Saturday, 14 October 2017

HSNO Regulation Economics

Rod Oram has highlighted two facts about New Zealand's EPA
  1. It is using a new method called "net benefit" to evaluating hazardous substances and new organisms; and
  2. There is no written description of the new method.
Gleaned from an interview with EPA Chief Scientist Jacqueline Rowarth, Oram tells us that the net benefit method entails...
weighing up the financial, health and environmental costs of a substance against its economic benefits. For example, Rowarth says, Roundup, and other forms are glyphosate, are highly beneficial to farmers because they reduce weeds and significantly increase crop yields.“It’s a very difficult calculation,” she says. However much economics goes into the analysis, “ultimately there is a final point that becomes subjective” about wider societal values.
Exactly how does that work? I asked her. “We’re trying to formalise that because more people are challenging decisions. 
Does the EPA have a guide to this methodology? “I’ve been exploring this in press articles…and a document is being prepared internally on this.”
So basically they're making it up as they go along. Instead of sticking with the precautionary principle, the EPA seems to weighing up things without measuring them first.

Worse, there appears to be no international guidance on how to do this. On the contrary, according to Rowarth...
The EPA is a world leader for its work on “net benefit” analysis.... “Our scientists are being invited abroad” to give presentations on the methodology,
To me, this sounds like another round of the "light-handed regulation" New Zealand used for privatised monopolies in the 1990s. Pushed by extremist economists, this ill-fated idea was also hailed as world leading. Not one country followed though, so we weren't so much leading the world as heading off down a blind alley. Eventually, New Zealand endured a very painful re-calibration of regulatory expectations, though not before the privateers locked-in $2-3bn of un-earned capital gains in the electricity distribution sector alone.

As part of the re-calibration, the Commerce Commission as regulator was instructed by Parliament to develop "input methodologies" setting out in detail how regulation was going to be applied, to consult on these methods and for its decisions to be subject to a specially-enabled High Court review. All of this effort was designed to avoid regulating "on the fly" as the EPA now apparently does.

So the omens are not good. But what are the economic arguments against the EPA's novel method? There are two: it is conceptually wrong; and it is fraught with severe practical challenges.

The Concept
Back in the late 90s I visited a colleague who was an economics professor at Stanford and learned that a common description of a hopeless student there was "not clear on the concept".  In this EPA matter, the concept is the precautionary principle which goes like this:
if an action or policy has a suspected risk of causing severe harm to the public domain (affecting general health or the environment globally), the action should not be taken in the absence of scientific near-certainty about its safety
Notice that the level of risk is not specified and could have a low, even a very low probability. What matters is that "severe harm" might ensue. Examples of potential severe harm in the HSNO context include widespread disruption of human endocrine systems, depression and other forms of mental illness, and unintended consequences from the outdoor release of GMOs. These all count as "severe harm". Under the precautionary principle the onus would be on those seeking approval to prove "scientific near certainty".

One economic rationale for this approach is that disasters are best avoided if possible. Disaster insurance might possibly help, but ask the good people of Christchurch if they feel properly compensated. Or try to buy insurance against the expected mega-quake from the Hikurangi Subduction Zone

In the HSNO context though, the potential disasters are not natural in the sense of being beyond human influence. Far from it, any disasters will be the result of decisions made by humans at the EPA to approve the release of hazardous substances and/or new organisms. We can't insure against these disasters, so the EPA should not expose us to them.

The effect of the precautionary approach is to place an onus of proof on those wanting approval for a HSNO and to set a high bar (near certainty) for that proof. This is fully consistent with the real options literature, which requires that the timing of irreversible decisions made under uncertain conditions be delayed until the upside benefits are well in excess of the potential costs.

By contrast, the net benefit approach is based on expected values. If the EPA's probability-weighted assessment is that benefits exceed costs, then it will release the HSNO.

Here's a way to think of the difference between these approaches. Imagine you were in charge of Christchurch back in the day, and you had to decide whether to allow houses to be built on liquifaction-prone land. The precautionary principle would direct you to avoid exposing people to disaster, so you'd decline the request, markets would adjust and houses would be built somewhere else.

The property developers who bought swampy land cheap and now want to build houses will be advocating the net benefit approach. They'll have plausible-sounding economists and power-points for Africa, and they'll be regaling you with all the massive benefits of seeing things their way. If you ask about the risk of disaster, they'll assign it a tiny probability so that an expected (probability-weighted) value analysis aligns with their request, while gently suggesting that perhaps you shouldn't be such an old worry wart. 

So, yes, there is a huge difference and no, the net benefit approach does not properly accommodate disaster risk. That's why we use the precautionary approach.

Incidentally, the precautionary principle is baked into the HSNO Act. The purpose of the Act (s4) is to
protect the environment, and the health and safety of people and communities, by preventing or managing the adverse effects of hazardous substances and new organisms
To me, if you 'protect' a thing or person you would ensure they were not exposed to even a small risk of disaster. Parents of young children are usually clear on this concept. 

Section 7 of the HSNO Act is titled "precautionary approach" and requires that
All persons exercising functions, powers, and duties under this Act ... shall take into account the need for caution in managing adverse effects where there is scientific and technical uncertainty about those effects.
Finally, Schedule 1AA of the Act is the Stockholm Convention on Persistent Organic Pollutants which is explicitly based on the precautionary principle.

Practicalities of Net Benefit Approach
You'll have gleaned by now that I think the EPA is barking up a blind alley, probably the one where Jacqueline Rowarth buries all the strawmen she keeps murdering in the popular press. For the record though, I should just finish with a few questions for the EPA in the event that it proceeds down this track. I'll focus on glyphosate for now....

1. How do you measure annual and cumulative exposure to glyphosate for New Zealanders?
2. What probability do you assign to a New Zealander contracting cancer from exposure to glyphosate?
3. How did you estimate that probability?
4. In your model, what is the cost of a New Zealander contracting cancer?


No comments:

Post a Comment