Estimates afford latitude

Here then, we begin to appreciate how the downsizing of estimates might matter in different ways to different stakeholders (see sections below). For many scientists, the single number is not a matter of concern. Accounts invoke latitude in their defence. Precision about the absolute number, an unknown, is said to be less important than the ‘order of magnitude’ and ‘trajectory’ of change in the target population, and its relation to other measures including treatment coverage. The advice is: ‘Do not look at a single number but look at the trajectory of what’s going on’ (P23). When doing so, ‘the magnitude of change is not that great’ (P21). For some, the number ‘makes not a jot of difference’ in practical terms (P23). As commented on the declining estimate of those left to treat:

There is a bit of latitude. I really don’t think things would have been very different whether it’s 100 000 or 80 000, whether the treatment uptake was 48% or 56%, it’s still a huge number of people, large proportion of people, that need to be reached in terms of the strategic goals and programmes. (P9)

If somebody says, ‘You’ve got 70 000 to treat or 85 000 left to treat’, I go ‘How does that change what I actually have to structurally do in my thinking and in my work?’ (P23)

In this account, the precision of the number per se ‘does not actually matter’. The latitude that surrounds the estimate does not alter ‘the reality that Australia is not on track’:

In some ways it doesn’t actually matter if the number goes down by 30%, there’s still a lot of people living with hep C who still need to be cured. (P26)

Even if this change places us closer to elimination, the point is we’re still not going to make it in time. […] The change in number won’t change government response materially. […] If I’ve got 70 000 people left to treat in Australia or 90 000 people left to treat in Australia or 120 000 people left to treat in Australia, okay, it's still tens of thousands. (P23)

The accounts of scientists emphasise the need ‘to get this number correct so that we can measure progress against the elimination targets’ as well as to ‘match’ it as ‘accurately as possible’ to the ‘real-world’ (P21; P9; P3), but at the same time, they accentuate the latitude and fluidity of estimation to caution against unrealistic assumptions of numerical stability and precision. The latitude in estimation is such that one scientist notes that while downsizing the estimate might be the ‘best thing to do’ it is ‘a brave thing’, because ‘we have no idea how accurate [estimates] are in how they represent the truth of the number of people living with hepatitis C, no one knows’ (P18). Users of the estimates (eg, those working in policy, advocacy and practice), some scientists suggest, ‘potentially do not understand that things do change, and that we need to be prepared for it’ (P21). Taken together, the denominator is said to have been afforded ‘too much power’, as if ‘Gospel’, appearing too much as a number, and not enough as an estimate:

There’s a bit of an own goal that’s happened over time where the original numbers have been considered to be as Gospel and, you know, really this is the number, when we should’ve always been talking about it as an estimate, and making it clear that it was just an estimate (P11)

The controversy over downsizing the estimate is here shifted as a problem of the perception that ‘the number is the number’. An appeal is made to the ‘fine print’, the ‘caveats’, that scientists say they communicate about the uncertainties of estimates to indicate that numbers should not be treated as such:

People have this view of the number is the number. Well, actually it’s not. The number is as best as at the moment we can work out. We always write caveats around it, if you read the fine print, but as most people don’t bother to read the fine print, they don’t go, ‘Oh, there’s caveats around here, and those confidence intervals are there for a reason’. But most people don’t know what confidence intervals really are. (P23)

Narratives of scale

Our analysis considers the number as a matter of narrative (how the story is made), and not simply calculus (how the calculation is made). The denominator ‘can be shifted in direction depending on the narrative that is around’ (P5). Whereas altering the estimate can be minimised as ‘ordinary’ and ‘mundane’, and within a latitude that does not present cause for concern, as noted above, it can also be narrated otherwise, as ‘big stuff’. For instance, in contrast to the accounts of some scientists, the accounts of some community stakeholders emphasised: ‘This isn’t a usual ticking over of the numbers, this is a major, this is a major change that’s proposed’ (P28). The controversy situates altered estimation as a site of ‘scale politics’, producing ‘scalar narratives’40 in the sizing and shaping of hepatitis C as a matter of policy and community concern. Scalar narratives are those that use scale as a resource to explain events and present an argument,41 42 here about hepatitis C. Here is one perspective on this deliberation:

I think what do we want this number to be is a tricky one, because what we want this number to be from the Department of Health is we want it to be small because it’s a low burden of disease. But what we want this number to be from resourcing of other areas might be that you want it to be bigger because that’s then going to ensure resourcing and allocation of staffing and things like that (P5)

In scalar narratives of elimination progress, the population estimate is prime. It is the ‘primary number that drives the response’ (P28). It ‘defines scale’ (P16). It is seen as the base from which ‘everything flow(s)’ when measuring testing, treatment and other targets. The number acts as a ‘proxy indicator of urgency’ (P12) and of ‘need for action’ (P4), shaping high-level policy (P10; P12) as well as the on-the-ground work plans of service providers (P7; P28; P29; P31). The number is circulated ‘everywhere’, in ‘every report’, and ‘gets used at every single thing’ (P28). Whatever the latitude of its estimation, the effects of this number are ‘real’:

Areas of action, and priority populations, and priority settings, within the strategies are sort of built around that number essentially, and if the number changes and that sort of gives rise to the downstream effects. (P10)

From a national number, it flows down like to people’s individual workplans. It is full-on how much this number is real. (P28)

This stuff matters. This is not mathematics published in a report somewhere that gathers dust, this modelling makes national policy. (P12)

Experience suggests that the number is ‘shockingly powerful’ and can make a material difference:

It’s been quite powerful for us in our advocacy to be talking about targets that had been missed. (P4)

The metrics-driven focus on progress and success is what perpetuates the yearning to get the number as ‘right’ and as ‘realistic’ as possible. There is said to be something of an ‘obsession’ with the base numbers that link to targets in the race to eliminate:

I work really closely with policymakers, including drafting things like the national strategies, and what I’ve noticed working with policymakers in this space is they’re quite obsessed with these numbers. (P4)

It’s important to understand the burden of disease in the population, and it’s important to get this number correct, so that we can measure our progress against the elimination targets. (P21)

As Australia enters the ‘endgame’ of hepatitis C elimination, scaling the problem intensifies. Measures of base population prevalence have come to matter more, and matter differently, since the arrival of DAA treatments, and with elimination becoming a probability, not mere possibility.13 Yet Australia’s progress against global elimination targets has ‘slowed’, moving from ‘fast-track’ and ‘early’ elimination to ‘missed’ targets and to no longer being ‘on track’.14 15 These shifts situate the controversy: ‘All of a sudden the number matters in a different way’ (P26); ‘The reason this [estimate] has such power is particularly the point that we’re at in Australia’ (P5). There is a ‘need to know’ (P13; P5):

We need to know […] if we’re getting close to or near endgame elimination. We know we have to try different approaches the closer we get to […] endgame elimination. (P20)

The concern here is how altered estimates reshape ‘policy narrative’, including bringing Australia back ‘on track’. A reducing base population may help to re-enact a scalar narrative of progress being made: ‘Our concern was like a major shift in the policy narrative from not being on track to kind of being on track’ (P4); ‘You change that number and you can model your way out of the problem’ (P28). It ‘suits the agenda now, for the estimate to be low’ (P12). We see contrast then, in how scalar narratives are imagined to align with imagined networks of expertise, with science and policy actors presumed more comfortable with a reducing population estimate than community actors concerned that this might undermine hepatitis C represented ‘big’ and ‘big enough’ going forwards. Any lessening of policy commitment is feared to reverse progress made (P16). A lesser target population ‘implies the problem is not as big as it is’ (P25). It creates ‘policy ambiguity’ (P4). In the endgame, uncertainty is unhelpful, because ‘the same, if not more resourcing is required’ (P7). Because ‘it actually gets harder and harder to achieve the same gains that you’ve seen previously […] it actually needs an increase in resources’ (P2). The endgame scalar narrative needs to perform things ‘big enough’ to maintain momentum: ‘This is a moment where you should hold your form’ (P23). There is, therefore, also a nuanced distinction in ‘scale politics’37 between the size of the population in need and the size of the problem at hand, with the reduction of the former indicating greater, if not the same, resource need going forward.

The event of controversy over population estimation can thus be seen as a deliberation on how big the matter of concern is represented to be, and what downsizing the population of people living with hepatitis C does to the narrative now:

We really need to be investing here. We need to be scaling-up. Now is the time in the epidemic. We’ve seven years left to achieve elimination, and a concern, not looking at the numbers in detail yet, but just knowing that it had been revised down, immediately was, ‘Shit! Well, what if this makes the narrative now?’. (P4)

This question of ‘what makes the narrative now?’ accentuates the potentials, and not just the pitfalls, of enumeration as an ‘evidence-making’ event35; that is, as a site not only of controversy but of opportunity.22 What potentials then, might different versions of numbers afford? Here, accounts emphasise how the downsizing of the target population intensifies the pressure to justify investment at a critical point in Australia’s elimination challenge, with the potential to ‘change the discourse’ in policy, planning and priorities:

Unfortunately, in government, those conversations can get pretty kind of simplistic and reductionist. So, where you’ve got clear evidence that a problem is big and expensive, it’s sort of easier to go and have arguments as to why investment should follow. I don’t think that necessarily leads to a scenario where if the number is revised down because of, you know, newer surveying and resourcing and updated modelling, that money goes away, but it does mean that you have to better articulate why […] maintaining investment to achieve that is such an important goal, and that what you’re investing in is going to get you there. So, in a situation where if the numbers drop substantially, […] I think it just changes the conversation in a way that people start to go, well, how much investment do you need? Why are you investing in it? It requires greater clarity.’ (P8)

Narratives of care

Narratives of scale intersect with those of care. The downsizing of the population estimate is accounted as a lessening which is not merely felt in numerical terms but as a form of devaluing and discounting in other ways too:

For people living with hepatitis C, all of a sudden it’s not as a big a problem, which for some people is a good thing, but for other people potentially minimises what they are experiencing. […] This has been done and all of a sudden there is less people. […] We’ve just changed the number of people overnight. As an individual, it feels like you’ve either just been taken out and tossed aside, or you know, put back in. (P5)

In this narrative, there are fewer people, and people who matter have gone ‘missing’: ‘40 000 people seemed to disappear from the prevalence number overnight, and there has been some questions raised about where did those people go’ (P5). This is the narrative that envisages the estimate as an absolute, as a number that represents actual individuals. In this articulation, the number is not a mere estimate, and moreover, it is more-than a number. When translated as not ‘just a number’, ‘It is people that we’re talking about, the numbers reflect people’ (P16). ‘People see this as people, they see it as themselves’ (P28). This is why getting the number ‘right’, ‘accurate’, ‘correct’ and ‘true’ is also a concern of those working face-to-face with affected people (P20). In this narrative, latitude and fluidity in estimation are part of the problem and cause of the controversy rather than part of the defence which seeks to moderate it. Because ‘the number really matters’ in representing lived experience—of people, care and treatment—the claim that estimates afford numbers a wide latitude that ‘does not matter’ in practice is contested. The contestation here then, is the navigation between an estimate that is also a number: The changing number is ‘not really real, but real’ at the same time. For instance, it is not that there is a misunderstanding that estimates are ‘the result of an equation’ that do not necessarily represent ‘individual people’—the science is relatively well understood—but that this is how they are felt in experiential terms. As commented:

I am a consumer of the numbers. […] So the numbers actually mean something to me. […] I’ve learnt recently that the numbers perhaps are not individual people. I think of them that there are 117 000 people with hepatitis C in Australia. I think of them as individuals and people (P7)

The downsizing of the denominator is thus materialised in these accounts as people that were once there but are now ‘never there to begin with’. While this downsizing in science-based narratives does not fundamentally alter the ‘reality’ of ‘not being on track’ and of what therefore needs to be done (see above), it is ontologically unsettling for some: ‘If the number is not real’, then what is? This is a lessening that is not merely numerical but translated as an absenting and disruption of experience, including in relation to care and capacity for action. Calculus translates as affective value, with ‘lost people’ no longer ‘there’ troubling the ‘real’ of care invested as well as projected. The number is not merely ‘the size the problem, it’s the justification for the work’; a ‘measure of the work that has been achieved’, and ‘the worthiness of the investment’ that, taken together, constitutes care:

They are out there doing stuff, and it’s hard work, and it’s frustrating work. […] We’ve treated 100 000 people, it’s something that we’re really proud of, and we should be bloody proud of it. Yeah, so the anger that I heard last year when suddenly 40 000 were just dropped off overnight. It was like, well, we don’t have to do anything from now on, because the number will just keep dropping, and we’ll reach elimination targets by 2030. It overlooks the effort. (P7)

The ‘big impact’ of altered estimates is recognised among some scientists, who reflect that they may ‘underestimate’ the social effects of how numbers become materialised in experience:

If I’m a person living with hepatitis C or if I’m an advocate who has been working in hepatitis C, I see the number of people living with hepatitis C drop from an estimated 120 000 down to 80 000 in one fell swoop. (P1)

We’ve got 30 000 less people now, so instead of 110 000, you’re saying there is only 80 000 people. Like, how can that be? Like, how can you just drop off a third of the population that we think has hepatitis C? (P9)

The ‘sudden’ loss of people is here positioned, by some, as ‘care-less’, representing a discounting of the care effort. The altered number presents as a ‘bombshell’ that is ‘blindsiding’, as not careful enough, and as a reduction that is ‘undermining’ the strategic response, as care-less. Not only is the timing of the downsizing in the number of people living with hepatitis C made contentious because it coincides with the launch of national endgame strategy, but it was said, by some, to have happened ‘overnight’ as a ‘surprise’ (P25), and presented as a ‘done deal’ (P4). The revision was likened to ‘shifting the carpet under us’ (P1). Coming too fast, the downsized estimate generates a sense of unease and lack of preparedness which in practical terms is not without risk. Concerns here included damage to reputation and credibility, for instance, in upending the capacity of stakeholders to have a narrative of confidence when accounting for a lesser population:

80 000 or 70 000 people no longer exist in that figure and that’s […] quite a significant thing to explain to people. (P16)

The ‘care-less’ effects of downsizing the denominator thus combine a sense of discounting across people, care experience and community-science relations. Here, the controversy is also enacted as a matter of respect and trust:

A disrespect for the rest of the sector, I think, that this is an important number, that we’re all in this together, and we’ve all got to understand where the number comes from. (P7)

Nobody’s disputing science changes or things change or anything like that, but you can't get around three times in 3 years, and change the number that significantly, and then stand there with a straight face to the community and be like, ‘we've got it’. (P28)

I think that the most unsettling part would be that there was just such a dramatic reduction in the numbers, and what does that say about the credibility of epidemiology itself? (P31)

With ‘community’ attributed ‘deep emotional attachment to these numbers’ because they ‘reflect people’—people in need, people cared, people cured—accounts appear to be in search of what constitutes a ‘good number’. A ‘good number’ here merges ideas of calculus (a good enough scientific estimate of people and population) with ideas of what numbers embody in their social meanings and values (measures that reflect actualised experience and capacity for action) with ideas of numbers that care (measures that are valued as fair and just). These versions of numbers each materialise or seek correspondence with reality, but they do so differently, and multiply at the same time. This emergent notion of ‘good number’ in its situation suggests an emergent ethics of enumeration for the field. Here is an account that begins to traverse some of these multiple versions of ‘good number’, in the face of unsettled community-science relations:

When the numbers change, you know, people want the numbers to change in a really good way. Because it’s them and it’s their families and it’s their people and they like want them to be good numbers. Now, nobody wants a bad, they don’t want bad numbers, they don’t want numbers to stay bad, their interest isn’t like ‘let’s keep people sick’ or anything like that. But they want people to be treated fair, and to be treated right, and to be done justice in the response. And they want to trust the thing. And I think part of this change process, it’s diminished the trust between the community and the numbers. (P28; emphasis added)