the cost of measures to reduce pollution according to the extent of either the damage done to society or the exceeding of an acceptable level (standard) of pollution’ (OECD Glossary of Statistical Terms).
← 3. The pollution haven hypothesis, or pollution haven effect, is the idea that polluting industries will relocate to jurisdictions with less stringent environmental regulations.
← 4. In opposition to the ‘polluter pays’ principle, the ‘beneficiary pays’ principle states that those who benefit from an action should contribute to the cost of that action (Hatfield-Dodds, 2006).
← 5. ‘Feebate’ instruments (or ‘bonus-malus’) levy a tax or charge (‘fee’) on activities or products below a given environmental threshold performance or limit, and provide financial support to activities or products that exceed this performance level (‘rebate’). A key example of a revenue-neutral ‘feebate’ instrument is the Swedish refund emission payment for NOx (OECD, 2013).
← 6. Regulatory capture occurs when a regulatory agency, created to act in the public interest, furthers the interests of groups that dominate the industry or sector it is charged with regulating.
← 7. Except for administration costs, amounting to around 0.7% of total revenue (OECD, 2013).
← 8. Due to the broad revenue neutrality of the instrument, relative product prices did not change, and therefore demand for such products did not alter. Moreover, as product prices do not change, the occurrence of negative distributional impacts is also prevented (OECD, 2013).
← 9. A hotspot may be actively created, for example through the concentration of tradable permits in space, time or sectors, or simply be a location, time or sector in which the marginal cost of pollution is particularly elevated, and not actively addressed by an instrument or instrument combination.
← 10. Most often applied in the context of increasing energy efficiency, the Rebound Effect occurs when increasing production efficiency reduces the cost per unit of service delivered, increasing consumption in accordance with the price elasticity of demand (Berkhout et al., 2000).
← 11. The externality is targeted directly. Also known as a ‘performance-based’ approach.
← 12. The externality is targeted indirectly, by the prohibition or mandating of certain practices. Also known as a ‘technology-based’ approach.
← 13. Broadly speaking, both impact-based and outcome-based approaches may be applied to both upstream and downstream actors.
← 14. The evidence suggests that for a well-known feebate example, the French Bonus-Malus system on new cars that aimed to reduce CO2 emissions, a pivot rate positioned too high, coupled with overly generous rebated, CO2 emissions actually increased.
← 15. Government (or other public body) is the ‘buyer’ of public good ecosystem services (e.g. CO2 sequestration).
← 16. Private actors are the ‘buyers’; with ecosystem services ‘purchased’ delivering (largely) private benefit.
← 17. The targeted regulation complements existing Danish nitrogen regulations, including those contained within the Food and Agriculture Package, adopted in December 2015.
← 18. As opposed to a forward auction, a reverse auction requires credit sellers compete to obtain business from the buyer. In this case, the MCD accepts bids from SWCDs, and then sets the maximum bid value after the first round of submissions.
← 19. WWTPs that engage in the system prior to the introduction of binding legislation require a single credit per pound of nutrient in excess of future binding limits, whilst those that purchase credits after the introduction of legislation will require up to three credits.
← 20. Including Personal Income Tax, Corporate Net Income Tax, Capital Stock and Franchise Tax, Bank Shares Tax, Title Insurance Company Premiums Tax, Insurance Premiums Tax and Mutual Thrift Institutions Tax (PSCC, 2011).