Skip to main content

Tom Kompas

University of Melbourne, Cebra, Faculty Member

Followers

76

Following

44

Co-authors

19

Public Views

Professor of Environmental Economics and Biosecurity at the University of Melbourne, working on optimal surveillance for early detection in biosecurity, the impact of climate change and large dimensional computational models in economics. Chief Investigator at the Centre of Excellence for Biosecurity Risk Analysis, School of Biosciences and School of Ecosystem and Forest Sciences.

less

Uploads

Papers by Tom Kompas

Productivity, Factor Accumulation and the Effects of Social Networks: Theory and Evidence

What vaccination rate(s) minimize total societal costs after ’opening up’ to COVID-19? Age-structured SIRM results for the Delta variant in Australia (New South Wales, Victoria and Western Australia)

PLOS Global Public Health

Using three age-structured, stochastic SIRM models, calibrated to Australian data post July 2021 ... more Using three age-structured, stochastic SIRM models, calibrated to Australian data post July 2021 with community transmission of the Delta variant, we projected possible public health outcomes (daily cases, hospitalisations, ICU beds, ventilators and fatalities) and economy costs for three states: New South Wales (NSW), Victoria (VIC) and Western Australia (WA). NSW and VIC have had on-going community transmission from July 2021 and were in ‘lockdown’ to suppress transmission. WA did not have on-going community transmission nor was it in lockdown at the model start date (October 11th 2021) but did maintain strict state border controls. We projected the public health outcomes and the economic costs of ‘opening up’ (relaxation of lockdowns in NSW and VIC or fully opening the state border for WA) at alternative vaccination rates (70%, 80% and 90%), compared peak patient demand for ICU beds and ventilators to staffed state-level bed capacity, and calculated a ‘preferred’ vaccination rate...

Optimal surveillance against bioinvasions: The sample average approximation method applied to an agent-based spread model

Trade-offs exist between the point of early detection and the future cost of controlling any inva... more Trade-offs exist between the point of early detection and the future cost of controlling any invasive species. Finding optimal levels of early detection, with post-border active surveillance, where time, space and randomness are explicitly considered, is computationally challenging. We use a stochastic programming model to find the optimal level of surveillance and predict damages, easing the computational challenge by combining the Sample Average Approximation approach and parallel processing techniques. The model is applied to the case of Asian Papaya Fruit Fly (PFF), a highly destructive pest, in Queensland, Australia. To capture the non-linearity in PFF spread, we use an agent-based model, which is calibrated to a highly detailed land-use raster map (50m×50m) and weather-related data, validated against a historical outbreak. We find that current surveillance levels are less than optimal.

Biofuels Subsidies and the Green Paradox

SSRN Electronic Journal, 2010

This paper develops sufficient conditions under which the Weak Green Paradox may (and may not) ho... more This paper develops sufficient conditions under which the Weak Green Paradox may (and may not) hold in terms of subsidies for biofuel production such that the supply-side responses by fossil fuel producers may more than offset the substitution to biofuels. Analytical results are derived and numerical simulations show that, under a wide range of parameter values, biofuel subsidies will increase the rate of extraction of fossil fuels in the short and medium term, and possibly bring climate-change damages closer to the present.

Rice land protection in a transitional economy: The case of Vietnam

Heliyon, 2021

Agricultural land protection (ALP) is a standard policy response to a desire for food security. H... more Agricultural land protection (ALP) is a standard policy response to a desire for food security. However, ALP may result in a misallocation of resources. Examining rice land policy in Vietnam, we determine the optimal level of rice land protected against other crops using a stochastic optimization model built on top of a general equilibrium framework, combined with sequential micro-simulations on household data. We find that converting part of protected rice land enhances economic efficiency. Nonetheless, the policy is relatively pro-rich, implying a trade-off between poverty reduction and economic efficiency, making some households in already poor areas worse off. Our approach can be applied to land-use planning generally, highlighting the relevant tradeoffs and the search for needed optimal land-use policies.

Optimal surveillance against bioinvasions: a sample average approximation method applied to an agent‐based spread model

Ecological Applications, 2021

Trade-offs exist between the point of early detection and the future cost of controlling any inva... more Trade-offs exist between the point of early detection and the future cost of controlling any invasive species. Finding optimal levels of early detection, with post-border active surveillance, where time, space and randomness are explicitly considered, is computationally challenging. We use a stochastic programming model to find the optimal level of surveillance and predict damages, easing the computational challenge by combining a Sample Average Approximation (SAA) approach and parallel processing techniques. The model is applied to the case of Asian Papaya Fruit Fly (PFF), a highly destructive pest, in Queensland, Australia. To capture the non-linearity in PFF spread, we use an agent-based model (ABM), which is calibrated to a highly detailed land use raster map (50m_50m) and weather-related data, validated against a historical outbreak. The combination of SAA and ABM sets our work apart from the existing literature. Indeed, despite its increasing popularity as a powerful analytical tool, given its granularity and capability to model the system of interest adequately, the complexity of ABM limits its application in optimizing frame works due to considerable uncertainty about solution quality. In this light, the use of SAA ensures quality in the optimal solution (with a measured optimality gap) while still being able to handle large-scale decision-making problems. With this combination, our application suggests that the optimal (economic) trap grid size for PFF in Queensland is about 0.7 km, much smaller than the currently implemented level of 5 km. Although the current policy implies a much lower surveillance cost per year, compared to the $2.08 million under our optimal policy, the expected total cost of an outbreak is $23.92 million, much higher than the optimal policy of roughly $7.74 million.

A practical optimal surveillance policy for invasive weeds: An application to Hawkweed in Australia

Ecological Economics, 2016

We propose a practical analytical framework which can help government agencies determine an optim... more We propose a practical analytical framework which can help government agencies determine an optimal surveillance strategy for invasive weeds, including cases of slow-growing or ‘sleeper weeds', and for all weeds at early stages of invasion where quantitative information is scant or rough. The framework consists of three key components: (a) a simple rule that can determine weed surveillance zones or where early detection is desirable, (b) a function that maps surveillance effort to early detection probability, and (c) a schedule to determine an optimal surveillance budget. A calibration to Hawkweed in Australia provides an example of the framework and shows that the optimal annual surveillance budget for this sleeper weed is substantial.

Capacity Reduction and Productivity: A Profit Decomposition for the Australian South East Trawl Fishery

Fisheries Buybacks

... The beneficial effects of the buyback and increased quota trading are in stark contrast to th... more

US biofuels subsidies and CO2 emissions: An empirical test for a weak and a strong green paradox

Energy Policy, 2014

Using energy data over the period 1981-2011 we find that US biofuels subsidies and production hav... more Using energy data over the period 1981-2011 we find that US biofuels subsidies and production have provided a perverse incentive for US fossil fuel producers to increase their rate of extraction that has generated a weak green paradox. Further, in the short-run if the reduction in the CO 2 emissions from a one-to-one substitution between biofuels and fossil fuels is less than 26 percent, or less than 57 percent if long run effect is taken into account, then US biofuels production is likely to have resulted in a strong green paradox. These results indicate that subsidies for first generation biofuels, which yield a low level of per unit CO 2 emission reduction compared to fossil fuels, might have contributed to additional net CO 2 emissions over the study period.

On marine reserves: rents, resilience and ‘Rules of thumb’

Using a bioeconomic model with two separate forms of uncertainty (a diffusion process and a jump ... more Using a bioeconomic model with two separate forms of uncertainty (a diffusion process and a jump process) a perturbation method is used to determine optimal reserve size for a harvested population with density-dependent growth. The results show that the benefits of reserves are understated in deterministic models and that, even with optimal harvesting, risk neutrality, a persistent population and without initial overexploitation, reserves have a positive economic value. Concavity of the value function with respect to optimal reserve size under uncertainty implies that, for many harvested populations, a positive economic return will arise from initially establishing small reserves, even when harvesting is optimal. The results also lead to general 'rules of thumb' when establishing reserves. Namely, that optimal reserve size rises with (1) the greater the rate of transfer from the reserve, (2) the greater the magnitude or the likelihood of negative shocks, (3) the more the actual harvest exceeds the optimal harvest, (4) the lower the intrinsic growth rate, (5) an increase in price or demand provided the population is between zero and its maximum yield, (6) an increase in marginal harvesting costs provided the population is between its maximum yield and carrying capacity and (7) the lower the discount rate.

Health and Economic Costs of Early, Delayed and No Suppression of COVID-19: The Case of Australia

We compare the health and economic costs of early (actual), delayed and no suppression of COVID-1... more We compare the health and economic costs of early (actual), delayed and no suppression of COVID-19 infections in 2020 in Australia. Using a fit-for-purpose compartment model that we fitted from recorded data, a value of a statistical life year (VSLY) and an age-adjusted value of statistical life (A-VSL), we find: (1) the economic costs of no suppression are multiples more than for early suppression; (2) VSLY welfare losses of fatalities equivalent to GDP losses mean that for early suppression to not to be the preferred strategy requires that Australians prefer more than 12,500–30,000 deaths to the economy costs of early suppression, depending on the fatality rate; and (3) early rather than delayed suppression imposes much lower economy and health costs. We conclude that in high-income countries, like Australia, a ‘go early, go hard’ strategy to suppress COVID-19 results in the lowest estimated public health and economy costs.

Epidemiological modelling of the health and economic effects of COVID-19 control in Australia’s second wave

Journal of Public Health, 2021

Background We investigated the public health and economy outcomes of different levels of social d... more Background We investigated the public health and economy outcomes of different levels of social distancing to control a ‘second wave’ outbreak in Australia and identify implications for public health management of COVID-19. Methods Individual-based and compartment models were used to simulate the effects of different social distancing and detection strategies on Australian COVID-19 infections and the economy from March to July 2020. These models were used to evaluate the effects of different social distancing levels and the early relaxation of suppression measures, in terms of public health and economy outcomes. Results The models, fitted to observations up to July 2020, yielded projections consistent with subsequent cases and showed that better public health outcomes and lower economy costs occur when social distancing measures are more stringent, implemented earlier and implemented for a sufficiently long duration. Early relaxation of suppression results in worse public health out...

Budgeting and Portfolio Allocation for Biosecurity Measures (Hawkweed, FMD, RIFA, PFF)

Catch, Efficiency and the Management of the Australian Northern Prawn Fishery

This paper is a study of the production technology and relative efficiency of vessels harvesting ... more This paper is a study of the production technology and relative efficiency of vessels harvesting banana and tiger prawns in the Northern Prawn Fishery (NPF), one of Australia's largest and most lucrative fishing areas. It is based on an unbalanced panel data set of 226 observations among thirtyseven vessels for the years 1990-1996 and employs a technique which specifies a stochastic frontier production function in order to decompose the variation among vessels in the harvest of prawns due to unbounded random effects beyond firm control from those that result in differences in technical inefficiency among fishing vessels in the industry. In other words, variations in maximum expected output can occur either as a result of stochastic effects (e.g., good and bad weather states), or from the fact that vessels in the industry may be operating at various levels of inefficiency due to mismanagement, poor incentive structures, less than perfectly competitive behaviour or inappropriate input levels or combinations. Estimation of this output frontier also provides key information on the relative importance of inputs in the production of banana and tiger prawns, output elasticities, returns to scale, possible variations in stock size and the economic performance of each fishing vessel, year to year. Likelihood ratio tests confirm that both stochastic effects and the extent of technical inefficiency matter, thus making traditional OLS estimates inappropriate. The level of technical inefficiency is shown to depend positively on gear headrope length and negatively on either the number of A-units or fuel expenditures. The point is especially relevant since A-unit restrictions over vessel size and engine power in the fishery during this period appear to have resulted in a substitution toward less efficient but unregulated inputs, such as gear headrope length. In this regard, the recent introduction of gear headrope length restrictions may be justified on two counts: both as a device to limit effort or catch and protect prawn stocks and as a way, given the final estimates in this paper, of improving economic performance by increasing the technical efficiency of vessels remaining in the industry. Nevertheless, it is important to emphasize that restrictions on an existing inefficient input may result in far smaller reductions in effort than projected, since the technical efficiency of vessels in the fishery will rise. With an increase in technical efficiency, gearrestricted fishing firms will harvest at points closer to their output frontiers. Moreover, with the removal of A-unit restrictions, 'effort creep' in the form of larger vessels and more powerful engines may more than compensate for any decrease in effort due to gear reduction. Acknowledgments: Paper presented at the AARES conference, February 2002. The authors would like to thank Steve Beare (ABARE), Janet Bishop (CSIRO), Bill Venables (CSIRO) and Peter Gooday (ABARE) for data access and very helpful comments on this work and the nature of the Australian northern prawn fishery in general. The views expressed herein are those of the authors and do not necessarily reflect the views of ABARE or ANU.

The value of information in biosecurity risk-benefit assessment: an application to red imported fire ants

Policy makers are confronted daily with uncertainty, especially in complex areas like biosecurity... more Policy makers are confronted daily with uncertainty, especially in complex areas like biosecurity. One way to improve decision-making and reduce uncertainties is to collect more information. Information is costly – whether the value of improved decision-making justifies the cost is a fundamental question facing policy makers. This paper addresses that question by making three practical contributions for binary choices (such as whether to implement or forego a particular policy). First, it analyses the determinants of the value of information, and how that value changes with features of the problem. Second, it uses this analysis to derive simple rules of thumb which provide upper bounds on the value of additional information. Third, it provides a practical application of the value of information in deciding whether to attempt eradication of the red imported fire ant.

Non-Market Values and Optimal Marine Reserve Switching

A stochastic bio-economic model is constructed to analyse the effects of marine reserve 'swit... more A stochastic bio-economic model is constructed to analyse the effects of marine reserve 'switching' between a 'no take' area and a harvested area. The model accounts for both market and non-market values of the fishery. Estimated parameters from the red throat emperor fishery from the Great Barrier Reef are used. Simulations show that an optimal switching strategy is, under a range of scenarios, preferred to fixed reserve and no reserve strategies. An important outcome is that the non-market values associated with the size of the fishery substantially affect both the returns from switching and the closure time.

A Rule of Thumb for Controlling Invasive Weeds: An Application to Hawkweed in Australia

We use a bang-bang optimal control model to derive a rule of thumb for an optimal management of i... more We use a bang-bang optimal control model to derive a rule of thumb for an optimal management of invasive weeds, in terms of the marginal benefits and costs of various control actions. Instead of determining the size of infestation under an optimal surveillance measure, the rule specifies the types of land where an invasive weed should be first prevented from establishment, and under what conditions control should be initiated. The types of land are modeled via the heterogeneous vulnerability of land to the weed and likely infestation. This easy-to-use rule is applied to determine how hawkweed should be controlled in Australia, across three potential control strategies: containment, eradication and no action. We investigate this rule-of-thumb in both deterministic and stochastic settings. With uncertainty, when calculating the threshold of when and how to act, we take into account the fact that delaying a control action will incur not only larger damage and a potentially larger sprea...

A Comparison of Parametric Approximation Techniques to Continuous-Time Stochastic Dynamic Programming Problems

We compare three parametric techniques to approximate Hamilton-Jacobi-Bellman equations via unidi... more We compare three parametric techniques to approximate Hamilton-Jacobi-Bellman equations via unidimensional and multidimensional problems. The linear programming technique is very efficient for unidimensional problems and offers a balance of speed and accuracy for multidimensional problems. A comparable projection technique is shown to be slow, but has stable accuracy, whereas a perturbation technique has the least accuracy although its speed suffers least from the curse of dimensionality. The linear programming technique is also shown to be suitable for problems in resource management, including applications to biosecurity and marine reserve design.

An Optimal Surveillance Measure Against Foot-and-Mouth Disease in the United States

Surveillance programs on farms and in the local environment provide an essential protection again... more Surveillance programs on farms and in the local environment provide an essential protection against the importation and spread of exotic diseases. Combined with border quarantine measures, these programs protect both consumers and producers from major health concerns and disease incursions that can potentially destroy local agricultural production and supporting industries, as well as generate substantial losses in trade and tourism. However, surveillance programs also impose costs in the form of expenditures on the surveillance program itself, along with the costs of disease management and eradication should an incursion occur. Taking border quarantine expenditures as given, this paper develops a stochastic optimal control model (with a jump-diffusion process) to determine the optimal level of surveillance activity against a disease incursion by minimizing the present value of the major direct and indirect costs of the disease, as well as the cost of the surveillance and disease ma...

od Today and None Tomorrow: The Economic Value of a Marine Reserve

Using data from what was once one of the world's largest capture fisheries the economic value of ... more Using data from what was once one of the world's largest capture fisheries the economic value of a marine reserve is calculated using a stochastic optimal control model with a jump diffusion process. The results show that with a stochastic environment an optimal-sized marine reserve can generate a triple payoff that (a), raises the resource rent even when harvesting is 'optimal', (b) decreases the recovery time for the biomass to return to its former state and smooths fishers' harvests and resource rents, and (c), lowers the chance of a catastrophic collapse following a negative shock.