The Scottish Government

« Previous | Contents | Next »

Listen

5. Case Studies of Conservation Auctions

This section reviews case studies of conservation auctions covering the USA, Australia, continental Europe and the UK. For each auction, we review the problem addressed, objectives, auction design, auction outcomes, and lessons learned. The schemes reviewed differ in their policy goals, their ways of setting reserve prices and their methods of assessing environmental benefits and ranking bids.

5.1 Results of postal survey

We conducted a postal survey of agricultural ministries in the EU-15. We asked respondents whether they were aware of any tendering schemes for conservation contracts in their country. We asked for the name of such schemes, if any, and the contact details of the person who may be able to answer questions about the design and operation of such schemes and experiences gained. We attached a one-page questionnaire specifying the points on which we would like some information. The purpose of the survey was to gauge lessons learned, to identify solutions to limitations and practical difficulties experienced in the running of conservation auctions, and to gauge opinions on ways of increasing the performance of auctions.

We received only two responses, from Sweden and Belgium, stating that no such mechanisms were used in these countries.

5.2 Conservation Reserve Program auction ( USA)

Problem

The aim of the CRP is more than just to obtain environmental services from landholders. It also aims to (i) reduce water and wind erosion; (ii) improve water quality; (iii) reduce sedimentation; (iv) preserve soil productivity; (v) improve habitat for fish and wildlife; (vi) curb production of surplus commodities; and (vii) provide needed income support for farmers. These goals are subject to the enrolment of lands for specific environmental policy priorities and in specific regions (Reichelderfer and Boggess 1988). Under this program, landholders receive government funds for retiring their lands from farm production for a period of 10 to 15 years ( FSA- USDA 1999). Farmers submit bids to the US Department of Agriculture ( USDA) for individual parcels of land, indicating the payment required to take the land out of agricultural production and put it to conservation use.

Design

Bid selection criteria have evolved over time (Francis 2001; HRSCEH 2000; Osborn, Llacuna and Linsenbigler 1995). Before 1990, bids were compared in terms of dollars per hectare of conserved land, implicitly assuming all hectares were of comparable value. All eligible bids at or below a predetermined level were accepted. Such a payment cap was not made public until the auction was completed. Current CRP auctions employ an environmental benefit index to compare bids. This index accounts for land quality heterogeneity and weights various environmental objectives according to their relative importance. Figure 2 shows the bid scoring system for the CRP.

The left hand column shows the scheme's six environmental objectives - from wildlife benefits to air quality improvements. The weights reflect USDA's preferences for the different attributes. Air quality improvements, for example, carry a weight of only 25, indicating that USDA considers this attribute significantly less important than erosion control, which carries a weight of 100. Each bid is ranked according to its contribution to each of these objectives. Bids are assessed on the basis of indicators (proxies) which relate either to certain observable attributes of the land (such as proximity to water bodies or to erodibility) or to certain land management practices (such as land cover type). Each attribute or land management practice carries a score which is multiplied by the weights shown in Figure 2. Bids would normally be ranked for acceptance according to the ratio of the overall score to the amount of the bid. However, the actual selection procedure for CRP bids is slightly different in that the bid amount enters the score as another attribute (with a negative sign). Bids that exceed a minimum overall score are accepted.

Figure 2: Bid scoring criteria for the Conservation Reserve Programme

Source: USDA, Economic Research Service

Results

In 1999, the programme enrolled 36 million acres of land (equivalent to 10 per cent of farmland in the US). In 2001, the annual bill for the program was US$1.5 billion. Currently (2005), approximately 33.5 million acres are enrolled in the CRP. Each contract covers an average of 74 acres with an average rental rate of $45.95 per acre.

Lessons learned

From its beginning in 1986, the CRP was conceived as a multiple sign-up scheme, whereby auctions would be carried out repeatedly over time. This has allowed landholders to learn where the (implicit) bid cap lies and to gradually adjust their bids to this cap (Reichelderfer and Boggess, 1988). Repeated auctions allow bidders to learn from previous rounds so that they can adjust their bids using this information. The result is an increased information rent, which reduces the efficiency of the auction. Hailu and Schilizzi (2004) have shown that even with the simplest learning, based only on one's own past results and without any knowledge of others' results, bidders can converge towards the marginal bid, turning the auction into something equivalent to a fixed price scheme. The implication is that auctions repeated identically over time erode their efficiency potential. To avoid this, the government agency can modify the rules of the game ( e.g. the bid selection criteria) from round to round, so as to introduce some uncertainty and reduce the relevance of information from past rounds.

5.3 BushTender Trial (Australia)

Problem

In 2001, the Victoria government (Australia) wanted to test the idea according to which auctions could efficiently purchase public environmental goods from private landholders. The good in question was biodiversity as captured through improved 'bush' management. 'Bush' in Australia refers to the original deep rooted ligneous vegetation prior to clearing and farming, which in agricultural areas survives today usually in isolated patches. Key questions included the farmer rate of participation and whether an auction could be more cost-effective, budget wise, than a traditional fixed price scheme.

Design

Several micro-regions in Victoria were designated, and a budget (not counting administrative costs) of A$400,000 was allocated in a first round (2001) and A$800,000 in a second round (Gippsland trials, 2002-03). Expressions of interest were first called for, then government officers visited the farms and the proposed land areas up for tender. Ecological data were collected on these areas and analysed by scientists to devise a Biodiversity Benefits Index ( BBI), defined as:

This BBI defined a benefit to cost ratio for the government. The Conservation Value Score ( CVS) reflected the ecological value of the site (its potential to yield ecological benefits) and the Habitat Amelioration Score ( HAS) reflected the value of the farmer's efforts in moving the current state of the area towards its ecological potential. Contracts were then negotiated one to one with each interested farmer, whereby a land management plan was set up for which the farmer would be paid, if successful, through the competitive auction mechanism. Contract durations of 3 years were offered in round 1 and of 3 or 6 years (with a possible commitment of 10 years using covenants) in round 2.

A sealed-bid discriminatory price auction was used to 'sell' the contracts to farmers and buy their pre-negotiated services. Bids were ranked according to the BBI ratio, from highest value per dollar bid down, until the budget constraint was hit. In the first round, information was not given to farmers regarding their BBI, whereas it was given in the second round, mainly for (government) learning purposes. Between the two rounds, laboratory experiments were carried out to explore the effect of withholding or not such information from landholders. Effects on collusion and other bidding distortions were observed to be small in the lab (Cason et al., 2003).

Results

A full analysis of both rounds of the BushTender experience is due to be available some time by end 2005. Meanwhile, the following results can be provided (Table 2):

Table 2: BushTender results

* The number of sites assessed was greater than the number of properties as most properties offered more than one site.
** A habitat zone represents a unique Ecological Vegetation Class ( EVC) / quality combination within a site.
*** Note that 1 ^st round and 2 ^nd round figures cannot be compared, as they relate to two different regions.
Source: Table compiled by the authors. Currency is in Australian dollars.

There was an acceptance rate of 97% of offered agreements by participants. In round 2, all but one landholder opted for a 6-year contract. Landholders demonstrated a strong preparedness to commit to a range of management actions, including stock exclusion, retaining trees and fallen timber and controlling weeds and introduced species (beyond current duty of care requirements). Fencing and supplementary planting recorded lower levels of uptake due to the site-specific nature of these actions. To date (mid 2005), almost 95% of participating landholders have complied with their contractual agreements.

Subsequent analysis of first round results by Stoneham et al. (2003) claimed a benefit of 700% of that which would have been obtained through a fixed price scheme, though this claim depends on the choice of the counter-factual fixed price to which it is compared. We will revisit this figure in Section 6.

Lessons learned

Gary Stoneham, of the Department of Primary Industries of the State of Victoria, Australia, has been kind enough to share his views on what he believes his team has learned through the BushTender experience over the last four to five years. His views are valuable as he was directly involved in the development and the running of the BushTender auctions. These are summarised in Box 6.

(1) The use of a discriminatory price auction does not reveal the opportunity cost curve, as it is optimal for bidders to shade their bids. As a result, the cost curve remains unknown. It can however be estimated using a model, like that of Latacz-Lohmann and van der Hamsvoort (1997), that allows one to work backwards from bids, assumed to be optimal, back to the original costs. Of course, one obtains estimated costs, not the real costs. The amount of error made in the estimation can be measured using controlled laboratory experiments, as has been done by Schilizzi and Latacz-Lohmann (2005a). One then obtains an estimate of the error made in the empirical cost recovery.

In addition, the following can be said:

• Carefully designed conservation auctions can work out in the field and yield non-negligible benefits, both in terms of ecology and in terms of budgetary outlays.
• Transaction costs for the first round of BushTender, which included on-site research, ecological scoring and auction administration costs, amounted to roughly between 50% and 60% of the amount used in the auction. Falconer and Whitby (1999) report a variation in Europe of between 30% and 80%). It appears that such levels of transaction costs are to be expected as part of both the government agency's and the farmers' learning investment. They should diminish as they gain experience over time. Assessment of the second round (Gippsland) could shed some light on the government side, although this round is administered somewhat differently from the first, so that the two rounds are not strictly comparable.
• The involvement of government officers and their dedication to explain to farmers the ins and outs of this new payment system are important for securing sufficient participation and thereby the level of competition necessary for the auction to play its efficiency role. The factors favouring and explaining landholder participation, or lack thereof, have been analysed by Ha et al. (2003).
• Field trials like BushTender have a learning function. They stand somewhere in between full scale policy implementation, like the Conservation Reserve Program which applied throughout the United States, and controlled laboratory experiments. In principle, the sequence running from theory to lab experiments to small scale field trials constitutes itself a learning process for full policy implementation. The question has been asked whether a previous stage can predict the outcomes of the following stage, and in particular, whether lab experiments can predict outcomes in the field. As Vernon Smith, who won in 2002 the Nobel Prize for pioneering experimental economics, has said, the issue is not whether lab experiments can predict outcomes in the field, but whether they allow us to make all the important mistakes at low cost, before field implementation. Smith reckons this is nearly always the case. Lab experiments remain very valuable, but field trials remain necessary before full scale policy implementation.

5.4 Auction for Landscape Recovery (Australia)

Problem

This auction field trial was part of a larger national programme in Australia, aimed at trialling various market-based instruments. Nine such projects implemented one variety or other of conservation auctions (see Appendix 3 for an overview). Unlike the BushTender trials, the Auction for Landscape Recovery ( ALR) aimed at securing multiple benefits from land management improvements, namely biodiversity enhancement, salinity control, and groundwater recharge abatement (recharge compounds salinity).

Design

The Avon River Catchment in the state of Western Australia was designated as the target area. As with BushTender, the programme was conducted over two rounds. Landholders could put up more than one bid each and were encouraged to put in joint bids. Before the bids were put forward for evaluation, an independent group of experts assessed if the proposed action would lead to the identified outcomes. A total of A$200,000 was available for farmer payments, with A$93,000 spent in round 1, leaving A$107,000 for round 2.

The ALR was conducted as a simple sealed-bid price discriminating auction, similar to BushTender. Landholders who had expressed interest were encouraged to submit a tender describing their proposed management activities, anticipated environmental outcomes and a bid. The tender process was communicated as rewarding those who deliver the greatest environmental benefit per dollar. Producers were reminded that the scheme is competitive. However, they could include a rent element in their bids if they wished.

Tenders were evaluated using a regional metric of 'biodiversity complementarity' (Faith and Walker, 1996) within a systematic conservation planning framework (Margules and Pressey, 2000). This metric, unlike the BBI, accounts for synergistic aspects due to number, size and distance of several areas; the BBI focuses on the individual value of each land area. However, as part of the research project, an environmental benefits index ( EBI) was also calculated, for comparability purposes with BushTender, namely in terms of cost-effectiveness (comparison yet to be completed).

Results

Full disclosure of results are still in the waiting (the second round was just carried out in March-April 2005). In the first round (April 2004), a total of 56 bids were received from 38 landholders - some landholders putting more than one bid. Over the two rounds, there were 89 tenders submitted by 57 landholders.

Out of 56 tenders submitted in round 1, a total of 10 were successful. These mainly offered to carry out fencing, revegetation and feral baiting. The relative weighting of these activities, implicit in the scoring method, was not known to bidders.

Lessons learned

From some preliminary analyses carried out by White and Burton (2005) at the School of Agricultural and Resource Economics of the University of Western Australia, some interesting outcomes emerge from this experience:

1) The building by natural scientists of a comprehensive scoring index for ranking multidimensional auctions is an exercise fraught with pitfalls. Subjectivity cannot be avoided, even if it is buried in the appearance of an objective measure (the scoring index). In particular, the relative weighting of different ecological benefits remains implicit and unknown to decision makers, and even to the scientists themselves if they do not have an explicit weighting procedure. If some of this information is to be communicated to farmers, it matters how this is done and how the information is to be interpreted. Thus, it must be acknowledged that the policy maker has some degree of discretion in choosing the relative weightings: they themselves form part of the (more or less implicit) policy preferences.

2) Calculations done using bid data from the first round seem to show that a number of landholders bid below their opportunity costs. Follow-up interviews suggested that some of them would have carried out conservation works even without payment, raising the issue of the justification of using a payment scheme altogether. The implication seems to be that government should consider three categories of landholders: those who are ready to carry out conservation works with little or no payment, or at any rate a payment below their opportunity costs; those who need to be paid at least their opportunity cost plus a rent; and those who will not enter into such agreements under any reasonable level of payment. From a budgetary point of view, this information can be of use to government, although it raises serious equity issues which cannot be ignored.

3) Again as an initial preliminary analysis, White and Burton (2005) were able to use data from the first round to benchmark the budgetary cost-effectiveness of the auction to that of an equivalent fixed-price scheme. They show that the cost-effectiveness of the ALR compared to that of a uniform price scheme varies between 315% and 207% in round 1 and 165% and 186% in round 2, depending on whether the fixed price scheme is input-based or output-based (see section 6.4 for details). They also show that comparing BushTender to an output-based scheme would considerably reduce the cost-effectiveness gains of 700% claimed by Stoneham et al. (2003). In addition, in a multi-round auction like the ALR (2 rounds), auction effectiveness can vary from round to round. This variability needs further research.

4) White and Burton (2005) conclude that "In general it is not possible to state if the uniform-price input scheme will generate lower or higher measures of efficiency for the auction: it appears to depend on the relative degree of heterogeneity in the opportunity costs, the environmental benefit , and the covariance between them." We would add: and the amount of bid shading by bidders.

5) They also suggest that "The choice of a counterfactual fixed-price scheme to measure auction efficiency should be guided largely by what is as a pragmatic alternative."

6) Fixed-price schemes are less vulnerable to rent-seeking than are auctions. In the ALR auction, there was evidence either of significant rent seeking or significant variations in the opportunity cost of labour (most likely the former).

7) In this case at least, there was no evidence to show that the auction imposed higher administrative costs than equivalent schemes using the same amount of information to underpin the selection process. This was because most of these costs were not linked to the specifics of running an auction, as indicated by the numbers that follow. About 70% of ALR costs were administrative costs, defined as all costs which are not payment transfers to farmers for on-ground works. They included all in-kind contributions, even if paid outside the project's budget. In addition, many costs were linked to the great distances involved and the remoteness of locations in rural Western Australia. Operational costs were estimated to total A$3291 per tender, split into A$1693 for variable costs and A$1598 for fixed costs (roughly a 50-50 split). Costs for Community Support Officers (on site) totalled A$1316 per tender and A$2055 per farmer. Research costs were estimated as A$1983 per tender and A$3096 per farmer. These costs only reflect project establishment: they do not include monitoring and compliance enforcement.

8) However, the amount available for payment transfers to farmers was only $200,000. It must be noted that the smaller the amount of transfer payments, the higher the proportion of administrative costs. This is a fixed-cost effect linked to the scale of the project. At the same time, such small scale 'pilot' projects represent a learning investment and part of an experimental investigation in a broad sense (see our general conclusions).

9) ALR administrative costs were incurred during five stages: scheme design, implementation, tender selection and evaluation. These stages appear to be common to most agri-environmental schemes (see Huylenbroeck and Whitby, 1999).

5.5 EcoTender (Australia)

Problem

EcoTender is an offshoot of BushTender and similar in intent to the Auction for Landscape Recovery. Like BushTender, it is carried out in the state of Victoria (Australia), and like the ALR, it broadens the scope of BushTender by aiming to secure multiple environmental benefits, including improvements in salinity control, biodiversity enhancement and water quality. It uses information from catchment-based modelling that can estimate both local and catchment-wide impacts on environmental outcomes as a result of changed land use and management. This reflects an educational goal, in that landholders are expected to discover how changes to native vegetation use and management can influence multiple environmental improvements.

Design

Under the EcoTender project, landholders are invited to submit bids, based on an agreed plan for improved native vegetation management and revegetation works on their properties. The design of the auction is similar to BushTender and the ALR, described above. However, the environmental benefit is defined somewhat differently. It is the sum of the biodiversity, saline land impact and in-stream water quality outcomes resulting from the proposed landholder management commitments, and estimated through the use of the catchment model. This allows the construction of an environmental benefit index ( EBI), which is combined with the cost represented by the farmer's bid in the following formula:

Successful bids will be those that offer the best value for money to the community, based on the amount of change in the environmental outcomes, the value of the assets affected by these changes and the cost. Successful landholders will receive periodic payments for their management actions under agreements signed with the Department of Sustainability and Environment.

Contracts are for a duration of 5 years, with an additional 5 years required for revegetation agreements. Fencing is required to remain in stock-proof condition for 10 years. Contracts come under common law. They are simply written and 3 to 5 pages long.

Bids can be lumped or separate; that is, a landholder can submit a bid for a number of areas or separate bids for each. (This constitutes a so-called combinatorial auction.) Pooled bids across several farmers are also allowed. Payments are annual, subject to satisfactory completion of agreed actions. Payments are not only input-based (management actions), but also include an output-based element: e.g. revegetation must meet certain performance standards relating to the number and diversity of plants established. In addition, an annual report must detail not only actions undertaken, but progress towards the agreed environmental targets. The regulator monitors a number of agreements each year. A breach may result in the cessation of annual payments.

This is an ongoing programme for which no results are yet available (2005).

5.6 Challenge Funds ( UK)

Problem

Challenge funding was introduced into Scottish forestry policy in 1997 - with the launch of the Grampian Challenge Fund and, a year later, the Central Scotland Challenge Fund. These funds operated under the umbrella of the Woodland Grant Scheme ( WGS) and offered additional grants to the standard WGS grants for extending the woodland area in specific geographical areas. They were competitive in that applicants were required to submit bids to the Forestry Commission ( FC) for this additional money. Under the Grampian Challenge Fund, the target was to plant 1,500 ha of new forest. No quantitative target was set for the Central Scotland Challenge Fund, but the aims here were to produce marketable timber, diversify land use and provide opportunities for countryside access and recreation. Both funds were closed for applicants in 2002 ( CJC Consulting, 2004).

There have been other Challenge Funds in Scotland, but currently only one challenge fund is in operation. This is the Woodlands In and Around Towns ( WIAT) Challenge - a new initiative aimed at stimulating sustainable management of woodlands near towns, thereby contributing to the regeneration of the urban environment and improving the quality of life for people living and working in urban areas ⁷ The challenge fund approach was used for WIAT to discover appropriate prices for hitherto un-priced goods and to ensure best value for money (Wright, FC Scotland, personal communication).

Design

Under the Grampian and Central Scotland Challenge Funds, landholders were asked to prepare and submit planting plans. Applicants had to pass a set of eligibility criteria relating to size, location and suitability of planting to deliver timber output. Bids were judged in terms of their value for money in relation to the aims of the Challenge Fund. A judging panel awarded grants to those applicants whose plans best met the aims of the Challenge and offered the best value for money.

Table 3 shows the bid scoring criteria for the Grampian Challenge Fund. The first four criteria relate mainly to costs and the likelihood of good establishment, the last four directly measure outputs. From these it is clear that four outputs are valued (recreation, landscape, habitat quality and productivity) with productivity given double the weight of the other three.

Table 3: Scoring system used for the Grampian Challenge Fund

Source: CJC Consulting (2004)

The panel selected high-scoring, low-cost bids first. Beyond that they traded off score against cost in a subjective way ( CJC Consulting, 2004). A similar scoring system was used for the Central Scotland Challenge Fund, with different weights given to the different outputs, and additional points being given for community involvement.

The WIAT Challenge Fund uses an alternative bid scoring method: Applicants are asked to rank their own proposals by answering (and scoring) a number of questions relating to the expected economic, social and environmental benefits of their proposed plans. ⁸ In the application documents, bidders are asked to consider that preference is given to proposals where the value for money is well demonstrated and arguments are logically and rationally presented. Applicants are also asked to support value judgments in proposals by evidence.

Results

The funds were very successful in rapidly expanding the land area under forestry and increasing harvestable timber output. Challenge funded planting increased the woodland area in both locations by almost 10%. Under the Grampian Challenge, operated for 6 years, 260 bids were received and 141 contracts awarded, extending the woodland area by over 3000 hectares - more than twice the target (Table 4). The Central Scotland Challenge, operated for 4 years, attracted 117 bids; 63 contracts were awarded. Bids ranged from £316 to £1,758 per ha in Central Scotland and £492 to £1,997 per ha in Grampian.

On average, the total payments in WGS grant aid (excluding Farm Woodland Premium Scheme) were £2,307 per ha in the Central Scotland Challenge Fund and £2,478 in Grampian (Table 5). The challenge top-up element was around 45% of the total grant aid in both areas.

Table 4: Areas planted under the Scottish Challenge Funds

Source: CJC Consulting (2004); Wright (personal communication, September 2005)

Table 5: WGS and Challenge Fund payments

Source: CJC Consulting (2004)

CJC Consultants carried out an economic evaluation of the two challenge funds in 2004. Respondents were asked whether they had deliberately tried to make their bids attractive to the Forestry Commission ( FC) given the competitive nature of the funding. With the exception of a few 'don't knows' all respondents answered 'yes'.

Garforth (2001)F undertook a study of challenge funding for the FC. He assessed the budgetary costs of each challenge fund as compared with a fixed-rate payment (a so-called location premium) set 20% below the bid price. The assumption was that applicants would have accepted a price 20% lower as a trade-off for the additional costs of bid preparation, the risk involved and some overbidding. His conclusions present a mixed picture: For the Grampian CF they were inconclusive, with a flat-rate grant calculated to bring in less land but at a lower cost. For the Central Scotland CF he concluded that a flat rate grant would have brought in the same area of land at lower costs!

CJC Consultants (2004) compare actual bids with a hypothetical flat-rate scheme. To secure all the bids with a fixed-rate premium would have required (with the benefit of hindsight) a premium of £1,758 in Central Scotland and £1,997 in Grampian. These were the highest bids received. Total payment would have been 36% higher in Central Scotland and 33% higher in Grampian. They conclude: "Such a fixed rate is clearly less efficient than challenge funding even allowing for some savings in transaction costs." (p. 63).

Some care has to be taken in interpreting these figures - and the conclusion. In computing the figures, the authors made the same mistake as Stoneham et al. (2003): They implicitly assumed that the bid curve ( i.e. the bids ordered from the lowest to the highest) represents the supply curve under a flat-rate scheme. We explained in section 5.3 why this is not the case: Bid curves contain an element of strategic overbidding and thus do not represent true opportunity costs! The study will thus have overestimated the efficiency gains of bidding.

Another important finding is this: FC staff reckon that operating the challenge funds took 20 per cent more staff time per application than fixed-rate incentives ( CJC Consultants, 2004). This makes the cost-effectiveness of the challenge funds appear questionable.

There has been no assessment of the WIAT Challenge Fund. There have been two bidding rounds, one in 2004 and another in 2005. FC received 38 bids in 2004 and 33 were successful. In 2005 there were 16 bids and 14 were successful (Wright, personal communication 2005).

Lessons learned

Douglas Wright of the Forestry Commission Scotland comments: "Benefits from our point of view are that it (challenge funding) ensures best value for money and it encourages good standards" (personal communication, 2005). This view is reinforced by the findings of CJC Consultants (2004, p. 67): "The challenge funds were an entirely appropriate incentive mechanism for delivering on the objectives of the schemes in the two forest areas. The challenges led to a rapid and substantial increase in planting. Additionality was very high and there was no evidence that bids were excessive. We conclude that the funds were a more effective and lower cost route for achieving substantial levels of planting in the two areas than fixed level grant aid."

On the other hand, one third of the respondents interviewed by CJC Consultants (2004) made comments that reflected a degree of dissatisfaction with a grant scheme based on tendering. The main comment was that it was 'unfair' in some way - either because neighbours received more cash or because it was 'hard to know what to bid' or because, in retrospect, entrants realised they had underbid. Some commented on the creation of animosity between neighbours. Others found it unfair because it suited those who could take the risks involved. ( CJC Consultants, 2004). Acccording to Douglas Wright, the main lesson is that because challenge funding is competitive there is less certainty for applicants. Unsuccessful applicants complain that there is a lot of work required in preparing bids. On balance, there was a consensus from stakeholders that challenge funds were too uncertain for the applicant and that they should be replaced by locational premia ( i.e. fixed rates per ha) although CF would be kept in the 'grant tool box'. As a result a new scheme was introduced in 2003, the Scottish Forestry Grant Scheme ( SFGS), replacing the former WGS and the two challenge funds. The FC has instituted locational premia of £2,000 per ha for farmed landscapes in Central Scotland and £1,500 per ha in Grampian. According to Douglas Wright, the Challenge Fund bids have informed the level of the locational premia.

5.7 Grassland Conservation Pilot Tender (Germany)

Problem

The focus of this pilot auction is on maintaining low-intensity grazing systems in the state of North Rhine-Westphalia, Germany. The conservation agency had initially offered fixed-rate payments. After only very few farmers had signed up, an auction was run to determine the excess payment required to induce broader participation. The objective thus was to establish whether an auction is a cost-effective mechanism to encourage broader participation in agri-environmental schemes. Funding was provided by the Ministry of Environment, Nature Conservation, Agriculture and Consumer Affairs of North Rhine-Westphalia. Three auction rounds were run between 2003 to 2005 in two counties (Holm-Müller and Hilden, 2004).

Design

Farmers were asked to specify in a sealed-bid process the amount of compensation the agency would have to pay them in excess of the fixed payment to induce compliance. A bid cap of 53 per cent of the fixed-rate payment was imposed for the first round. That is, bidders who demanded extra compensation in excess of 53 per cent of the standard rate were rejected. All bids below that reserve price were accepted. The reserve price was not known to the bidders before auction, but it was made public afterwards. The reserve price for the second round was set at 43 per cent of the standard payment (Holm-Müller and Hilden, 2004).

Results

In each of the bidding rounds, 15 landholders submitted bids covering an area of 335 and 353 hectares, respectively. First-round bids ranged from €47 to €150 per hectare, with an average bid of €92/ha. Second-round bids were lower on average (€46/ha), with a slightly wider range (€20 to €195/ha). Nine of the 15 landholders who had applied in the first round were accepted with a total area of 218 ha. Payments (in excess of the fixed rate) ranged from €47 to €80/ha (average €73/ha). In the second round, all 15 bids were accepted. The extra payment thus was €46/ha on average. Apparently, the reserve price (of 43% of the fixed payment) was not applied in the second round (Holm-Müller and Hilden, 2004). The third round was carried out in 2005. No results are available as yet.

Lessons learned

Fewer farmers than expected submitted a bid, implying that the scheme was not effective in encouraging broader participation in agri-environmental management. A survey was conducted after the second round to establish why farmers were not interested in the scheme although top-up payments had been offered through the auction. Many farmers said that, while the level of payment is an important factor, it is not the only factor determining their willingness to participate. Land was very scarce and lowering the intensity of their grassland was an option many could not afford. Uncertainty over the yield and energy losses from de-intensification was another deterrent to participation, as was uncertainty over the impacts of the latest CAP reform. A few farmers expressed general concerns about competitive bidding, feeling that such a mechanism was 'unfair'.

Interestingly, four of the rejected first-round bidders subsequently entered the scheme at the standard rate ( i.e. without top-up payment), indicating that their bids had not been based on marginal cost considerations.

The survey also tried to establish whether there had been prior consultation or even collusion among bidders. No such evidence was found. The survey revealed, however, that second-round bids were informed by first-round bids. Bids for the second round were revised downwards on average, with no bid exceeding the 53% bid cap imposed in the first round (Holm-Müller and Hilden, 2004).

5.8 Auction trial with outcome-based payment scheme (Germany)

Problem

The aim is to reward landholders for the provision of environmental services in the form of intensive cultivation crop retirement and replacement with grassland of high ecological and floral biodiversity quality. Payment is made not according to actions undertaken, but to the quality of grassland achieved. Grassland quality is categorised into three classes (I, II and III), with class III being the highest quality.

A specific aim of this program was to generate data on the transaction costs of farmers participating in such auctions. These costs include the time to gather information so as to be able to form a monetary valuation for the bid; the hiring of advice e.g. consultants; and the filling out of forms. Transaction costs were measured by surveying farmers who had participated in the auction. They were asked to fill in detailed questionnaires.

This project takes place in Lower Saxony, Germany (county of Northeim), funded by the Federal Ministry of Education and Research. (Rüffer, 2004)

Design

This is a budget-constrained (€30,000), sealed-bid, discriminatory price auction. The auction was held in 2004. The contract duration is for one year. The benefit criterion is represented by the grassland category and the cost by price per hectare. Performance assessments ( i.e. which grassland category has actually been achieved) are carried out by 31 July 2005, with payments made subsequently. This auction clearly is of an exploratory and even of an experimental nature (Groth, 2005).

Results

Out of the 38 farmers who submitted bids for a total of 199 plots covering 350 hectares of land, 28 farmers with 159 plots totalling 288 hectares were accepted for payments. The range of prices per ha requested by farmers are:

• Grassland I: 40 - 145 € (average 85 € /ha)
• Grassland II: 55 - 300 € (average 142 € /ha)
• Grassland III: 100 - 350 € (average 203 € /ha)

A comparison with a fixed price scheme has been done for class I grassland. Instead of 85 €, the envisaged uniform cost would have been 130 €/ha, 53% higher.

In terms of transaction costs, a farmer's valuation of his time for preparing a bid varied between 10 and 25 € per hour, with an average of about 17 € per hour. The average time spent being 4 hours, the average cost of preparing a bid was calculated to be about 67 €. This compares with an average payment per farmer of 875 € for grassland I (7.6%) and of 2090 € for grassland III (3.2%) (Groth, 2005).

Lessons learned

The output-oriented auction seems to have been popular with farmers and policy administrators alike. Groth (2005) concludes that the auction has worked well in that it has revealed a wide range of bid prices which gave the policy administrator much scope to select the most cost-effective producers.

It seems from this exercise that farmers' transaction costs are relatively small compared to payments made, but in this case, we have a very simple scheme based on a simple categorisation of output quality. This is unlikely to be representative of the majority of conservation schemes.

It would also have been instructive, given common practice to date and the experimental nature of this pilot, to include in the program a 'control group' paid on inputs (management actions) rather than outputs (grassland quality). This would have provided some data as to the relative merits of each approach in a very simple case.

5.9 Other conservation auction schemes

Auctions have also been used in areas of natural resource management other than agri-environmental policy. For example, contracts for the decommissioning of fishing vessels are usually allocated through competitive bidding: fishers are asked to nominate in a sealed-bid process the amount of compensation required for permanently removing their vessel from the fishery (Schilizzi and Latacz-Lohmann, 2005a, Latacz-Lohmann, 2001).

Auctions have been used to buy back water abstraction licenses from farmers in order to preserve minimal instream flows in rivers for environmental or recreational purposes (Laury, 2002). For example, the Flint River Drought Protection Act in the state of Georgia, USA, requires that state authorities use an "auction-like process" to pay farmers to suspend irrigation in declared drought years. The Environmental Protection Division is required to assess the risks of an upcoming drought and to determine consequently the number of hectares which must be taken out of irrigation to maintain acceptable instream flows. An auction is organised whereby farmers may offer to voluntarily forgo irrigation of all land covered by a specific water-use permit for the remainder of the cultivation year in exchange for a lump sum payment (Cummings et al., 2002). The development of such auction schemes to water restriction allocations is studied by Hailu and Thoyer (2005b).

« Previous | Contents | Next »