Home » Posts tagged 'risk' (Page 2)

Tag Archives: risk

IDRC Davos 2012

Tomorrow I’m off to the International Disaster Risk Conference 2012 in Davos, Switzerland.

I’ll be volunteering there and checking out as much of the latest research and best practice in disaster risk management as I can. It’s particularly encouraging to see sessions on cascading mega disasters, urban risks, the future of risk management and broader governance approaches in a post-Hyogo Framework for Action environment.

I hope to post a couple of updates during the conference and a longer recap on my return to Kathmandu. In the meantime check out the conference website or follow them on twitter.

Emergency Services Levy: Towards a risk based approach?

The New South Wales Government has recently released a discussion paper on changes to the way the emergency services are funded in the State. Fire and Rescue NSW, the NSW Rural Fire Service and the NSW State Emergency Service currently cost about $1billion a year to run. The current model sees contributions from the insurance industry, local government and the state government towards the funding of the three services. The insurance industry contributes to 73.7% of their total budget.

This state of affairs has been widely acknowledged as being inefficient, inequitable and counter-productive. Insurance taxes have been widely acknowledged to reduce rates of insurance. Un-insurance rates in NSW are some of the highest in the country. A move towards a property based level is strongly recommended in the discussion paper which is asking for community views on the design.

Most of the questions relate to how a fair and efficient property based tax can be levied. I wish to focus on a different aspect: If the current levy disincentivises risk management measures (ie. insurance) could the future levy be designed to incentivise risk management measures?

In theory the levy could be based on a service delivery model (which do exist for the various services). However any decision to do so would be primarily based on equity grounds – this does raise the tricky question of whether you would charge on the basis of the service provided (which are generally lower in rural areas) or the cost of providing the service (which are generally higher in rural areas).

The paper dismisses the use of a risk based approach to determine a property levy as impractical, but I think it deserves closer examination. How could a risk based approach work? Would it be feasible to implement? And would it actually lead to risk reduction?

Property risk comes from two sources – the site of the property (ie. its exposure to hazards) and its construction/use (ie. its vulnerability to hazards). For hazards like urban fire and hailstorm the site doesn’t matter too much. Other hazards, such as flood and bush fire depend on both building construction and location.

Exposure to hazards

The component of the risk attached to the site of the property is very difficult to move. It’s unclear what the effect of a site-based risk levy would be. On one hand, a higher levy could increase property prices and lead to more wealthy folk moving into the area. On the other hand higher rates for disaster prone properties combined with high insurance premiums, could just end up reducing the income and wealth of those who are disaster-prone, increasing their vulnerability.  Although there has been substantial research on the effect of flooding on property prices, there remains substantial methodological flaws in much of this research and more work still needs to be done. Changes to insurance costs are not a good indicator, if the prices get too high, consumers can just opt-out.  And council rates are generally based on property values, not the other way around, so they too make a poor guide of how individuals would respond.

What about the construction of the property? Here retrofitting, property maintenance, the installation of safety systems and other practices (particularly for commercial and industrial sites, which I will leave out in this discussion as they’re already highly regulated and only represent a small portion of the overall risk) could make a difference.

As the list of potential anthropogenic and natural hazards is rather long I’m going to concentrate on the major ones our emergency services respond to: urban fire, bush fire, hail and windstorm and flood.

Urban Fire

Around two-thirds of all residential fires in NSW (in 2006/07, the most recent year for which data is available) were caused by some form of human action, whether through negligence, misadventure or malice. Only 8.37% of fires were due to short circuits and other electrical failure, but this is the leading cause of fires due to equipment and design failures (which make up 15% of all residential fires).

These causes are reflected in the majority of home fire advice which relates to individual and family preparedness. Additional guidance is targeted at a few electrical items such as Halogen down-lights, but even this advice is heavily weighted to maintenance and inspection. Attempting to incentivise good behaviour through behaviour based discounts to a property levy would be virtually impossible to implement (at least without your house into some sort of AI surveillance machine to make sure you’re doing things right). With few structural and non-structural measures that can be implemented to reduce fire risk that leaves fire response equipment.

Smoke Alarms are already mandatory in NSW, so no additional incentive is needed to install them. However Fire and Rescue NSW does recommend the installation of Home Sprinkler Systems, which have been proven to be very effective at preventing deaths in house fires. Given that this is a very specific measure the best form of incentive may be some form of rebate on the installation of home sprinkler systems, rather than discounts to a property levy.

Bush Fire

As with urban fire much of the potential for mitigation of bush fire relates to maintenance and inspection. However design and construction are just as important. In NSW new dwellings and renovations must comply with Planning for Bush Fire Protection. However much of the risk is associated with existing dwellings in bush fire prone areas. There is the potential to retrofit existing dwellings using the construction methods outlined in this guide published by the Victorian Country Fire Authority. Some simple measures on existing dwellings such as installing steel screens on doors and windows, enclosing under-house spaces, fitting a rooftop sprinkler system and installing gutter guards are relatively cheap and practical for almost all dwellings.

As the methods vary depending on fire hazard and existing construction – a broad incentive could be effective in letting home-owners select their own retrofitting methods. A risk rating and scoring system for various mitigation measures could form modifications for a property based levy. The number of bush fire prone properties is only a fraction of the total building stock in NSW, a carefully targeted and capped levy ‘surcharge’ with reductions based on some sort of checklist of mitigation options might work – it needs further investigation.

Hail and Wind

Roof damage due to hail and windstorms (I’m going to ignore water ingress due to poorly maintained gutters and downpipes – another maintenance issue) is one of the major areas of response for the SES and a significant cause of property damage from natural disasters.

There are a variety of trade-offs when it comes to different roof types including cost, longevity, maintenance needs, damage potential and ease of repair. Corrogated Fibro is probably the worst roofing material due to its fragility and toxic nature. However I haven’t found any quantitative research on the costs and benefits associated with different roofing materials.

Roughly three quarters of all new dwellings are constructed with tiled roofs with sheet metal and slate making up the majority of the other quarter. Tiles and slate are the most vulnerable roofing materials to hail as shown in the following table:

Hailstone Size (cm) Damage
3.0–4.0 Glass and plastic roofing broken
4.0–5.0 Old slate 100+ years old, Old tiles 50+ years old, cracked
5.0–6.0 Old slate tiles broken, new tiles crack
6.0–7.5 New concrete tiles and terracotta tiles break
7.5–8.5 Sheet metal dented – all other roofing broken
8.5–9.0 Sheet metal dented – all other roofing smashed
>9.0 Sheet metal roofing penetrated/ cracked

Clearly sheet metal is the most hail resistant form of roofing, and if appropriately tied can be remarkably wind resistant too. Roofs are most likely to be replaced during renovations or at the end of the roof lifespan, or after damage. Although it’s unclear what the ‘best’ material is, incentives could encourage people to replace roofs (or build with a better roof in the first place) if a levy presents a clear cost disincentive when keeping the previous material. Gathering data by visual inspection or even remote sensing should be relatively affordable and not need to be repeated too often.

Flood

If a property is to be built or renovated in a flood prone area (noting that in some areas there can be a substantial difference between a 1% AEP flood, the usual residential building standard, and the probable maximum flood) there are a number of design and construction features that can substantially reduce flood damage. These include using flood compatible materials, installing the electrical wiring higher than normal and considering the use of ground floor rooms.

For existing dwellings some of these methods can also be employed, or there are other means including a variety of temporary flood protection products that can be retrofitted on existing buildings (usually designed to keep <1m of slow moving water out of the house).

The most expensive method is house raising. There are funding sources for the raising of houses in some jurisdictions, including NSW, however these are generally only in areas with approved programs. Anecdotally there is demand for funding for these projects outside of these areas. A levy reduction could help people pay for this type of project, however I’d argue that making funds available for any property that meets a particular criteria would be a more effective method.

Although there has been substantial work done on cost benefit of around flooding, I haven’t seen any work specifically relating to property based incentives/disincentives for these measures. As with bush fire and storm, plenty of potential for mitigation – but uncertainty if a property levy approach would work.

Conclusion

There is some work happening in building design and construction in relation to climate change, including looking at regulatory mechanisms to enhance mitigation and retrofitting and incentives. However more work is needed to better understand the risk, effectiveness of retrofitting measures and community response to incentives.

The Insurance Council of Australia commissioned Deloitte to investigate a number of models for a property based levy. It did propose a risk based model calculated on an LGA-wide basis with the risk measure based on previous fire incidence in that LGA. This would not create significant differences for individual properties to create incentives. It does recommend that additional research could be conducted, and models prepared to investigate a levy raised on a per-property basis.

With significant existing risk associated with natural disasters in Australia and new development increasing risk too, all possible measures to encourage property mitigation need to be examined. A risk-based emergency services levy with discounts for mitigation could be expensive and difficult to implement, and may not encourage enough mitigation to balance out the cost of the approach. However where this balance lies has not been established by research – I think it is worth further detailed investigation.

The Census and Emergency Management

Yesterday saw the release of the first batch of data from the 2011 Australian Census and my inner statistics nerd was very excited. But the census is not just for demographers and statistics nerds – it has real practical uses for emergency managers. Here’s just a few areas where it comes in handy:

Risk Management

Risk is often characterised as Hazard X Vulnerability. Hazard is generally easy to determine and there are substantial studies on natural hazards in particular. Vulnerability can be much more difficult to determine, especially when talking about people. What makes someone vulnerable will ultimately depend on a host of complicated factors and relationships, but demographic data can provide some coarse indicators.

A description of a community and its vulnerabilities is grounded in solid demographic data, the best of which comes from the census. It can suggest specific issues in a community and allow researchers to target further work to better reveal and understand its vulnerabilities.

The census reveals a raft of data about socioeconomic status, education, age, family characteristics, motor vehicle ownership, English language ability and employment. It can help identify communities where there are concentrations of people who may be more vulnerable in the event of a disaster and thus at higher risk. This can inform risk assessments and ultimately target mitigation and preparedness measures at these communities.

Evacuation

Modern evacuation management depends on demographic data to ensure that the population of an area can be evacuated safely in the event of an emergency. It is particularly relevant for emergencies where there is some warning such as floods, tropical cyclones and bush fires. Here are some key items useful in evacuation planning which can be derived from census data:

Total Population: This number isn’t as important as you might think, but it’s a good starting point. Evacuation centres need to be able to cope with evacuees who go there, but most people who evacuate prefer to stay with friends and family. Still the total population can give an indication of the number of people evacuation centres may need to assist.

Number of households: Although technologically based systems, such as Emergency Alert in Australia, are being increasingly utilised by the emergency services doorknocking remains a mainstay of evacuation warning. It is an effective means of warning, particularly when combined with other methods such as mass media and new technology. Doorknocking is resource intensive in terms of personnel and time. To know how many teams you need or how long it will take you need to know the number of doors that need to be knocked. The census provides answer in the number of households in an area.

Number of cars: In Australia and most other developed nations motor vehicles are the traditional method of evacuation. Australian emergency managers pioneered simple methods for calculating the time required to evacuate an area along a limited number of routes. The number of motor vehicles is a key factor in this calculation. Most households will take all their cars with them, so you can’t just rely on household numbers. With the rise in number of motor vehicles per household, knowledge of the number and growth over time in an area is critical to ensuring safe evacuation.

Number of households without cars: This is a critical factor to ensure that sufficient alternative transport is provided to evacuate those without a vehicle.

Number of vulnerable people: Those who are elderly, have a disability or are from a Non-English Speaking background may find it difficult to either evacuate on their own, or understand the evacuation warnings. These people are present in all areas, but if the census identifies a particular concentration this can allow for emergency planners to take their needs into account and plan accordingly.

Census data can also show changes in population in an area over time indicating when evacuation routes may reach capacity. This should trigger the need to either curb development or increase the capacity of these routes.

Community Engagement

Community engagement material and programs will function best if they are targeted and tailored to the at risk communities. Some of the useful census data for community engagement includes:

Language: A common and relatively cheap option is to reprint preparedness materials in different languages and then distribute these materials among the communities who speak that language.

Non-Private Dwellings: The census doesn’t just count people at home. It counts people in hospitals, prisons and other institutions. This can help emergency managers identify areas with a large concentration of facilities like nursing homes and hotels, which need to be targeted with education materials specific to their circumstances.

Private Dwelling types: The types of private dwellings in an area and the number of people in them can also suggest how community engagement should be targeted. For example caravans are a high risk group combining both high hazard (caravan parks are often in hazard prone areas) and high vulnerability (long term residents of caravan parks often have low socioeconomic status and may also be elderly or have a disability). Census data can identify at a broad scale areas with large concentrations of caravan residents enabling emergency managers to locate the individual parks and target education efforts.

New features in the 2011 census

There have been some changes to the way the census data is packaged for use. Of particular relevance to emergency managers are changes to the geographical areas on which census data is reported have improved in granularity. This allows emergency managers to examine demographic data on much smaller areas. Over time this will enable approaches that are increasingly customised towards communities.

Disaster Risk: Objectives?

The definition of risk in the Risk Management International Standard is the effect of uncertainty on objectives. This definition works quite easily in the corporate context, where a business has a clearly defined set of objectives, but what about in disaster management.

Communities are a bit messier than a business. They don’t get together and articulate their objectives over the next ‘corporate planning period’. Although governments do attempt gather general views and aspirations of the community, these efforts are coloured by the purpose of the engagement and more vocal individuals that they tend to capture.

Most disaster risk management experts would characterise their objective as the reduction of disaster risks. Plug that into the risk definition and you get something that’s pretty close to being circular. Unpack it a bit further, though and disaster risk reduction boils down to this:

  • Protecting communities from harm; and
  • Minimising the immediate and long-term impacts of that harm on communities.

Applying the risk definition to these objectives shifts the focus from the hazardous events to our ability to anticipate, protect and recover from the community impacts of hazardous events.

The risks that we’re trying to manage are not floods, storms and chemical leaks. They’re that our levees may be full of holes, our logistical coordination arrangements are inadequate and that the community is not prepared to shelter-in-place during an emergency.

This approach offers new opportunities. It allows emergency managers to:

  • Focus on actions that are much more within our control;
  • Identify potential failure points in our emergency management system before the disaster occurs; and
  • Prioritise improvements to existing systems, equipment and infrastructure rather than re-inventing the wheel.

Knowledge of hazardous events is still important, though. Knowing how disasters will test our capabilities mean understanding both and their interaction.

Applied a capability focussed risk management process? Let me know in the comments.

Welcome to Casus Calamitas

Welcome to Casus Calamitas, a blog about disaster risk, risk reduction and resilience in the modern world. Casus Calamitas loosely translates from latin to mean ‘chance of disaster’.

Latin doesn’t really have a word for risk, at least not in the modern sense. So what is the difference between chance and risk?

Central to modern systems of disaster management is the concept of ‘risk’. The International Standard on Risk Management (ISO 31000) defines risk as the effect of uncertainty on objectives. In the disaster context this uncertainty can relate to:

  • When will the hazardous event occur;
  • how intense will it be;
  • how will it impact on communities and the environment; and
  • how well will those communities respond to the impact.

Chance just refers to the timing of the event. Risk involves much more than chance. Unfortunately, people often focus just on this element missing the opportunity to do much more about minimising their risk. Mature risk management systems are able to bring in all the dimensions to strive towards a much broader and objective view of the disaster risks.

I’ve talked about the uncertainty associated with disasters, but that’s just half the ISO31000 definition. What are the objectives that this uncertainty impacts on?

That’s something I’ll focus on next time.