One of the big challenges our Government has had to deal with has been the devastating Christchurch earthquakes. The Prime Minister has made the recovery and rebuild of Christchurch one of the four key priorities for our Government and all credit to the work of Gerry Brownlee and my Canterbury colleagues on the job they are doing to achieve this.

The post-earthquake challenge I want to address this morning is how part of our legacy needs to be a more resilient nation to future earthquakes. We have had a few reminders in recent weeks with the shakes in Kaikōura and Wanaka of how seismically active our country is. We have also witnessed on television the devastation in Nepal of how deadly poorly designed older buildings can be in an earthquake, with their death toll from the earthquake now over 7500.

Working out a sensible and balanced approach for regulating the earthquake risk of older buildings throughout New Zealand is one of the more complex policy problems we need to resolve this year. If we are too soft, we may risk the lives of hundreds of our fellow citizens whenever the next major event occurs. If we are too hard we will gut dozens of our country towns, lose hundreds of our heritage buildings and impose huge costs on farmers, apartment owners and businesses.

In looking to the future, it is useful to reflect on our past. New Zealand has experienced seven fatal earthquakes since 1840 with a total of 473 lives lost. The events of Napier in 1931 costing 256 lives and Christchurch in 2011 with 185 lives lost reinforces that earthquakes are our greatest natural disaster risk.

But to keep this discussion in context, we also need to consider other risks. We are 100 times more likely to die in a car accident, 50 times more likely to drown and 10 times more likely to die in a house fire. On the basis of costs and benefits, simple measures like smoke detectors in homes is going to trump earthquake strengthening of buildings, but that is an issue I will pursue another day.

The history of regulating buildings for earthquake resistance in New Zealand started in 1931. There have been several iterations of the standard for new buildings as the science of earthquakes and how we build to resist them has evolved. It is worth noting that whereas the death toll in Napier equated to one person in 100 being killed, 80 years later this ratio dropped in Christchurch to one in 2000. To put it another way, the equivalent death toll in Christchurch would have been 4000 people. The improved science and regulation of buildings between Napier and Christchurch reduced the fatality rate by about 95 per cent.

The comparatively easy public policy issue is ensuring all new buildings meet the very best of standards in respect of earthquake resistant design. I have some work to do in tightening up the regulation of the profession of engineers, but this is reasonably straight forward.

The far more challenging political issue is what we do with our older building stock. Our country has over 1.7 million existing buildings and it would be impractical and over the top to require all of them to be brought up to today’s earthquake resistance standards. We need to make decisions, over which buildings must be upgraded, to what standards and on what timetable.

We have had a thorough process on these issues with the Royal Commission of Inquiry providing some guidance in 2012, a public discussion paper process in 2013 and decisions by Cabinet later that year. The proposals have caused considerable angst for building owners, local government, engineers and economists all raising major questions about whether the approach has the right balance. I’ve spent hundreds of hours since taking up the portfolio last year, working through options for refining our policy.

The first substantive issue is where we draw the line on defining when a building needs to be upgraded. Today I want to confirm the policy of 34 per cent of new building standard. I accept it is arbitrary and that there is still a level of risk for buildings above 34 per cent but note that no one was killed in Christchurch from a building that met this level. Building owners may choose to go beyond this bare minimum and I commend owners, particularly when doing other major renovations, to take the opportunity to improve their buildings seismic resistance above this level where economic.

The second substantive question is over to what degree we nationally, as compared to locally, regulate. The 2004 Building Act effectively left it entirely to each of our 67 councils. The Royal Commission of Inquiry was quite critical that this had not worked and the issue had largely been put into the too hard basket. It recommended a single nationally consistent approach, which the Government largely adopted.

Today I am announcing a more refined approach. I don’t believe it is efficient or effective to have every one of our 67 councils going through the expensive technical and drawn out public process of developing their own policy on earthquake-prone buildings, but I do not believe a single national timetable appropriately deals with the different level of earthquake risk across the country.

The frequency of earthquakes likely to cause fatalities varies hugely across the country. We expect one in Wellington every 120 years, in Queenstown every 320 years, in Christchurch every 720 years but in Dunedin every 1700 years and Auckland only every 7400 years.

Another way to express this data is predicted future fatalities, across New Zealand from earthquakes. Eighty per cent are expected in greater Wellington. Eleven per cent in Christchurch. Provincial cities like Palmerston North, Tauranga, Napier, Hamilton and Nelson make up a further eight per cent. Auckland only makes up 0.5 per cent despite being a third of the country’s population. In short, you are over 50 times more at risk from an earthquake in Wellington as compared to Auckland. The most at risk places are Arthurs Pass, Hamner Springs and Milford Sound. The least risk places are Balclutha, Oamaru and Northland.

These different earthquake risk levels are taken into account in the new building standard but also need to be considered in the timeframes for assessing and upgrading older buildings.

These timeframes have a significant impact on costs. Older buildings are gradually being replaced and so a longer timeframe means fewer are being forced to upgrade or demolish. It is also far more economic to do earthquake strengthening work when other changes are being made to a building.

So rather than requiring all buildings to be assessed in five years and upgraded within 15, today I am announcing a more targeted approach.

New Zealand is to be categorised into three seismic zones of risk from high, to medium to low. High risk areas will need to be assessed within five years and strengthened within 15. Medium areas assessed within 10 and strengthened within 25 and low risk areas assessed within 15 and strengthened within 35 years.

The second change is targeting education and emergency buildings by requiring that in high and medium seismic risk areas they be identified and strengthened in half the standard time. We are prioritising all education buildings regularly occupied by 20 people or more like classrooms because they have a higher concentration of people than other buildings, the people are younger, the law compels attendance, and because schools are commonly used as civil defence centres. We also want to ensure buildings like hospitals can maintain essential services in the aftermath of a significant earthquake.

The third significant change in policy is reducing the scope of the buildings requiring assessment to those that pose the greatest risk. The current policy already excludes over 1.3 million stand-alone homes, but it includes 260,000 farm buildings. It is difficult justifying to the farmer that the hay shed which he might spend an hour a week in requires an earthquake risk assessment, but the farm house which he sleeps in every night does not.

The new policy will exclude farm buildings, retaining walls, fences, monuments, wharves, bridges, tunnels and storage tanks. The new methodology for identifying earthquake prone buildings will ensure the focus is on the older unreinforced masonry buildings that pose the greatest risk.

The main tool in the current policy to get buildings upgraded is this requirement that those assessed under 34 per cent must be strengthened or demolished in the specified timeframe.

The revised policy adds new tools. If an older building under 34 per cent is having a significant alteration or upgrade the earthquake strengthening will need to be done then. This is similar to the existing Building Act that requires fire egress and disability upgrades to occur simultaneously.

We are also working to drive upgrades through people being more aware of the earthquake resistance of buildings. There will be a publicly available register and website listing all earthquake-prone buildings. Buildings will be required to have notices at their entrance stating where they are below the minimum standard. We also intend such information to include the actual percentage of code with a red notice for those under 20 per cent and orange below 34 per cent. This reflects that the risk is substantially greater, as much as 10 times, for buildings well under the 34 per cent definition of an earthquake-prone building. These notices and extra information will influence customers and tenants and as a result building owners to strengthen buildings ahead of the maximum timeframe.

Other aspects of the policy that we are retaining are provisions for an extension of 10 years for listed heritage buildings and exemptions from strengthening for some low risk, low occupancy buildings like country halls in parts of rural New Zealand.

These policy changes I am outlining today are to be followed through into amendments to the Building Act and to regulations that we will be consulting on later this year. There is a lot of work ahead for us in working with engineers, building owners and local government in getting the detail right.

Let me summarise the practical effect of these policy changes.

Firstly a lot fewer buildings will require an assessment. Rather than 500,000 the focus will be on the 30,000 buildings that pose the greatest risk.

Secondly, this approach puts the pressure on upgrading of schools and hospitals in our most earthquake prone areas. They will need to all be above the 34 per cent minimum within a decade.

There is a lot less pressure for building upgrades in places like Auckland, Oamaru, Northland and Dunedin which will now have up to 50 years rather than 20 to bring their older buildings up to standard.

It is true that this revised earthquake-prone buildings policy is more gradual than the first proposal, but it is also more ambitious than how other seismically active countries deal with older buildings. In California and Italy, the only older buildings that are legally compelled to be upgraded are schools and hospitals.

I have been clear in refining the policy that the Government wants to retain the safety gains but get the costs down. This policy will result in an estimated 330 fewer deaths and 350 fewer serious injuries from earthquakes over the next century, the same as the earlier proposals. However, by taking a more targeted approach, the cost is $777 million rather than $1360 million – a saving of over $500 million.

Can I conclude by saying there are no easy answers to the seismic risk posed by thousands of older buildings in New Zealand. We cannot completely eliminate the risk to life, nor save every heritage building, nor avoid a substantial strengthening bill. This revised policy strikes a better balance between safety, cost, the retention of heritage and practicality.