Numerical Maximisation of Likelihood: A Neglected Alternative to EM?

There is by now a long tradition of using the EM algorithm to find maximum‐likelihood estimates (MLEs) when the data are incomplete in any of a wide range of ways, even when the observed‐data likelihood can easily be evaluated and numerical maximisation of that likelihood is available as a conceptually simple route to the MLEs. It is rare in the literature to see numerical maximisation employed if EM is possible. But with excellent general‐purpose numerical optimisers now available free, there is no longer any reason, as a matter of course, to avoid direct numerical maximisation of likelihood. In this tutorial, I present seven examples of models in which numerical maximisation of likelihood appears to have some advantages over the use of EM as a route to MLEs. The mathematical and coding effort is minimal, as there is no need to derive and code the E and M steps, only a likelihood evaluator. In all the examples, the unconstrained optimiser nlm available in R is used, and transformations are used to impose constraints on parameters. I suggest therefore that the following question be asked of proposed new applications of EM: Can the MLEs be found more simply and directly by using a general‐purpose numerical optimiser?     

Pension plan solvency and extreme market movements: a regime switching approach

We develop and test a new approach to assess defined benefit (DB) pension plan solvency risk in the presence of extreme market movements. Our method captures both the ‘fat-tailed’ nature of asset returns and their correlation with discount rate changes. We show that the standard assumption of constant discount rates leads to dramatic underestimation of future projections of pension plan solvency risk. Failing to incorporate leptokurtosis into asset returns also leads to downward biased estimates of risk, but this is less pronounced than the time-varying discount rate effect. Further modifying the model to capture the correlation between asset returns and the discount rate provides additional improvements in the projection of future pension plan solvency. This reduces the perceived future risk of underfunding because of the negative correlation between interest rate changes and asset returns. These results have important implications for those with responsibility for balancing risk against expected return when seeking to improve the current poor funding positions of DB pension schemes.     

Optimal surveillance against foot‐and‐mouth disease: the case of bulk milk testing in Australia

Previous foot‐and‐mouth disease (FMD) outbreaks and simulation‐based analyses suggest substantial payoffs from detecting an incursion early. However, no economic measures for early detection have been analysed in an optimising framework. We investigate the use of bulk milk testing (BMT) for active surveillance against an FMD incursion in Australia. We find that BMT can be justified, but only when the FMD entry probability is sufficiently high or the cost of BMT is low. However, BMT is well suited for post‐outbreak surveillance, to shorten the length of time and size of an epidemic and to facilitate an earlier return to market.     

Assessing greenhouse gas emissions from electric vehicle operation in Australia using temporal vehicle charging and electricity emission characteristics

ABSTRACT

Significant interest exists in the potential for electric vehicles (EVs) to be a source of greenhouse gas (GHG) abatement. In order to establish the extent to which EVs will deliver abatement, however, a realistic understanding of the electricity and transport sector GHG emissions impacts arising from different approaches to integrating EVs into the power system is required. A key issue in this regard is the extent to which GHG emissions are a function of where and when EV charging will be enabled (or disabled) by the provision of recharging infrastructure and implementation of charging management strategies by the electricity industry. This article presents an investigation of the GHG emissions arising from electricity and gasoline consumption by plug-in hybrid EVs under a range of standard EV-power system integration scenarios. An assessment framework is presented, and GHG emissions from EV use are assessed for the New South Wales (NSW) and South Australian (SA) pools of the Australian National Electricity Market (NEM) using retrospective electricity system generation data for 2011. Results highlight that there is a range of possible outcomes depending on the integration scenario and emissions accounting approach used. This range illustrates value of a temporally explicit assessment approach in capturing the temporal alignment of electricity sector emission intensity and EV charging. Results also show the importance of a clean electricity generation mix in order for EVs to provide a GHG abatement benefit beyond what would be achieved by a hybrid (but non-plug-in) vehicle. The extent to which overnight charging in NSW is observed to produce higher emissions relative to unmanaged charging also illustrates a possible trade-off between GHG emissions and benefits for electricity industry from EV charging at times of low demand.     

Absorptive Capacity, Coauthoring Behavior, and the Organization of Research in Drug Discovery

We examine the interface between for-profit and publicly funded research in pharmaceuticals. Firms access upstream basic research through investments in absorptive capacity in the form of in-house basic research and 'pro-publication' internal incentives. Some firms also maintain extensive connections to the wider scientific community, which we measure using data on coauthorship of scientific papers between pharmaceutical company scientists and publicly funded researchers. 'Connectedness' is significantly correlated with firms' internal organization, as well as their performance in drug discovery. The estimated impact of 'connectedness' on private research productivity implies a substantial return to public investments in basic research.     

The economics of household waste management: a review

In recent years reducing the amount of waste generated by households has become an important policy issue in industrialised economies. It is no longer acceptable to discard waste without concern for environmental and natural resource issues. In an effort to reduce household waste various policy instruments such as kerbside charges, deposit-refund schemes, integrated sales tax exemptions and virgin material taxes, have been proposed and/or implemented. This article reviews the economics literature that has addressed household waste management. It is argued that a comprehensive modelling framework is necessary if the complex policy environment is to be accurately described.     

Targeting Monitoring Resources to Enhance the Effectiveness of the CAP

From its inception the operation of the CAP has required the compliance monitoring of farming activities and anticipated increases in the complexity of the monitoring tasks with the introduction of the SFP will stretch monitoring resources even further. We outline the scope for using targeting to enhance effectiveness of monitoring resources in dealing with non-compliant behaviour. Targeting resources can increase information on particular groups making identification of inappropriate behaviour easier. However, in reallocating monitoring resources towards the targeted sub-group, there is likely to be an associated reduction in resources allocated to monitoring non-targeted agents and therefore an increase in their incentive to cheat. To reduce this problem we suggest balancing the targeting-induced reduction in the probability of being caught cheating for non-targeted agents with an appropriate increase in the penalty for those agents if they are caught cheating. This increase in penalty results in an increased divergence between the outcomes from being caught or not caught cheating for non-targeted agents, which increases the overall perceived risk of the cheating option. And for risk-averse agents, this increase is a disincentive to cheating. On this basis we argue for greater use of target-based compliance monitoring as part of the revised and evolving CAP.