Mind your Ps and Qs! Variable allowance supply in the US Regional Greenhouse Gas Initiative

https://doi.org/10.1016/j.jeem.2022.102620Get rights and content

Abstract

We investigate the new institutional design for the US Regional Greenhouse Gas Initiative (RGGI). The scheme incorporates two allowance reserves that adjust the supply of allowances in the event of unexpectedly high or low allowance demand. These reserves are enacted if the clearing price breaches a predetermined set of trigger prices. Our experiment finds trigger prices act as focal points: the distribution of clearing prices is bimodal and aligns with the trigger prices. We find the existence of focal points can be explained by both the institutional design and subjects’ behavior. Further, decreasing the range between trigger prices increases total revenue but decreases allocative efficiency. The regulation is more sensitive to changes in trigger prices than reserve quantities.

Introduction

A common feature within existing pollution markets is the ability of a regulator to adjust allowance supply. These supply policies are typically incorporated within the initial allocation mechanism of each market: usually, pollution auction rules are adjusted so that the allowance supply can vary in the face of unexpected allowance demand. An increase in supply may assist in reducing firms’ abatement costs whereas a reduction in supply may incentivize innovation in pollution abatement. Such adjustable allowance policies are used in the US Regional Greenhouse Gas Initiative (RGGI), European Union Emissions Trading Scheme (EU-ETS), and the California Cap-and-trade Program (Shobe et al., 2014, Perino and Willner, 2016, Burtraw et al., 2018).1

In this article we investigate one prominent design that has recently been adopted by the US Regional Greenhouse Gas Initiative (RGGI): the use of dual allowance reserves “Qs” and trigger prices “Ps”. An allowance reserve is a separate quantity of allowances that can be used to adjust the initial supply either up or down (Murray et al., 2009, Burtraw et al., 2018). Consequently, a reserve will be used to increase supply when the initial allowance price is sufficiently high: the initial price has to be larger than a ‘trigger’ price, which has been predetermined by the regulator. An allowance reserve can also be used to reduce supply, which works in a similar but opposite manner: if the initial allowance price is too low – below a set trigger price – the reserve will withdraw allowances from the supply.2 We find that the regulator’s choice of “Ps and Qs” has a pivotal effect on the regulatory outcome. In particular, we find the regulator’s choice of trigger prices “Ps” act as focal points for the firms, which determines the allowance clearing price and thus influences revenue generation and allocative efficiency.

The actual RGGI design, which commenced at the beginning of 2021, has been developed to not only contain firms’ costs (by maintaining a pre-existing upper allowance reserve) but also to contain emissions (by introducing a lower allowance reserve). This is observed within Fig. 1.

RGGI’s initial allowance supply is projected to be around 13–15 million allowances per auction (where each allowance accounts for one ton of CO2).3 The upper reserve denoted by the Cost Containment Reserve (CCR) holds a stock of 7,514,778 allowances that can be released to the auction if the initial clearing price is higher than the trigger price of $13 per allowance. The lower reserve – denoted by the Emissions Containment Reserve (ECR) – has a capacity of 6,845,333 allowances that can be withdrawn from the market if the initial clearing price is below $6 per allowance. Although this novel regulatory design commenced in 2021, it is a priori unclear how the design of multiple allowance reserves – with corresponding trigger prices – will affect the functioning of the auction and the regulatory scheme in general. In particular, it is unknown how the interaction and combination of these two allowance reserves will affect the clearing prices, revenue, and efficiency. This is important to study given the economic and the environmental impacts of these regulatory features. Our paper is a first step in this direction.

In this article, we develop an experimental analysis of variable allowance supply, which underpins the core institutional design of RGGI. In our stylized framework, we examine how the design of both an upper and lower allowance reserve (with associated trigger prices) alters the auction outcome. Our aim within this article, then, is to investigate how the regulator’s design choices – of trigger prices, initial supply, and allowance reserves – impact the auction outcome. We find evidence that the regulator’s chosen trigger prices act as focal points: the distribution of clearing prices is bimodal, which aligns with the trigger prices. We show that this is not just an institutional design feature of having two allowance reserves but is exacerbated by buyers’ bidding strategies. Furthermore, the equilibrium clearing prices are positively related to independent changes in both trigger prices.

We also find the range between the two trigger prices – what we denote as the trigger-price collar – has significant implications for the outcome of the scheme. As the width of the trigger-price collar decreases, i.e., the “soft” price floor and ceiling move closer together, we find this increases total revenue and the amount of allowances sold, but decreases allocative efficiency. Throughout our analysis, we find the scheme is more sensitive to changes in trigger prices than changes in quantities. Holding the maximum number of allowances constant, we find evidence that increases in the initial cap (with a smaller CCR and constant ECR) decreases the final clearing price as well as increasing the number of equilibrium allowances sold. However, we find that changes in quantities have no significant effect on revenue or allocative efficiency.

As our evidence highlights the focal-point characteristics of the regulator’s trigger prices, we complement the prominent literature that investigates the so-called focal-point hypothesis (e.g., Isaac and Plott, 1981, Coursey and Smith, 1983, Scherer and Ross, 1990, Knittel and Stango, 2003, Gode and Sunder, 2004, Engelmann and Müller, 2011, Rey and Tirole, 2019). This hypothesis argues non-binding price controls – usually “hard” price ceilings – may create a focal point in pricing decisions and thereby reduce competition. Although empirical evidence does exist for focal points in a limited number of markets (e.g., Knittel and Stango, 2003), the experimental laboratory evidence generally finds no collusive focal-point effects (Isaac and Plott, 1981, Coursey and Smith, 1983, Engelmann and Müller, 2011). In contrast to this literature, we find evidence of focal-point effects within a trigger-price system. This strengthens the argument for the existence of focal points. In addition, it widens the support for the focal-point hypothesis from traditional non-binding “hard” price ceilings as observed in the literature to “soft” price floors and ceilings (where the clearing price can be above or below the predetermined trigger prices). We show that these focal-point effects arise from both the institutional design and from bidding strategies in the auction.

Our main innovation centers on the analysis of dual allowance reserves, which has not been explored by researchers thus far. In doing so, we are able to analyze how the regulator’s institutional design parameters – both separately and in combination – affect the operation of the regulatory scheme. More generally, our research also advances the literature on pollution market supply-side policies. The original idea of combining both prices and quantities to regulate can be traced back to Roberts and Spence (1976) and Weitzman (1978).4 Since then a broad literature has developed to investigate how alternative supply-side polices perform with respect to cost effectiveness. In particular, Murray et al. (2009) provide arguments for the use of allowance reserves and a number of studies have consequently investigated how a single allowance reserve – to ensure cost containment – would perform (Burtraw et al., 2010, Fell and Morgenstern, 2010, Grüll and Taschini, 2011, Fell et al., 2012, Shobe et al., 2014, Stranlund et al., 2014, Kollenberg and Taschini, 2016, Khezr and MacKenzie, 2018, Salant et al., 2019). Overall, this literature tends to find that adjusting allowance supply assists in cost containment and can improve welfare.5 Yet in our analysis, with dual allowance reserves, we find this may not always hold: the focal-point effects could mean that firms’ costs increase if the regulator decides to raise either of the trigger prices.

In terms of methodologies, the majority of this literature uses either theoretical or simulation modeling. For example, in relation to California’s cap-and-trade market, Borenstein et al. (2019) show, using simulations, that the distribution of clearing prices may be skewed towards either the price floor or the ceiling due to high levels of uncertainty of the business-as-usual emissions as well as the relative price-inelasticity of pollution abatement. Hence equilibrium prices are likely to be concentrated at trigger prices. This study provides useful insight as to the workings of emission markets and in particular the California program and suggests that trigger prices can potentially become focal points. It may be hard, however, to use this to infer policy implications for RGGI because the new features of the dual allowance reserve system – and the corresponding distinct incentive effects – are unique to RGGI. Our goal therefore is to examine the use of dual allowance reserves and trigger prices as adopted by RGGI. In contrast to Borenstein et al. (2019), our approach focuses on examining the behavioral response to these design features, using experimental methods. Behavioral reactions to market design features can be difficult to capture using simulations. Further, investigating the behavior of bidders using observational data can be extremely challenging, as individual bidding behavior and actions are private information, as are valuations, and the counterfactual is not known, which makes it hard to obtain a meaningful comparison.

Consequently, there is a branch of literature that focuses on the experimental evidence of supply-side policies.6 The major benefit is that this can provide evidence of agent behavior within a controlled environment with known valuations in order to test the impacts of regulatory design (Shobe et al., 2010, Shobe et al., 2014, Holt and Shobe, 2016, Perkis et al., 2016, Burtraw et al., 2018). To date, however, the analysis of dual allowance reserves has been neglected. Although Burtraw et al. (2018) focus on analyzing three ECR treatments relative to a no-ECR policy, they do not provide an investigation of the ECR in combination with a CCR nor the ramifications of the regulator’s institutional design parameters. Further, Stranlund et al. (2014) consider price controls and banking within an allowance market but their price controls are “hard” price floors and ceilings, which allow unlimited changes within allowance supply.

Our contribution is therefore in providing a greater understanding of the role and ramifications of dual allowance reserve policies and the associated focal points that are created. We find that the regulator’s predetermined choice of initial allowance supply, trigger prices, and allowance reserves, have important implications for the functioning of the regulatory scheme. Our evidence suggests the price of carbon can be heavily influenced by the regulator’s choice of trigger prices rather than the cost of pollution control (which may have long-run efficiency impacts). The findings from this study will provide further knowledge for the designers of prominent allowance markets.

Our article is organized as follows. In Section 1.1, we provide a background on the institutional design within RGGI. In Section 2 we outline the experimental design. Section 3 presents results at the market level as well as examining individual bidding behavior. Section 4 provides some concluding remarks.

The US Regional Greenhouse Gas Initiative (RGGI) is a regional cap-and-trade market that commenced in 2008 and regulates emissions from the utility sector in the Northeastern US states. The majority of allowances within RGGI are allocated via a quarterly auction, where, on average, 43 firms participate in bidding (2017–2019 inclusive) and has generated a cumulative revenue of approximately US$3.5 billion since 2008.7 The auction format used is the multi-unit uniform-price auction, which is the predominant method to distribute pollution allowances (Lopomo et al., 2011). Each individual auction generates around US$75 million in auction revenue, which is invested in mainly energy and consumer programs.

As the program has developed over time – with two major program reviews in 2012 and 2017 – the auction rules have also evolved. Unlike conventional allowance auction designs that usually offer a fixed amount of allowances at each auction, the auction rules of RGGI have been developed to vary the supply of allowances. Within the first program review in 2012 (and implemented in 2014), the Cost Containment Reserve (CCR) was established to reduce significant price increases (RGGI, 2013). As the name suggests, this was implemented to dampen any cost changes due to changes in demand shocks and to provide a more ‘hybrid’ approach (combining price and quantity measures), to allowance supply. The CCR was originally a fixed amount of 10 million allowances with the trigger price being initially set at US$4 per allowance with increases year-on-year (the initial allowance supply was also continually decreasing year-on-year). Within the 2017 Program Review (RGGI, 2017), a similar but opposite reserve was developed. From 2021, RGGI will implement the Emissions Containment Reserve (ECR). The rationale for establishing an ECR is to reduce supply of allowances at sufficiently low prices, to ensure a sustained incentive to invest in abatement technology. The CCR and ECR have annual capacities – 10% of the regional (non-adjusted) cap – so if the allowances are exhausted (filled to capacity) in a previous auction then these reserves are no longer available for use until the next calendar year.8

As described earlier, and shown in Fig. 1, the RGGI design in 2021 is fairly symmetric in terms of the size of allowance reserves with both the ECR and CCR containing approximately half of the number of allowances contained in the initial supply. The trigger prices for the ECR and CCR will initially be set at $6 and $13 per allowance, respectively, and will increase by 7% per annum. Yet, note although these trigger prices increase at a constant rate each year, the absolute size of the trigger-price collar increases over time due to different baselines. This is shown in Fig. 2.

Given the dual reserves are primarily aimed at containing costs (dampening prices) as well as ensuring that emissions are contained, it is interesting to investigate how these perform. Our experiment can directly examine how prices change within this new regulation and the extent to which emissions are contained. Our results imply that as the trigger-price collar increases, ceteris paribus, we would expect to see increases in allowance prices, a reduction in allowances sold (hence a lower level of emissions) but an increase in allocative efficiency.

Interestingly, within the development process of the ECR, discussion arose over the potential for focal-point effects from the implementation of allowance reserves. In contrast to our findings, the general conclusion was that the anchoring of bids was unlikely to occur.9 Yet it is clear from our analysis that how the trigger prices are chosen – both separately and collectively – have important implications for revenue and allocative efficiency. What our evidence suggests is that, within RGGI, the regulator could make improvements in its criteria of importance (such as revenue generation, efficiency, etc.) by adjusting the trigger prices in both an absolute and relative manner. However there may exist tradeoffs. A tighter trigger-price collar, may increase revenue but at the expense of reduced allocative efficiency.

Section snippets

Dual allowance reserves

In this section, we develop a stylized framework that reflects the key institutional design features proposed within RGGI. This framework then informs our experimental design.

Results

In this section, we analyze results at the aggregate market level and at the individual level. We focus first on the fundamental auction outcomes relating to the allowance clearing price and the number of allowances sold. We then study two measures of auction performance: total revenue raised in the auction (i.e., price times the number of allowances sold), and allocative efficiency, which is defined as the percentage of the maximum possible surplus actually achieved in the auction. Summary

Concluding remarks

In the majority of pollution markets an increasing trend has been to allow for variable allowance supply. Adjusting allowance supply can assist firms with their abatement costs as well as spur innovation in pollution abatement technologies. A prominent design – used within the US Regional Greenhouse Gas Initiative (RGGI) – is the use of allowance reserves that can either increase or decrease the initial supply. Allowance reserves are fixed quantities of allowances that can either flow in or out

References (57)

  • KhezrP. et al.

    Permit market auctions with allowance reserves

    Int. J. Ind. Organ.

    (2018)
  • KhezrP. et al.

    Revenue and efficiency in pollution permit allocation mechanisms

    Energy Econ.

    (2021)
  • KollenbergS. et al.

    Emissions trading systems with cap adjustments

    J. Environ. Econ. Manage.

    (2016)
  • LiCalziM. et al.

    Tilting the supply schedule to enhance competition in uniform-price auctions

    Eur. Econ. Rev.

    (2005)
  • McAdamsD.

    Adjustable supply in uniform price auctions: Non-commitment as a strategic tool

    Econom. Lett.

    (2007)
  • McKitrickR.

    A derivation of the marginal abatement cost curve

    J. Environ. Econ. Manage.

    (1999)
  • PerinoG. et al.

    Procrastinating reform: The impact of the market stability reserve on the EU ETS

    J. Environ. Econ. Manage.

    (2016)
  • PizerW.A.

    Combining price and quantity controls to mitigate global climate change

    J. Public Econ.

    (2002)
  • RobertsM.J. et al.

    Effluent charges and licenses under uncertainty

    J. Public Econ.

    (1976)
  • ShobeW. et al.

    Elements of emission market design: An experimental analysis of California’s market for greenhouse gas allowances

    J. Econ. Behav. Organ.

    (2014)
  • StranlundJ.K. et al.

    An experimental analysis of compliance in dynamic emissions markets

    J. Environ. Econ. Manage.

    (2011)
  • StranlundJ.K. et al.

    Price controls and banking in emissions trading: An experimental evaluation

    J. Environ. Econ. Manage.

    (2014)
  • AusubelL.M. et al.

    Demand reduction and inefficiency in multi-unit auctions

    Rev. Econom. Stud.

    (2014)
  • BorensteinS. et al.

    Expecting the unexpected: Emissions uncertainty and environmental market design

    Amer. Econ. Rev.

    (2019)
  • BurtrawD. et al.

    Quantities with prices

  • CamererC.F.

    The Promise and Success of Lab-Field Generalizability in Experimental Economics: A Critical Reply to Levitt and List

    (2011)
  • CasonT.N. et al.

    Complying with Environmental Regulations: Experimental EvidenceDiscussion Papers Series 637

    (2020)
  • CourseyD.L. et al.

    Price controls in a posted offer market

    Am. Econ. Rev.

    (1983)
  • Cited by (8)

    • Adjustable emissions caps and the price of pollution

      2023, Journal of Environmental Economics and Management
    • Investment Incentives in Tradable Emissions Markets with Price Floors

      2023, Journal of the Association of Environmental and Resource Economists
    View all citing articles on Scopus

    We thank Zhengyang Bao for excellent research assistance and Monash University and the University of Queensland for research funding. We are also grateful for helpful comments provided by audiences at the Monash Environmental Economics Workshop, ESA World Meetings (online) and QUT and QUEX (joint initiative of UQ and Exeter) seminars. The research has ethics approval from the University of Queensland, Business, Economics and Law, Low and Negligible Risk Ethics Sub-Committee (Approval Number 2018000828).

    View full text