erik millstone food thinker september 2016

The evolution of food safety policy-making:

progress and retrogression

Erik Millstone

SPRUe.p.millstone@sussex.ac.uk

My topic is the evolution of food safety policy regimes over the interval since the mid-1970s - when I stumbled into this field. The discussion will comparatively outline changing regimes, in:

• the UK, • the EEC/EC/EU, • and the USA.

I had been teaching the history and philosophy of science at Sussex for less than 6 months when I read Daniel Zwerdling’s ‘Death for Dinner’ in the New York Review of Books, February 1974.

Food safety will be understood to include 3 main types of risks:

• Microbiological• Nutritional, and • Toxicological.

One key unit of this analysis is what the WTO SPS agreement refers to as:

the chosen level of protection (or CLP)

ie how safe?

Almost nothing is perfectly risk free. Life is intrinsically risky, in relation to food safety policy one question is:

what levels of risk (and uncertainty) have been deemed acceptable? A more interesting question is:

how have levels of risk (implicitly) deemed acceptable been changing? Which I will interpret as:

have they been rising or falling remaining unchanged?

A distinctive feature of the changing food safety policy regimes is that despite their frequent changes, they have almost invariably been legitimated by reference to eg:‘sound science’ or ‘independent scientific advice’.

Part of my analysis will consider the extent to which the evolution of the three policy regimes have been driven by:the state of, or changes in, scientific knowledge.

[Answer: not a lot.]

To appreciate the relevance of

‘science’, it helps to understand

the evolution of understandings,

and representations, of the role of

science in policy-making.

Science

Policy Making

Model 1: the technocratic model policy is based (only) on sound science

Facts

Technocratic narratives can be very appealing, eg to ministers and to expert advisors.

That was true not only of eg Thatcher & Gummer on BSE, but also eg Blair on GM foods and Dobson at the start of the FSA.

“I...think it is important for us to take on board all the issues relating to GM food. The only thing I have said…is that it is important for the whole debate to the conducted on the basis of scientific evidence, not on the basis of prejudice...the biotech industry in this country is an immensely important industry...I do worry that there are voices...that are not prepared to give enough consideration to the potential benefits as well as to the potential downsides of this...for the future ...it is important that this is conducted on proper scientific grounds.” (Blair, Hansard 18 June 2003)

The technocratic model is flawed, because scientific evidence is often incomplete and equivocal, so judgements are uncertain.

Eg carcinogenic risks from saccharin range between 0.22 and 1,144,000 extra cases of cancer in US popl. from ~70mg/cap/day.

Science in policy does not speak with certainty or with one voice.

One reason I am so critical of attempts to justify policies with technocratic narratives is that research has shown that:those who most enthusiastically articulate technocratic narratives are often best placed to know how misleading they are.

Even if per impossibile all scientific uncertainties were eliminated, science still could not decide food safety policy. You cannot derive an ‘ought’ from an ‘is’.

Policy judgements are concerned with the acceptability of possible risks (CLsP) and uncertainties in exchange for anticipated benefits. Those are normative culturally-contingent value judgements – they are policy matters, not scientific issues.

Model 2 – The Red Book ‘decisionist’ model

Risk assessment

Risk management

Risk

communication

Science

Values interests and practicalities

Social science

That model is the new orthodoxy. It was a step forward but it is incomplete because it ignores how non-scientific considerations frame scientific representations of risk.For example, choice of target ‘risks’, target ‘groups’, and judgements as to what is deemed ‘relevant evidence’.

The source of many of the uncertainties, and conflicting risk assessments, are those non-scientific framing assumptions.Codex’s crucial innovation was recognising those assumptions, and calling them: ‘risk assessment policy’.

The CODEX Alimentarius Commission (in 2003) and all its 189 Member States (in 2007) abandoned the Red Book Model in favour of what science policy analysts called a co-dynamic model – ie a model in which science and policy-making mutually influence each other.

Socio- economicand political

factors

Scientificfactors

Technical, economic, social

and political factors

Risk management

framing assumptions:

aka ‘risk assessment

policy’

Risk management(eg trade-offs

and judgements of acceptability)

Expert

Assessment

Model 3 - the co-dynamic model:reciprocal links between science and policy

Codex refers to Risk Assessment Policy in the following terms:“Determination of risk assessment policy should be included as a specific component of risk management.Risk assessment policy should be established by risk managers in advance of risk assessment, in consultation with risk assessors and all other interested parties… The mandate given by risk managers to risk assessors should be as clear as possible.”

Under traditional arrangements, risk assessment policies were routinely decided by nominally ‘scientific’ risk assessors, or by those who selected members of risk assessment bodies.Under those conditions, policy judgements were left entirely implicit and were misrepresented as if they were purely scientific. In EFSA and the FSA that remains the case.

Making risk assessment policy explicit means being as rigorous about the choice of questions asked as about the choice of answers given. RAP issues are always in play, not just wrt food safety and not just wrt risks, but they are often not explicit.

The UK’s Food Standards Agency (FSA) and the European Union’s European Food Safety Authority (EFSA) were both supposed to be constructed on a Red Book model.

In practice they have failed to do so.

They were supposed to ensure two separations:

- regulation from sponsorship, - science from policy.

Responsibility for protecting consumers and public health would (we were told) be separated from departments with responsibility for promoting the commercial interests of the food and/or farming industries, and would be re-located in departments responsible for public health and/or consumer protection.

Secondly, the reforms were supposed to ensure the separation of the science of food safety from any and all policy considerations.

In the prevailing idiom: the science of risk assessment would be entirely independent of, and uncontaminated by, risk management policy considerations.

Neither of those separations has been achieved. Regulation remains (~covertly) subordinated to sponsorship, eg in both ‘assessment’ and ‘management’ fora, and defended in terms of eg ‘competitiveness’ and ‘growth’.The FDA, the EFSA and FSA are all subject to varying levels of ‘corporate regulatory capture’.

Scientific risk assessments are invariably framed by prior non-scientific judgements about eg what counts as a ‘risk’, what counts as relevant evidence, how much of which kinds of evidence is necessary &/or sufficient to justify recommending acceptance, restriction of banning. Both the FDA and the FSA advise themselves, and they never receive advice that conflicts with their policy preferences. The Commission wants EFSA to be:“…independent, but not out of control…”

12 July 2010

Nov 2010

In the USA, CLsP have changed as indicated

Risks 1970s 1980s 1990s 2000-2009

2010-now

Microbiological ↑ ↔Red Book

↔ ↔ ↓AMR

Nutritional ↓ ↓ ↓ ↓ ↓

Toxicological ↑>Delany (1958)

↓IARLGAsp.

↓GMOs

↔ ↔

In the EEC/EC/EU:

Risks 1970s 1980s 1990s 2000-2009

2010-now

Microbiological ↓ ↓ BSE

↔nvCJD

Supply chain complexity

↔/↓DG-SANCO

& FVO

↓DG-SANTE

& FVO

Nutritional ↑ ↓ ↓ ↓ ↓

Toxicological ↔SCF

↔SCF

?Mut/Reg?

↓SEA

↔/↓EFSA

↓Asp &

Glyphosate

In the UK:

Risks 1970s 1980s 1990s 2000-2009

2010-now

Microbiological ↔ ↓ Salm etc

BSE

↑>3/96

E.Coli 0157

↔FSA

↓AMR

MRSA EColi

Nutritional ↑ ↓NACNE

↓ ↔FSA

Traffic lights

↓DoH

DEFRA

Toxicological ↔ ↓ ↔ ↓ ↓

Increasing rates of resistance (%) to key antibiotics in E. coli blood-poisoning infections in England, Wales and Northern Ireland

Prevalence of obesity among adults aged 16+ years in England

Health Survey for England 1993-2014 (3-year average)

The underlying diagnosis was faulty.

The problem is not that regulation and sponsorship interact, but that regulation remains covertly subordinated to sponsorship, rather than the other way round.

Especially in relation to support for R&D and innovation trajectories.

If standards are to rise, it will be necessary to:

•Explicitly implement CODEX RAP undertakings, accountably.

“Chairs of...advisory committees …[should ensure]…that the proceedings of the committee...are properly documented…so that there is a clear audit trail showing how the committee reached its decisions....decisions should include an explanation of where differences of opinions have arisen during discussions and why conclusions have been reached...They should also explain any assumptions and uncertainties that are inherent in their conclusions.”

But also, in conclusion:

•No longer pretend risk assessment is purely scientific, but make the linkages to policy explicit and accountable.

•Pursue a trajectory of predictably rising standards, to incentivise and drive R&D and the direction of innovation.

•Ensure that sponsorship is subordinated to regulation, rather than vice versa.

• Assess not just ‘risks’ but also ‘benefits’ and juxtapose them explicitly.

• Make risk assessments comparative in terms of safety – we want safer foods.

• Do not just regulate ingredients, but also products (and maybe corporate portfolios?).

• Ban all genotoxic carcinogens, and progressively reduce all in vivo and in vitro carcinogens, mutagens, and endocrine disrupters.

• Reverse the allocation of the benefit of the doubts.

• Oblige risk assessors to prioritise identification of false negatives over false positives.

• Subordinate agricultural policy to public and environmental health benchmarks.

• Ban all genotoxic carcinogens, and progressively reduce all in vivo and in vitro carcinogens, mutagens, and endocrine disrupters.

Brexit and chaos in the UK.

ConclusionOperationalising a co-dynamic model, and explicitly implementing the CODEX RAP provisions nationally and internationally could help create the conditions under which food safety policy-making would raise food safety standards, as well as achieve and reconcile both scientific and democratic legitimacy.