Working Paper

Soft and Hard Price Collars in a Cap-and-Trade System: A Comparative Analysis

Jun 9, 2011 | Harrison Fell, Dallas Burtraw, Richard D. Morgenstern, Karen L. Palmer, Louis Preonas


We use a stochastic dynamic framework to compare price collars (price ceilings and floors) in a cap-and-trade system. Sources of uncertainty include shocks to baseline emissions, affectingcorresponding abatement costs, and shocks to the supply of offsets. We consider a continuum between soft collars, which have a limited volume of additional emission allowances (a reserve) available at theprice ceiling, and hard collars, which provide an unlimited supply of additional allowances, thereby preventing allowance prices from exceeding the price ceiling. For all cases considered, we set the price floors and ceiling such that the expected cumulative emissions net of offsets are equal to the cumulative allowances. Consequently, increasing the size of the allowance reserve requires higher price ceilings and floors, and a lower probability of reaching the ceiling. Across most parameter values examined, we find that increasing the size of the allowance reserve leads to lower expected net present values of compliance costs, although the differences are not large. However, when offset supply shocks are highly persistent and exhibit strong (negative) correlation with baseline emission shocks, hard collars deliver noticeably lower expected costs, though with a wider range of emission outcomes than the soft collars.

A new RFF paper looks at how different options for controlling the costs of climate regulation in the United States fare in terms of expected emissions and overall program costs.

A “price collar” on emissions allowances has emerged as a favored approach for containing the potential costs of a U.S. cap-and-trade system regulating greenhouse gases. By setting both a ceiling and floor on allowance prices, a collar would reduce year-to-year volatility in abatement costs and prevent the overall costs of the cap-and-trade program from being either much higher or much lower than anticipated. Over the long term, the expectation is that it would lower the costs of achieving a given emissions target.

In practice, the collar would work by making additional allowances available to emitters to keep prices from rising above the ceiling, while removing allowances to keep prices from falling below the floor. Two variants of a price collar have emerged: 

  • A “hard collar” places no limits on the number of allowances that can added to the system at the ceiling price.  While it contains costs, it also reduces the probability that the long-term emission goal will be met. 
  • A “soft collar” sets a limit on the number of allowances that can be added to the system by creating a fixed allowance reserve.  This would provide some of the benefits of reduced price volatility while giving greater assurance that environmental goals will be met.

A soft collar mechanism is included in the climate legislation recently introduced by Senators Kerry and Lieberman. In a new RFF Discussion Paper, “Soft and Hard Price Collars in a Cap-and-Trade System: A Comparative Analysis,” Fellow Harrison Fell and colleagues at RFF explicitly consider the effects of uncertainty in the market for emissions allowances on the selection of a particular type of price collar. They find that for a given price collar, a soft collar tends to lead to higher expected costs and price variability, while a hard collar tends toward higher emissions volatility.

Perhaps the most important finding is that the lion’s share of cost savings is achieved with the initial introduction of additional allowances under a price collar. As further additional allowances are made available (that is, the collar is hardened) additional cost savings are achieved, but they come at a diminishing rate. Under most plausible scenarios the cost savings from a hard collar are relatively modest.  In one possible case, however, a hard collar can result in significantly greater cost savings than a soft collar. 

An important feature of the underlying problem is the degree of uncertainty about future costs. The model characterizes two sources of uncertainty: the year-to-year variation in the cost of emissions reductions in the U.S. economy and the variation in the costs of international offsets. Further, the model accounts for the fact that variation in these costs might be related. For example, a domestic economic boom might increase the cost of domestic emissions reductions, but at the same time it also might be associated with higher demand in global commodity markets, which would limit the supply of international offsets. In this case the availability of offsets might amplify the uncertainty about future costs, even as they lower overall costs. Another important parameter is the persistence of shocks to domestic abatement costs and the international supply of offsets.  With persistent and highly negatively related cost shocks, a hard price collar will achieve lower expected costs than a soft collar.