Austerity has killed at least 334,000 people so far

[lpm]

Decreasing life expectancy is a slightly different metric. A few deaths of the young has a bigger impact than lots of deaths of the old.

To be brutal about it, economically it's better for average life expectance to fall from 80 to 79 via all pensioners dying one year earlier, than via pensioners living to the same age but an increase in mortality for <65. The policy answers are different - e.g. investment in end of life care vs investment in mental health.

[Fishnut]

The Conversation has a piece on the research, putting it into some context. A few quotes,

Before this new study, some disputed the idea that austerity could be blamed for the increase in deaths by pointing out that most of those who had died prematurely were old and so had benefited from the “triple lock” of the UK state pension... Between 2010 and 2020 the average UK pensioner saw their real-terms weekly income (after housing costs) rise by only £12, to £331 a week. That represents a meagre 3.8% rise over the whole decade, which works out to £1.71 extra per day. This in no way compensates for rising fuel and other costs that especially hurt poorer pensioners...

Poorer pensioners and those in most need were also most harmed by cuts to state services. They lost their adult social worker and carer visits, local government help, and so much else that existed in 2010 – but was largely gone by 2019...

in August 2022 the Financial Times published estimates suggesting a large proportion of recent non-COVID-related deaths could be ascribed to just one aspect of austerity: waiting over 12 hours to be seen in accident and emergency departments.

[Millennie Al]

As noted earlier, in a country with the wealth of the UK, the extent of poverty is a political choice.

The definition of poverty is also a political choice.

Don't give them ideas.

I don't need to. Different definitions are already in use.

Clearly there are a variety of distinct concepts you can stick the word "poverty" onto. And within each concept, typically you also have to choose a specific numerical level.

We use a best practice definition of poverty in this country, suitable for the situation of a developed country, based on the median income. That's a concept of poverty that is more closely related to an idea of inequality and social exclusion, rather than a concept based on survival. Bare survival should not be an issue in a developed country,...

This article is referring to poverty as the cause of many deaths. Therefore it should use a definition which is appropriate for that usage. Absolute poverty causes deaths. Relative poverty doesn't (unless people feel so excluded that it leads to suicide or they have such an immense urge to be richer that they do very dangerous things to make money - but neither should be solved by reducing relative poverty).

Furthermore, there are obvious flaws in the article, which is really a piece of political advocacy rather than a scientific paper. Even if the data were correct and correctly analysed, it's assuming correlation is causation.

[IvanV]

This article is referring to poverty as the cause of many deaths. Therefore it should use a definition which is appropriate for that usage. Absolute poverty causes deaths. Relative poverty doesn't (unless people feel so excluded that it leads to suicide or they have such an immense urge to be richer that they do very dangerous things to make money - but neither should be solved by reducing relative poverty).

Furthermore, there are obvious flaws in the article, which is really a piece of political advocacy rather than a scientific paper. Even if the data were correct and correctly analysed, it's assuming correlation is causation.

I think you are underestimating the effect of income on health and lifespan in this country. In reality, we see really large life expectancy differences between higher and lower income areas in this country. Of the order of 10 years (BMJ). The same source notes differences in extent of life lived in good health can be as large as 19 years between the most deprived groups and others.

So, that's the data, showing really large differences in health and life expectancy between different income groups in this country. How do we explain it? It's clearly not to do with access to insufficient food, as the deprived often have high rates of obesity these days. I think I saw something recently suggesting it wasn't much to do with access to health care, or health care of different qualities. So what is it about? I would very much doubt it is to do with the causes you mention. Perhaps you can track down some literature and come back and tell us.

[Millennie Al]

Absolute poverty causes deaths. Relative poverty doesn't

I think you are underestimating the effect of income on health and lifespan in this country. In reality, we see really large life expectancy differences between higher and lower income areas in this country. Of the order of 10 years (BMJ). The same source notes differences in extent of life lived in good health can be as large as 19 years between the most deprived groups and others.

What makes you think that this is due to relative rather than absolute poverty? Note that if life expectancy is reduced by relative poverty and you impoverish the richer people down to the level of the poorest, the average life expectancy will rise. However, if the cause is absolute poverty, it will fall. It seems highly implausible that relative poverty is responsible.

So, that's the data, showing really large differences in health and life expectancy between different income groups in this country. How do we explain it? It's clearly not to do with access to insufficient food, as the deprived often have high rates of obesity these days. I think I saw something recently suggesting it wasn't much to do with access to health care, or health care of different qualities.

If someone was claiming that Covid vaccination made people's hair turn blue, it would be dismissed as having no reasonable mechanism and that therefore any correlation must be due to something else. But when it's about inequality, people don't seem to care abut explaining the underlying mechanism - presumably because they want it to be true and don't want to look too closely.

So what is it about? I would very much doubt it is to do with the causes you mention. Perhaps you can track down some literature and come back and tell us.

That's no small job. Especially if people have been ignoring causative mechanisms so have not been collecting the relevant data.

[Bird on a Fire]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

[plodder]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

Is this an invocation of the predictive powers of ecology, or economics?

[IvanV]

Absolute poverty causes deaths. Relative poverty doesn't

I think you are underestimating the effect of income on health and lifespan in this country. In reality, we see really large life expectancy differences between higher and lower income areas in this country. Of the order of 10 years (BMJ). The same source notes differences in extent of life lived in good health can be as large as 19 years between the most deprived groups and others.

What makes you think that this is due to relative rather than absolute poverty?

It's clearly more complicated than observing that these people live a long way above the absolute poverty lines of less developed countries, where starvation, shelter, etc have direct effects on life expectancy. Rather, we can expect life expectancy to increase with income level for some distance above those minimum income levels, and hence be related, over some interval, to absolute income levels. So, the question is, are these people still living within that band of incomes where you would expect absolute income to be driving life expectancy, or are they above that.

I think it fairly obvious that they are well above that. And that is because we see people living longer lives in many other countries at lower income levels than the relatively deprived communities in Britain.

Countries like Greece and Portugal have higher life expectancies than all of Britain. They must have many more people at income levels at or below those of British relatively deprived communities. Countries with really much lower incomes, like Thailand, Sri Lanka and Ecuador, have life expectancies below the UK average, but still much higher than the relatively deprived sections of British society. These countries are so much poorer than Britain, that most of the population must have incomes below those of the relatively deprived in Britain.

Note that if life expectancy is reduced by relative poverty and you impoverish the richer people down to the level of the poorest, the average life expectancy will rise. However, if the cause is absolute poverty, it will fall. It seems highly implausible that relative poverty is responsible.

I don't think anyone would expect that chopping a large amount of income out of the better off all of a sudden is going to do very much in the short run. These things establish themselves over the longer term. I would suspect that they would be related more immediately to the structure of society, and more slowly to income levels.

[Bird on a Fire]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

Is this an invocation of the predictive powers of ecology, or economics?

Not predictive powers, but pattern-spotting abilities. Economies and ecologies are both examples of complex systems, so there are similarities.

[plodder]

I think you are underestimating the effect of income on health and lifespan in this country. In reality, we see really large life expectancy differences between higher and lower income areas in this country. Of the order of 10 years (BMJ). The same source notes differences in extent of life lived in good health can be as large as 19 years between the most deprived groups and others.

What makes you think that this is due to relative rather than absolute poverty?

It's clearly more complicated than observing that these people live a long way above the absolute poverty lines of less developed countries, where starvation, shelter, etc have direct effects on life expectancy. Rather, we can expect life expectancy to increase with income level for some distance above those minimum income levels, and hence be related, over some interval, to absolute income levels. So, the question is, are these people still living within that band of incomes where you would expect absolute income to be driving life expectancy, or are they above that.

I think it fairly obvious that they are well above that. And that is because we see people living longer lives in many other countries at lower income levels than the relatively deprived communities in Britain.

Countries like Greece and Portugal have higher life expectancies than all of Britain. They must have many more people at income levels at or below those of British relatively deprived communities. Countries with really much lower incomes, like Thailand, Sri Lanka and Ecuador, have life expectancies below the UK average, but still much higher than the relatively deprived sections of British society. These countries are so much poorer than Britain, that most of the population must have incomes below those of the relatively deprived in Britain.

Note that if life expectancy is reduced by relative poverty and you impoverish the richer people down to the level of the poorest, the average life expectancy will rise. However, if the cause is absolute poverty, it will fall. It seems highly implausible that relative poverty is responsible.

I don't think anyone would expect that chopping a large amount of income out of the better off all of a sudden is going to do very much in the short run. These things establish themselves over the longer term. I would suspect that they would be related more immediately to the structure of society, and more slowly to income levels.

These other countries with better life expectancy: how has their “austerity” or lack of it (needs defining) changed their life expectancy relative to ours? Al is correct here, “researchers” aren’t bothering to properly explore this stuff because they already “know” the answers.

[plodder]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

Is this an invocation of the predictive powers of ecology, or economics?

Not predictive powers, but pattern-spotting abilities. Economies and ecologies are both examples of complex systems, so there are similarities.

There are similarities in that neither is any good at pinning down causation.

[Gfamily]

“researchers” aren’t bothering to properly explore this stuff because they already “know” the answers.

Sounds like you're a "...people are tired of experts..." fan.
So it should be easy for you to demonstrate (with data) that there's no causation behind the correlation.

[IvanV]

Al is correct here, “researchers” aren’t bothering to properly explore this stuff because they already “know” the answers.

And you know this because? You have no knowledge of what researchers have done or not done in this area.

We were recently referred to that excellent twitter thread with 40 important concepts. I forget the name for that very useful thesis that proposals without evidence can be rejected without evidence. It is not for the mainstream position to be forever repeating its evidence again to those who can't be bothered finding it out, when they present ideas that contradict it without evidence.

I've sketched the outlines of some standard positions. I have given some broad reasons as to why those are the standard positions, and why they have some plausibility.

I can sometimes take my education a bit further. But this is not a literature I know well, so I am afraid I cannot quickly acquaint you with it.

So if you or Al want to throw some stones, it really is your job to find out what you might be throwing stones at before chucking them from a position of ignorance.

[Bird on a Fire]

Is this an invocation of the predictive powers of ecology, or economics?

Not predictive powers, but pattern-spotting abilities. Economies and ecologies are both examples of complex systems, so there are similarities.

There are similarities in that neither is any good at pinning down causation.

Shots fired!

I think there's possibly some truth in that, at least with a restrictive view of "pinning down causation": the systems are too large to be contained within a controlled experiment, and macro patterns can be difficult/impossible to predict from micro relations.

So you tend to end up with explanations that have a lot of moving parts, and are somewhat context dependent. Even then a well -evidenced reckon can help inform a cautionary approach to interventions.

Empirically understanding the impacts of top-down policies like austerity should be fairly straightforward, in that there are lots of countries with different responses to lots of crises. You just have to control for the fact that every case is unique.

[Millennie Al]

Countries like Greece and Portugal have higher life expectancies than all of Britain. They must have many more people at income levels at or below those of British relatively deprived communities. Countries with really much lower incomes, like Thailand, Sri Lanka and Ecuador, have life expectancies below the UK average, but still much higher than the relatively deprived sections of British society. These countries are so much poorer than Britain, that most of the population must have incomes below those of the relatively deprived in Britain.

Well, while that may show it is unlikely to be absolute poverty, it doesn't show it being relative poverty either. Looking up the List of countries by income equality we find this ordering by Gini coefficient with life expectancy from the first table in List of countries by life expectancy.

1. 0.328 Portugal LE=80.98
2. 0.331 Greece LE=81.09
3. 0.35 Thailand LE=77.34
4. 0.351 UK LE=80.90
5. 0.393 Sri Lanka LE=77.14
6. 0.473 Ecuador LE=77.22

So Sri Lanka and Ecuador have worse relative poverty and worse absolute poverty than the UK, so their life expectancies should be terrible. Their poor people should be it even harder by relative poverty than the UK and also hit by the absolute poverty. In particular, Sri Lanka and Ecuador have very similar life expectancy but significantly different Gini coefficients. The other way around, Thailand and the UK have veyr similar Gini coefficients, but quite different life expectancies, suggesting that either relative poverty is unimportant or at least that other factors are much more important.

Note that if life expectancy is reduced by relative poverty and you impoverish the richer people down to the level of the poorest, the average life expectancy will rise. However, if the cause is absolute poverty, it will fall. It seems highly implausible that relative poverty is responsible.

I don't think anyone would expect that chopping a large amount of income out of the better off all of a sudden is going to do very much in the short run. These things establish themselves over the longer term. I would suspect that they would be related more immediately to the structure of society, and more slowly to income levels.
[/quote]

I made no suggestion that the effect would be sudden. But you have still failed to come up with any plausible mechanism by which relative poverty could reduce life expectancy.

[dyqik]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

Is this an invocation of the predictive powers of ecology, or economics?

No, it's not. At least, it's not specific to ecology or economics.

Tipping points exist in most systems beyond a certain level of complexity and non-linearity. System analysis shows this - instabilities occur in most systems, complexity theory is the study of this, and the butterfly effect in chaos theory is an extreme example of a tipping effect.

See e.g. http://www.complexity.soton.ac.uk/theor ... Points.php

[plodder]

Basic stuff.

Complex systems often have tipping points and non-linear responses. It's entirely possible that absolute poverty matters up to a point, after which the impact of relative poverty is greater.

And you don't need to understand mechanisms to identify patterns.

Is this an invocation of the predictive powers of ecology, or economics?

No, it's not. At least, it's not specific to ecology or economics.

Tipping points exist in most systems beyond a certain level of complexity and non-linearity. System analysis shows this - instabilities occur in most systems, complexity theory is the study of this, and the butterfly effect in chaos theory is an extreme example of a tipping effect.

See e.g. http://www.complexity.soton.ac.uk/theor ... Points.php

I mean this is just junk when we’re talking about specific policy decisions over a relatively short length of time

[lpm]

But you have still failed to come up with any plausible mechanism by which relative poverty could reduce life expectancy.

Relative poverty makes people feel like losers.

Drug takers are predominantly losers.

Drug mortality includes young people so has outsized impact on the life expectancy stat.

[plodder]

But you have still failed to come up with any plausible mechanism by which relative poverty could reduce life expectancy.

Relative poverty makes people feel like losers.

Drug takers are predominantly losers.

Drug mortality includes young people so has outsized impact on the life expectancy stat.

Lovely, I’m sure that’s clearly identified in the data

[lpm]

Yes, of course it is. Look at any of the data and for people above 50 it's the usual heart disease and cancer. But below 50 and the mortality is suicide and drug poisoning.

The other diseases of poverty - smoking and alcohol - will kill more by numbers but tend to kill at over 50, so don't have such a magnified impact on life expectancy.

[IvanV]

Well, while that may show it is unlikely to be absolute poverty...

Good step forward!

...it doesn't show it being relative poverty either...

I shall look forward to further steps in the investigation of this.

[EACLucifer]

Is this an invocation of the predictive powers of ecology, or economics?

No, it's not. At least, it's not specific to ecology or economics.

Tipping points exist in most systems beyond a certain level of complexity and non-linearity. System analysis shows this - instabilities occur in most systems, complexity theory is the study of this, and the butterfly effect in chaos theory is an extreme example of a tipping effect.

See e.g. http://www.complexity.soton.ac.uk/theor ... Points.php

I mean this is just junk when we’re talking about specific policy decisions over a relatively short length of time

As opposed to what, your chinny reckons and what that bloke at the pub who won't ever shut up had to say about it? Because that's the quality you are bringing to this thread right now.

[Bird on a Fire]

Is this an invocation of the predictive powers of ecology, or economics?

No, it's not. At least, it's not specific to ecology or economics.

Tipping points exist in most systems beyond a certain level of complexity and non-linearity. System analysis shows this - instabilities occur in most systems, complexity theory is the study of this, and the butterfly effect in chaos theory is an extreme example of a tipping effect.

See e.g. http://www.complexity.soton.ac.uk/theor ... Points.php

I mean this is just junk when we’re talking about specific policy decisions over a relatively short length of time

I was responding to the discussion around Al's assertion that "Absolute poverty causes deaths. Relative poverty doesn't."

The discussion is clearly broader than specific policy decisions over a relatively short length of time because Al is now looking at relative inequality in developing countries. I'm sure he doesn't think the coalition government did things that made UK society similar to Thailand or Ecuador - he's still grappling with the basics.

[Bird on a Fire]

The actual paper, as you'd expect, does delve into mechanisms in quite some detail.

From the introduction:

Although a number of different contributory factors were initially proposed, a considerable body of evidence now demonstrates that UK Government’s ‘austerity’ policies are the main cause of these pre-pandemic changes.12–17 This includes a recently published, large-scale, critical assessment of all the relevant evidence.18 Such policies—introduced from 2010 onwards, and following ‘the great recession’ of 2008—have removed billions of pounds from both social security and vital services, and have thus particularly impacted on poorer, more vulnerable, populations.10 15 18 Similar adverse effects of austerity measures on population health have been seen in other high-income countries.19–22

There is also evidence that the reductions in social security income and loss of services have disproportionately affected women in the UK. This is for a number of important reasons including: more women being in receipt of social security payments in the first place; the disproportionate effects of cuts on particular, female-dominated, groups such as lone parents; the contraction in public sector jobs where women are likely to be employed; and inequalities in caring responsibilities (and the associated need for local government services and social care in particular).23–29 Among the elderly, the fact that more women live alone and are unable to share financial burdens may also be relevant.30

(All those numbers are references to other studies providing evidence for the statement in question.)

Further discussion of mechanisms, from the Discussion:

Although our analyses have produced mixed results in terms of the main hypothesis under investigation, the worse mortality experience of women living in the most deprived neighbourhoods might well reflect a more adverse, and gendered, financial impact of government policy. UK Government’s austerity cuts to social security payments and eligibility, as well as to important services, have been shown in a number of analyses to affect women more than men. This is for a number of reasons. First, women make up the majority of social security recipients in the UK, and have thus been more affected by the enormous cuts in spending in that area;23–26 40 41 that said, some of this has been contested as it can be difficult to disentangle impacts on an individual’s income from that of the broader household.27 28 Second, some individual social security cuts have disproportionately affected particular groups: lone parents (the vast majority of whom are female); single pensioners (more of whom are women) and people with disabilities (many of whom are looked after by female carers).23–27 40 41 Third, women are more likely to be poorer and on lower incomes—this is linked to inequalities in the labour market and in the division of caring responsibilities—and the cuts have been shown to be highly regressive, impacting most on those low-income groups.23–26 Fourth, as caring responsibilities (eg, for children, elderly and disabled people) are undertaken much more by women, they are not only affected by reductions in income, but also by cuts to, and loss of, public services that support such caring activities.25 26 40 41 Fifth, as the majority of public sector employees (especially part-time employees) in the UK are female, women are more likely to have been affected by both the public sector pay freezes that were introduced, and by public sector job losses.25 26 40 A range of other factors have been cited including the roll-out of the ‘Universal Credit’ social security benefit (as it disincentivises second income earners in a household, the vast majority of whom are female),40 cuts to legal aid/advice which have discouraged the challenging of discriminatory employment practices affecting women (eg, pregnancy/maternity discrimination),41 and cuts to specific services such as those for young children and in relation to gender-based violence.25 26

Analyses of austerity-related trends in other health outcomes also present a mixed picture in terms of differences between sexes. Healthy life expectancy (which combines mortality data with information on self-assessed general health) decreased by 2 years in Scotland following implementation of austerity, but the decline was similar for men and women.45 However, Thomson et al showed that following the introduction of austerity, levels of poor mental health among those aged 25–64 years old in England worsened to a greater degree among women than men46; and although trends were only analysed up to 2014, subsequent analyses of the same data over a longer period (to 2018) showed worsening mental health among both men and women, but still to a greater degree among women.47 In contrast, Wickham et al, investigating the effect of one component of austerity on mental health—the introduction of the ‘Universal Credit’ social security benefit—showed similar adverse effects for men and women.48

The more adverse recent trends in premature (0–64 years) mortality observed in Scotland compared with England (for both men and women) shown in online supplemental figures 1 and 2 are largely explained by greater increases in drug-related deaths in Scotland in this period18 50; however, those trends have themselves been exacerbated by the same UK Government’s austerity policies discussed here.15 18

So, we have demonstrated impacts on care, mental health and drug use. All predictably damaged by cuts to relevant services, and those services were implemented in the first place to reduce morbidity and mortality. Seems like a plausible causative pathway, backed up by lots of analysis. Might be a good place to start, rather than amateur epistemology hour.

[dyqik]

Is this an invocation of the predictive powers of ecology, or economics?

No, it's not. At least, it's not specific to ecology or economics.

Tipping points exist in most systems beyond a certain level of complexity and non-linearity. System analysis shows this - instabilities occur in most systems, complexity theory is the study of this, and the butterfly effect in chaos theory is an extreme example of a tipping effect.

See e.g. http://www.complexity.soton.ac.uk/theor ... Points.php

I mean this is just junk when we’re talking about specific policy decisions over a relatively short length of time

The thing about tipping points is that they cause rapid changes, often unexpectedly.

There's probably a whole bunch of examples within the Government's CoVID responses, for example.