Conservation and Policy Conservation Performance Payments for Carnivore Conservation in Sweden ASTRID ZABEL∗ AND KARIN HOLM-MULLER†¨ ∗Universitaetsstr. 16, Environmental Policy and Economics (ETHZ), 8092 Zurich, Switzerland, email [email protected] †Institute for Food and Resource Economics (ILR), Rheinische Friedrich-Wilhelms-Universitaet, Nussallee 21, 53115 Bonn, Germany Introduction discuss this approach as an alternative strategy to con- ventional ex post compensation to alleviate carnivore- Many carnivores require vast territories, and as human livestock conflicts. population increases, more pristine natural areas are be- ing developed and converted into agricultural land. Un- surprisingly, carnivores that live at the fringe between Conservation Performance Payments wild and agricultural land occasionally prey on livestock. Predation of livestock can result in severe economic In search of new solutions to alleviate carnivore-livestock losses (Mishra 1997; Thirgood et al. 2005; Woodroffe conflicts, a performance-payment scheme was developed et al. 2005). Herders, whose livelihoods depend on live- and implemented in Sweden. Conservation performance stock, often seek to kill predators to prevent further dam- payments are monetary or in-kind payments made by a age. Conservationists, on the other hand, engage in mea- paying agency to individuals or groups and are condi- sures to protect endangered carnivores because they are tional on specific conservation outcomes (Albers & Fer- appreciated as an important component of biodiversity. raro 2006). Performance payments are made on a strict Viable solutions to make coexistence of wildlife and live- quid pro quo basis, and the amount depends on the level stock acceptable to conservationists and livestock own- of conservation outcome. Their focus is completely on ers are much needed and are likely to be increasingly outcome; the actions that led to the conservation out- sought after as human sprawl increases. come are not relevant. In the context of carnivore conser- Schemes that provide ex post compensation to live- vation, the conservation outcome can readily be defined stock owners for losses to predation have been imple- as the number of carnivore offspring in a certain area. mented in many places around the world, but they have This conditionality concept gives the paying agency the not proven to be widely successful. Most of the schemes’ possibility to pay exactly and solely for the conservation deficiencies can be ascribed to one or several of the goal it strives for and is thus an interesting solution to the following problems: moral hazard (Cozza et al. 1996; prevailing principal–agent problem. Swenson & Andren´ 2005), high transaction costs (Saber- A well-known problem of ex post compensation sche- wal et al. 1994; Blanco 2003), long time lags (Fourli mes is moral hazard (Cozza et al. 1996; Swenson & 1999; Madhusudan 2003), and problems of trust and Andren´ 2005), which leads to suboptimal levels of live- transparency (Montag 2002; Western & Waithaka 2005). stock protection. Conservation performance payments Mainly due to these problems, practitioners and analysts do not give rise to moral hazard because the payments are have denounced ex post compensation schemes as inad- contingent on conservation outcomes and not livestock equate, fraudulent, and cumbersome (Naughton-Treves losses. et al. 2003). Drawing on empirical insights into a con- The conservation performance payments are issued for servation performance-payment scheme in Sweden, we carnivore offspring and the amount is calculated to offset Paper submitted May 14, 2007; revised manuscript accepted October 25, 2007. 247 Conservation Biology, Volume 22, No. 2, 247–251 C 2008 Society for Conservation Biology DOI: 10.1111/j.1523-1739.2008.00898.x 248 Conservation Performance Payments all the future damage that the animals are expected to (Swenson & Andren´ 2005). Our recent survey revealed cause. Hence, compensation can be counted on and there that, on average, herders lose nearly 20% of their reindeer are no time lags, which is not always the case in ex post flocks to carnivore attacks each year. compensation. Furthermore, herders do not have to bear The conservation performance payments are made by the transaction costs of searching for animals killed by the Swedish state to Sami villages contingent on the carnivores and the subsequent filing for compensation. number of carnivore reproductions that are certified on Thus, the incentives to keep carnivores alive and let them the villages’ reindeer grazing grounds. The payments are reproduce are likely to be higher under a performance- made irrespective of actual predation incidents. Incen- payment scheme than under an ex post compensation tives to apply optimal levels of livestock protection are scheme. Nevertheless, the assessment and verification of not distorted and consequently the scheme does not give the defined conservation outcome, for example carnivore rise to moral hazard. Furthermore, there are no prob- offspring, are likely to give rise to substantial transaction lems with time lags because payments are made for costs for the conservation agency. In certain cases this carnivore offspring (i.e., while the animals are too may be a serious drawback of these schemes. young to cause damage). The amount of payment is de- Another problem arises when the responsibility for the termined according to the monetary damage that the off- conservation outcome cannot be attributed directly to spring are expected to cause throughout their lifetime. individuals. In these cases groups rather than individ- In 2007 the payment per certified wolverine and lynx re- uals can be granted the conservation performance pay- production was SEK200,000 (approximately US$29,000). ments. This approach was chosen in the Swedish scheme In Sweden payments are also made for the regular and discussed later. Rewarding groups can in turn give rise occasional occurance of lone wolverines and lynx. These to problems of collective action. Finding solutions to payments of SEK70,000 and SEK35,000 (approximately these problems is a crucial prerequisite for a conserva- US$10,150 and US$5,075), respectively, are lower than tion performance-payment scheme to work. the payments for offspring. Once the money has been After a short description of the Swedish system of paid, the Sami villages have the authority to decide on performance-based compensation we present our anal- the use and internal distribution of the money (Regerin- ysis of strategies developed in Sweden to circumvent the gens Proposition 2000/01:57). collective-action problem. We do this with reference to a set of criteria defined by Ostrom (1990). Our aim was to evaluate the common-pool regimes of the Swedish reindeer herders and to provide an assessment of the Internal Management of Payments workability of the ex ante compensation scheme. If payments are set high enough to assure full compensa- tion, the internal payment-distribution scheme needs to create a situation in which each individual herder has an The Swedish Case incentive to refrain from killing carnivores to reduce the risk of predation incidents. Poaching carnivores would In 1996 the Swedish government implemented a new simultaneously reduce the likelihood of obtaining off- performance-payment strategy to attain and maintain sta- spring and performance payments for these offspring in ble populations of wolverines (Gulo gulo), lynx (Lynx the next year. This makes strategies to solve the pris- lynx), and wolves (Canis lupus) within the country. oner’s dilemma of collective action necessary. Our main Wolverines, which are listed as vulnerable on the World interest was to determine whether the Sami villages in Conservation Union Red List of endangered species Sweden have set up common-pool regimes to manage the (Mustelid Specialist Group 1996), and lynx roam the distribution of the payments toward collective action in wilderness areas in the northern parts of Sweden. They conservation and whether there are interdependencies share the premises with reindeer (Rangifer tarandus among the villages’ structural characteristics and their tarandus) that are herded by the indigenous Sami peo- modalities of distributing the money. ple. Reindeer herding has been central to Sami livelihoods Theindicatorswechosefortheidentificationof for centuries and is still deeply enrooted in their culture. common-pool regimes were derived from Ostrom’s anal- Today about 20,000 Sami people live in Sweden and ap- ysis of several long-term, common-pool systems of re- proximately 2,500 work full time in the reindeer business source management around the world. Ostrom (1990) (Jordbruksdepartementet & Sametinget 2004). devised 7 design principles for institutional arrangements Wolverines and lynx prey on reindeer (Persson 2005; that are hypothesized to help make systems of resource Danell et al. 2006) and especially during the winter are management persist over the long term. The 4 principles assumed to essentially depend on the abundance of rein- we decided to investigate in the Swedish context were deer (Pedersen et al. 1999). Roughly estimated, each clearly defined boundaries, minimal recognition of rights wolverine and lynx annually kills around 40 reindeer to organize, collective-choice management, and congru- Conservation Biology Volume 22, No. 2, 2008 Zabel & Holm-Muller¨ 249 ence between appropriation and provision rules and local mentioned that they were aware of unsatisfied members. conditions.