Misunderstanding the policy process

Policy development happens through a complex interplay of politicians' directives, civil servants' work, and civil society input. 

Bryce Edwards is director at The Integrity Institute, which has been pretty critical about anyone's involvement in policy processes other than the bureaucracy and anyone considered to not have a financial stake in an outcome. 

Last week, BusinessDesk found the Institute's trust deed. Or, rather, decided it was worth checking over at the Company's Office to see what it says. And the whole document is really weird. Things that ought to be operational decisions for the Director are instead hard-coded into the deed. The rest is a combination of direction to do the obvious and specific targets they want investigated, like going after Transparency International. 

I'd threaded the thing; you can't link directly to a Companies Office filing. 

There's been a fair bit of tu quoque over this. But the best take I think thus far is from Deb Te Kawa. A snip:

Sometimes, those who claim to defend democracy are the ones who misunderstand it most. The Integrity Institute’s recent campaign, ostensibly about exposing undue influence over policy, has revealed something more troubling: a fundamental confusion about how policy advice, participation, and legitimacy actually work in democratic systems.

The policy advisory literature has long since moved past the myth of the neutral bureaucracy. As Craft and Howlett (2012) argue, modern advisory systems are pluralised and layered. Policy advice no longer flows solely from within the state. It emerges through dynamic interaction between political actors, ministerial advisers, public servants, and external stakeholders: including iwi, hapū, academics, commentators, lobbyists, consultants, contractors, industry and community interest groups, and civil society. This is not a breakdown in integrity. It is a shift in how democratic knowledge is produced and contested; a shift that has been underway since 1996, when we began correcting for the distortions of the 1980s’ new public management reforms.

Rather than engage with this shift, the Integrity Institute appears to reject its premise. In targeting Federated Farmers, the Institute frames visible, declared advocacy as inherently suspect. Yet publishing policy platforms, meeting with ministers, or advocating for sector interests is entirely within the bounds of democratic practice. As Craft and Halligan (2020) remind us, robust advisory systems must accommodate both internal and external sources of advice. To treat advocacy as corruption is to misread the core architecture of modern policy-making.

This conflation is not just technically inaccurate. It is democratically dangerous. Legitimacy in policy does not come from insulation. As I have been exploring in the Waitangi Tribunal Thursdays series, and arguing in The Practical State, it comes from contestability, transparency, and deliberative engagement.

Whether it’s iwi asserting rangatiratanga, unions calling for fairer conditions, or academics and researchers offering empirical insight, the presence of diverse voices is a safeguard, and not a threat. As the policy advisory literature insists, multiple advisory channels are vital for balanced decision-making in complex societies.

The real question, then, is not whether influence exists. It always does. The question is: what kind of influence, under what conditions, and with what visibility? The distinction between transparent, procedural engagement and opaque, privileged access is not semantic. It is constitutional. Effective oversight requires more than tracing contacts. It demands a grounded understanding of procedural fairness, institutional independence, and the layered nature of advisory input.

With that particular critique in mind, let's have a look at a different campaign. 

One that Guyon Espiner has been running, on the public dime, at Radio New Zealand, and at RNZ-Newsroom co-production The Detail. 

In his telling, shadowy alcohol industry influence stymied admirable public health efforts to adopt new and stricter Canadian low-risk drinking guidelines.

He's had a lot of airplay on this. If you're reading this in New Zealand, you've been forced to pay for it. 

So what is that shadowy influence?

David Farrar has the correspondence. 

The Brewers Association wrote to Ministry of Health asking for details on a review of the low-risk drinking guidelines. And there was also correspondence on the use of the alcohol levy - a small levy imposed on every bit of alcohol sold, used to fund various harm-reduction efforts. Producer levy schemes tend to have producer involvement. 

The Brewers also pointed out what looked like an error on the Health NZ website (recall that Health NZ is the operational arm, Ministry is policy). They thought that some proposed Canadian guidelines had actually been implemented. 

Health Canada had commissioned a third party to produce revised alcohol guidelines. That was a couple of years ago. Those proposed guidelines have not been adopted or ratified by Health Canada. And two different Ministers of Mental Health and Addictions, in late 2024 and early 2025, have confirmed that the 2011 guidelines remain the ones in place. 

Health NZ thought that Canada had tightened its guidelines and was taking this as basis for tightening ours. 

The Brewers pointed out an error. The Ministry of Health saw that error corrected. And Guyon Espiner deemed the whole thing an example of influence that needs to be stopped. Of course, on his podcast interview with The Detail, he hedges a bit - saying he's only raising questions and noting how interesting it is that a framework convention bars industry discussion with government in the case of tobacco but not for alcohol. Not that he's campaigning to get the framework convention extended to alcohol. 

Let's go back to Deb's piece. 

As Craft and Howlett (2012) argue, modern advisory systems are pluralised and layered. Policy advice no longer flows solely from within the state. It emerges through dynamic interaction between political actors, ministerial advisers, public servants, and external stakeholders: including iwi, hapū, academics, commentators, lobbyists, consultants, contractors, industry and community interest groups, and civil society. This is not a breakdown in integrity. It is a shift in how democratic knowledge is produced and contested;

The Brewers found an error. Min Health made the same mistake Espiner did, potentially based in motivated reasoning - wishing that the Canadians had given NZ an excuse to tighten guidelines here. 

That error was corrected, because we don't stick bureaucrats in towers and ask them to hand down advice from on-high. There is interaction. It is helpful. It makes things suck less. 

Some campaigners, like Edwards, and like Espiner, seem not to like it when that interaction results in policies that they like less. They seem to view it as inherently corrupt. At least Edwards isn't doing it on public funding. 

David Farrar's post with the correspondence went up on the 25th. 

The day before that, I submitted a column to Newsroom on the topic, because I'd first caught this on their site - they co-produce The Detail. I sent it through on Wednesday of last week for my usual slot on Tuesday - earlier than usual, because it was critiquing some of their work and I wanted to give them time with it. 

After asking that I add a lot more detail on the evidence around moderate drinking, they decided not to run it. There was an in-house view that Espiner had sufficiently couched what he'd said on The Detail. Perhaps there was a background worry that publishing a critique would make a likely press council / BSA complaint about The Detail piece more viable; I wouldn't know. 

I brought the piece back to 800 words; it was in Monday's Post as a full-page print piece (and presumably Press etc). An ungated version is here. I confirmed bits of the correspondence with the Brewers Association in the interval. 

The full piece I'd sent to Newsroom is below; it's a lot longer than the version at the Post, in part because they asked me to add a fair bit of content. 

I only saw Deb's piece after all this. But it's been rolling around in my head since then. The policy process really isn't what some of these campaigners seem to wish it were. 

Anyway - the piece that Newsroom declined. I declined to write a substitute piece for them for this week. 

Confusion about Canada’s alcohol guidelines

Some things are legitimately confusing.

Canada’s low-risk drinking guidelines are not confusing. They are readily available on Health Canada’s website for anyone who cares to check.

But the discussion around them has been confusing.

Last week, RNZ’s Guyon Espiner claimed that “the maximum number of drinks recommended per week to stay low-risk (for men and women) is set at two in Canada” and that safe drinking guidelines in Canada had reduced in recent years.

Earlier this week, on The Detail, Espiner noted that “there’s some debate about what’s official and what’s not official.”

Canada’s official guidelines are easily found. Simply go to Google and ask for Health Canada’s low-risk alcohol drinking guidelines. If your Google is like my Google, it’ll send you to this Government of Canada page. The page is current; it was last updated in March of this year.

That page will tell you two things.

First, the official Canadian low-risk guidelines are similar to the New Zealand guidelines that Espiner cites — no more than 10 standard drinks per week for women; 15 for men.

[Canada’s guidelines, as reported by Health Canada]

But there’s a catch to that. And that’s the second thing you’ll find. Canada’s standard drinks are larger than ours. There, a standard drink is 13.45 grams of pure alcohol. A New Zealand standard drink is only 10 grams. But you have to know that a New Zealand standard drink is only 10 grams to notice that …detail.

When you put those together, it is obvious that Canada’s official low-risk drinking guidelines allow more consumption than New Zealand’s. 15 standard New Zealand drinks means 150 grams of alcohol. 15 standard Canadian drinks are 201.75 grams of alcohol, or just over 20 standard New Zealand drinks.

To put things in further perspective, a 330mL bottle of Panhead Supercharger (the closest bottle to hand here at the office) measures 1.5 NZ standard drinks or 1.1 Canadian standard drinks.

Espiner damns New Zealand’s guidelines for being out of step with Canada’s and blames industry influence.

But the Canadian guidelines he cites are not the official guidelines. Nor are they ones that Health Canada has proposed.

Instead, they are guidelines suggested by the Canadian Centre for Substance Abuse. They are a research and advocacy organisation that receives government funding and provides recommendations, like many New Zealand NGOs undertaking temperance-tinged research into alcohol on the government’s dime.

Health Canada did commission the work, but has not taken up the recommendations.

Last November, Yuval Daniel, director of communications for the federal Minister of Mental Health and Addictions, in response to a media query, stated that,

“The Canadian Centre for Substance Abuse and Addiction’s proposed guidelines have not been adopted by the Government of Canada. Canada’s 2011 low-risk alcohol drinking guidelines remain the official guidance.”

The Minister’s office reaffirmed the 2011 guidelines again in January of this year.

"Canada’s 2011 low-risk alcohol drinking guidelines remain the official guidance to Canadians and Health Canada’s website continues to reflect this," the statement said.

"The CCSA’s proposed guidelines are part of discussions with a broad range of stakeholders to determine the best approaches to communicating on the risks of alcohol consumption to the public."

It is always possible that the Canadian government will change its guidance in response to the CCSA’s recommendations. Before February, the governing Liberals could have felt constrained by an impending election and views that Trudeau’s government had been overly ‘nanny state’. The CCSA recommendations were at least somewhat controversial and contested.

But Espiner was wrong to damn New Zealand’s guidelines for being old and out-of-step in comparison to Canada’s.

And it seems wrong to damn industry influence in New Zealand for preventing us adopting proposed Canadian guidelines that even Canada has not adopted. A late update to Espiner’s story at RNZ, which criticised industry’s checking of our Ministry’s reliance on Canadian examples, includes comment from the Brewers Association’s Dylan Firth:

"It is also important to note that the (RNZ) story highlights concern raised with reference to the Canadian guidelines with the Ministry of Health. But does not point out that the Canadian guidelines have not in fact been changed. Hence the issue being raised with the Ministry with what was present on their website at the time," Firth said.

The podcast at The Detail has not been updated. Listeners to it, and readers of the column in Newsroom accompanying it, could be forgiven for believing that there is some sort of current crisis in alcohol consumption.

But the numbers suggest that, if consumption today is a crisis, it’s amazing that New Zealand survived the 1980s.

Statistics New Zealand tallies the number of litres of pure alcohol available for consumption per person. You can check the figures yourself, but Stats NZ doesn’t make it easy. You have to go to Infoshare, select Industry Sectors, then select Alcohol Available for Consumption.

Apologies for the jargon here. Stats calls it “available for consumption” because that is how it gets its data – when excise is levied and alcohol is made available for consumption. Stats cannot know whether your bottle of scotch will last a few weeks or a few years. The numbers are hard to find if you didn’t already know where to look.

Their statistics go back to 1986. In 1986, there were 11.6 litres of pure alcohol available per person aged 18 and over. That works out to about 2.5 standard New Zealand drinks per day per adult. Since some people don’t drink, others must have been consuming considerable amounts.

Through the 1990s, Kiwis consumed between nine and ten litres of pure alcohol per capita per year. Consumption blipped up toward the end of the 2000s, back up to ten litres. And it’s been on a steady decline since 2011, when New Zealand’s current guidelines came in.

In the year to March 2025, only 8.1 litres of pure alcohol were available per person – over a 30% drop since 1986.

Average consumption isn’t enough to tell the story on its own. It is possible to have an increase in harmful consumption despite overall decline if light and moderate drinkers mostly quit drinking while others shift to heavy consumption.

Fortunately, the New Zealand Health Survey has some data on that. Its most recent survey, for 2023/24, looks at a range of indicators. Changes in questions between 2015/16 and 2016/17 mean we can’t go as far back as we can with average consumption.

In 2016/17, 79.5% of us consumed alcohol in the past year. That declined to 76% by 2023/24.

Over the same period, ‘hazardous drinking patterns’ declined from 24.9% to 21.9% among those who consumed any alcohol.

The survey defines ‘heavy episodic drinking’ as six or more standard NZ drinks on one occasion. But remember that a single bottle of Supercharger is 1.5 standard drinks. So, four bottles of Supercharger on a single evening will get you there.

On that definition, the proportion of drinkers engaging in heavy episodic drinking at least monthly dropped from 27.5% to 23.4%, and the proportion of drinkers doing so at least weekly dropped from 14.9% to 10.8%.

So fewer among us are drinking. And fewer of those who do consume alcohol are doing so in ways that are considered harmful. All of these reported changes between 2016/17 and 2023/24 were considered statistically significant – that’s what the little down arrow on the table below represents.

 

It’s always possible to say that those figures are still too high. But surely the substantial decline in that kind of consumption ought to be mentioned somewhere, lest those unfamiliar with the stats be led to believe that we’re in some kind of industry-induced crisis, given the reporting on it.

I think the current guidelines are fine, but obviously that’s a more contestable position. In Canada, Brock University’s Professor of Health Sciences Dan Malleck pointed to problems in the CCSA’s recommendations. He notes that their research review set guidelines that ruled out all but a handful of studies and certainly was not a cost-benefit assessment.

Alcohol and health gets complicated because alcohol increases the risk of some disorders, particularly when consumed heavily, while reducing the risk of other disorders, including at moderate levels of consumption.

People who want to emphasise the health benefits of drinking could point to solid evidence on cardiovascular disease. Rimm and Moats found that those benefits are present in a prospective study of people with baseline healthy lifestyles.

People who want to emphasise the health risks of drinking generally point to cancer, where alcohol increases the risk.

Mumakal and coauthors pointed to some of the potential associations in a now decade-old article pleading for a long-term randomised control trial of moderate drinking’s effects.

 

 

Advocates pointing out problems in one set of studies (either showing health benefits or harms) sometimes to ignore similar problems in the other set of studies. The same standard of evidence should hold regardless of whether a study accords with one’s policy preferences.

This is not medical advice; I am not that kind of doctor. But unless you have a strong family history of one sort of problem or the other, it may be safer to play the averages by looking at the combined effect across all disorders.

Looking at disorders one-by-one is a problem. The level of drinking that minimises your risk of heart disease increases your risk of other diseases.  

All-source mortality risk is a decent way of tallying things up across all disorders, or at least relative to available alternatives. That’s the risk of dying in any given year, of any cause.

It’s important to exclude former drinkers in these kinds of studies. And studies have done that at least since 2006; my go-to from that era is this metastudy by DiCastelnuovo, Donati et al. The risk of dying from anything is minimised at consumption up to around a standard drink per day. Risk rises from there. If you’re having three drinks per day, you have about the same risk of dying as someone who has never consumed alcohol. And if you drink more than that, you take on more risk.

Work since then has flattened the J-curve somewhat, with a smaller reduction in mortality from light drinking. But there just does not seem real-world risk worth worrying about for consumption within the current New Zealand guidelines.

And there is risk involved in putting out nonsensical advice. It trains a lot of people to ignore all advice. And some advice is very much worth heeding. Heavy drinking in pregnancy is very bad.

New Zealand’s guidelines seem generally fine. Nevertheless, our public broadcaster, on public funding, which I am forced to contribute toward through money taken from me every fortnight, left listeners with the impression that we are in some kind of alcohol crisis and that New Zealand’s guidelines are out of step with Canada’s.

But Canada’s actual official guidelines allow higher consumption than New Zealand’s because their standard drinks are larger than ours.

The partial late correction was welcome but should not have been needed.

Statistics New Zealand’s data shows a long-term decline in average consumption and a more recent sharp decline. Their availability-based estimates correspond with a substantial decline in harmful consumption patterns in the New Zealand Health Survey.

Some things are legitimately confusing. Canada’s alcohol guidelines are not confusing. But what Radio New Zealand thinks it is doing in its current temperance campaign is at least a little confusing.