Creating Public-Private Partnerships in Intensive Forest Management

Paul Comeau[1], Laurie Gravelines and David Lewis


Despite the potential of Intensive Forest Management (IFM) to deliver higher per-hectare timber yields, and to allow the transfer of more forest land from commercial to recreational and wilderness uses, risk and uncertainty inhibit the rate of growth in the use of IFM techniques. This paper presents a newly developed process by which government and industry can join forces to find win-win risk-sharing partnerships that accelerate the pace at which IFM technologies and procedures are brought into practice. The new process rests on the premise that neither the public nor the private sector can or should modify its fundamental objectives in deference to the profit or conservation goals (respectively) of the other. Instead, the process is one in which the two sides individually invest in IFM while sharing investment risk in such a way as to improve the position of each party in relation to its particular objectives. The process involves methods of conventional and "Bayesian" (subjective) probability as a means of facilitating consensus on matters of scientific and commercial uncertainty and acceptable risk.


"After you Alfonse" is the classic line from a play in which a pair of intrepid treasure hunters chivalrously bow back and forth, each insisting on the other’s privilege to be first into a dark and shadowy house long celebrated for the mystery and uncertainty of its hidden secrets. What lies within? Unimaginable wealth? Sudden and horrible death? Just a lot of dust?

As a metaphor for intensive forest management (IFM), After you Alfonse pretty much describes the dilemma of the public sector (explorer one) in search of risk-free environmental, conservation, and economic benefits, and the private sector (explorer two) in search of risk-free shareholder returns. A game theorist might describe the dilemma as two players locked in a game whose rules inevitably lead them to an outcome that serves the objectives of neither. In the game of IFM, governments play to liberate forestlands to wilderness and recreational purposes while absorbing zero risk of either subsidizing or otherwise indemnifying industry against the financial risk of diminished harvestable timber supply[2]. Private firms play the IFM game to reap the prospective rewards of stronger future yields with zero risk of present-day reductions in allowable cut. The game thus endows both players with the certainty of costs now and the uncertainty of benefits later. It is not surprising therefore that IFM expenditures exhibit anemic long-run growth and, in some jurisdictions, a distinct downward trend (see Figure 1).

Figure 1: IFM expenditures (’000$) on Crown and private land 1985 - 1996.

A New Game: Investment-based Risk Sharing

As any game theorist will attest, collusion is one means of resolving a dilemma such as IFM outlined above. With collusion, and no cheating, play can yield outcomes that leave both sides better off. In the real nexus of free markets and international trade agreements, however, neither side can enter into a collusive relationship, not in the absence of a third- party empowered to govern the rules of fair play - and probably not even then. Closely monitored international treaties and the days of all-powerful third-party domestic regulators long-gone mean that we must find another way. Some observers flirt with the idea of "enlightened private self-interest" whereby progressive corporate management is encouraged to modify its near-term profit-maximizing behaviour in deference to the benefits of long-run conservation and environmental sustainability. While "good corporate citizenship" has its place, it imposes on industry an "ethical overload" that inevitably burns off any serious move to modify the risk-reward relationship that best serves the interests of a company’s lenders (banks) and equity shareholders. IFM suffers accordingly.

The method of resolution we discuss in this paper, Investment-Based Risk Sharing (IBRS), rests on the premise that neither the public nor the private sector can or should modify its fundamental objectives in deference to the profit or conservation goals (respectively) of the other. Instead, the game is one in which the two sides individually invest in IFM while sharing investment risk in such a way as to improve its own position as compared with the no-investment alternative. Some call this approach "public-private partnering," a catchy but misleading phrase unless we recognize that the term "partnership" does not imply a sharing of objectives, only a sharing of risk and rewards of co-production. We also prefer the "IBRS" rubric because the sharing of risk does not imply the sharing of cost (i.e., subsidy). Subsidy is a mode of public policy rejected by most domestic governments and forbidden in most bilateral and multilateral international trade agreements.

The essence of IBRS is illustrated at Figure 2. Each forest management regime "option" represents the combination of two complementary but independent sets of IFM investment initiatives by industry and government, respectively. Return on investment is measured from the unique perspective of each party. In the case of industry, returns are measured in terms of firm-level productivity gains, cost avoidance or other direct financial gain to shareholders, net of the firm’s life-cycle investment cost. In the case of government, returns are measured in terms of public policy outcomes[3], principally non-timber values, net of the government’s life-cycle investment costs. Non-timber values might include public benefits such as:

Under Option 1, in which industry invests $30 million and government zero, private returns, after adjusting for risk, lie at three percent, an insufficient margin by most benchmarks to warrant execution. Result: No IFM. Under Option 2, the government considers investing $5 million of the Option 1 IFM initiative, reducing industry’s risk exposure to $25 million. This improves the private risk-adjusted rate of return to 10 percent. From a government perspective, the $5 million investment is worthwhile (the 14 percent return exceeds most public sector investment hurdle rates). For industry however, the outcome, at 10 percent, is still marginal. Result: up to $5 million of IFM investment, but only to the extent that those sub-initiatives are mutually exclusive and amendable to parsed implementation.

Under Option 3, the government ups its investment to $10 million, reducing the industry’s risk exposure by a further $5 million. Under this risk-sharing arrangement, both sides stand to gain risk-adjusted rates of return in the order of 20 percent, well in excess of benchmark hurdle rates in both sectors. Result: fully $30 million of IFM investment and a win-win "ROI" proposition for the public, for lenders, and shareholders alike. The outcome might be described as a "partnership," but only in the sense that, through a cooperative examination of each other’s prospective risks and rewards, the two sides have found a risk-sharing strategy that maximizes the self-interest of its stakeholders, and its stakeholders alone.

Figure 2: Government-Industry Risk-Sharing Template: The Search for a Win-Win Value Proposition



Expenditure / Investment

Return on Investment

Value in Savings And Benefits

Risk Adjusted ROI

Expenditure/ Investment

Return on Investment

Risk Adjusted ROI

Investment Option 1








Investment Option 2








Investment Option 3








Getting to Yes in IBRS: The Role of ForestRAP[4]

Finding resolution, "partnership," within the framework depicted at Figure 2 demands cooperation and consensus, but not in relation to objectives. As we have seen, IBRS recognizes that the disparate goals of government and business are not up for compromise. Where cooperation and consensus must occur is at the informational level. In short, each side must believe that the numbers are right and that neither player is manipulating quantitative risk profiles in order to shift more exposure to the other. The Ontario Ministry of Natural Resources (MNR) is working with a process called "ForestRAP" (for Forest Risk Analysis Process).

How ForestRAP Works

Three factors underpin a sound ForestRAP process:

Organizing Public-Private Deliberations for Scientific Transparency and Flexibility

Although the consensus literature has long proclaimed the importance of openness and flexibility, these principles are rarely extended to the scientific level. Yet this is the level at which the seeds of mistrust and indecision are sown when it comes to risk sharing among diverse parties to a high-stakes investment decision. To address the problem, ForestRAP insists on openness and flexibility in relation to:

i. Scientific models; and
ii. Consensus-Based probabilistic assumptions.

(i) Open Scientific Models. In a departure from traditional forecasting procedures, ForestRAP insists that scientific models are not simplified for convenience of quantification or known versus unknown empirical knowledge. Outcome variables (such as growth and yield curves and non-timber values) and their hypothesized causal factors are diagramed non-mathematically to reveal to non-scientists the internal structure and logic of IFM science. In contrast to traditional model representations, these "structure and logic" diagrams do not limit the forecasting framework to relationships requiring mathematical representation or empirical observation. This allows the introduction of political risk factors and other "soft" realities that often elude traditional modeling. The only restriction is that any factor to be included can, in principle, be assigned a probability distribution, if subjectively so (see below).

(ii) Consensus-Based Probabilistic Assumptions. The structure and logic diagrams serve to facilitate panel sessions organized to elicit expert and public-private stakeholder beliefs about model logic and numerical assumptions. For each variable and coefficient (numerical expressions of causal linkages) identified in the model, panelists provide ranges, or probability distributions, that characterize uncertainty about them. To those unfamiliar with probability and statistics, this task may sound onerous. However new techniques and software programs are designed specifically to make the application of probability analysis accessible and user friendly. The MNR employs software (also called "groupware") that enables a panel to visualize the implications of any particular supposition about the probability range for a particular variable or relationship. This facility sharply accelerates the achievement of technical consensus.

Blending Objective and Subjective Data

Where do the judgments about probability come from? Wherever possible, the starting point is empirical, scientifically obtained frequency data. Of equal importance however are subjective probabilities obtained within the "Bayesian," or subjectivist paradigm[5]. Whereas a "frequency-based" probability represents the incidence of observed outcomes, a subjective probability is the degree of belief held by an informed person that a given outcome will occur. The use of subjective probability in ForestRAP blends the subjective beliefs of experts and stakeholders with objective, scientific knowledge.

ForestRAP stakeholder panels are facilitated by an independent third-party who employs the computer groupware referenced earlier to elicit subjective probabilities from panelists for each factor in the structure and logic model. A key attribute of the facilitation process is that stakeholders are never drawn into a debate about who is right and who is wrong. Extreme views may be assigned low probabilities, but this is wholly different from impugning an individual’s view as being unworthy of consideration. Stakeholders may be expected to present technical arguments that differ sharply from the mainstream but are not provably incorrect. In dismissing one view while accepting another, traditional forecasting approaches foster polarization and encourage divisive and unproductive debate. ForestRAP, on the other hand, embraces any reasoned view, albeit with different degrees of probability.

The elicitation of subjective probabilities is thus no mere roundup of arbitrary judgments. To be legitimate, a subjective probability must obey the same axioms as frequency-based probability. When an expert argues that a given assumption has a 30 percent probability of occurring, he/she must assign a probability of 70 percent to the probability non-occurrence. Using subjective probability, ForestRAP can thus shed quantitative light on IFM scientific uncertainties that resist traditional methods of quantification, not because of limited expert knowledge, but because of limited field research.

Accounting for Simultaneously Occurring Risks

With consensus regarding scientific logic and factor probabilities in-hand, simulation software allows all elements of uncertainty and risk to vary simultaneously (some with non-zero covariance) so as to yield the probability distribution for both the public and the private ROI in relation to a given IFM investment option. This is very different from the traditional approach to forecasting in which individual factors are varied by arbitrary amounts, one at a time, in a series of "what-if" scenario experiments. "What-if" experiments do not prompt consensus because they do not represent reality. What ForestRAP provides is the probability range for public and private ROI and net benefits under the real-world reality that everything will depart simultaneously from our best scientific assumptions according to non-arbitrary probabilities. This approach builds consensus by appealing to stakeholders’ commonsense conviction that no one can be sure of the future; that assumptions do not veer from forecast values one at time; and that the best one can hope for is an engaged and systematic risk assessment of what might happen. With ForestRAP, we hope that stakeholders in IFM investment will get beyond scientific debate and move on to the deliberation of policy and decision on risk-sharing IFM strategies.

ForestRAP is no magic bullet. The development of more intensive forms of forest management has been plagued by a lack of hard scientific data (although great strides have been made in growth and yield modeling on sample plots for example) in the presence of extensive expert knowledge in the field of forestry. When combined with highly controlled experimental data, the use of subjective probability often yields wider ranges for variables of interest. Yet this reflects the reality of what is known and unknown and as such creates a basis for honest discussion and consensus on the nature and significance of the risks and how to share them. The resulting risk profile thus reflects both the state of knowledge and science at any given time and permits each party to invest in IFM productively and reasonably. Furthermore, by conducting multi-dimensional sensitivity analyses using the risk analysis model, factors that drive economic and commercial outcomes can be ranked so as to maximize the value of new information and thereby guide further research efforts.

Piloting ForestRAP

In collaboration with Tembec Inc. and the Forestry Research Partnership, MNR is piloting Investment-Based Risk Sharing by employing ForestRAP as part of the "Northeastern Ontario Enhanced Forest Productivity Study." The study is a two-year investigation into the effectiveness and business case for alternative methods of silviculture through intensified forest management. The facilitated ForestRAP panel process is being applied to a wide range of scientific models and risks, including growth and yield forecasting, non-timber values and the entire range of IFM input costs. The results are to be widely reported and employed in helping formulate the Ministry’s long-term approach to silviculture in the context of conservation and environmental policy.


HLB Decision Economics Inc. A Business Case Framework for Intensified Forest Management on Crown Land, Report prepared for Ontario Ministry of Natural Resources, November 2001.

[1] Vice President, HLB Decision Economics Inc., Suite 500 - 1525 Carling Avenue, Ottawa, Ontario K1Z 8R9, Canada. Tel: (613) 234-7575; Email: [email protected]; Website: www.hlbde.com
[2] The government’s strategy is further compounded when public lands are integral to the support of existing forest products industry in largely single industry communities.
[3] Including, in the case of public lands, government as owner and beneficiary of a commercial timber producing forest.
[4] ForestRAP was developed by HLB Decision Economics for the Ontario Ministry of Natural Resources under the "1999 Ontario Forest Accord."
[5] Bayes was an 18th century mathematician who gave rise to the subjectivist school of statistical analysis. Bayesian statistics is a mainstay of analysis in various fields, including pharmaceuticals and bio-medical research.