Author | Lightnews

Policy Sketchbook @policysketch.bsky.social · 2d

I wonder if a useful heuristic here is centralised vs decentralised. The centralised infrastructure of datacentres is less likely to have consumer surplus than the decentralised infrastructure of e.g. fibre optics, mass internet access. So what if anything is decentralised in AI infrastructure?

1 2

Policy Sketchbook @policysketch.bsky.social · 2d

Oh so it's The Spectator then.

1

Policy Sketchbook @policysketch.bsky.social · 3d

As a result, the system is good at managing and averting crises, but much less good at doing other things like improving productivity, improving experience, changing service models, planning ahead etc.

2 26

Policy Sketchbook @policysketch.bsky.social · 3d

I think this observation gets to the heart of something. The NHS is run as a crisis service, with everyone right up to the chief executive of the hospital overwhelmingly focused on managing and averting crisis over the next week or two.

1 8 66

Reposted by Policy Sketchbook

Sam Freedman @samfr.bsky.social · 3d

New post just out:

"On the edge"

On my unexpected week in hospital and five things I learnt about the state of the NHS.

(Free to read)

open.substack.com/pub/samf/p/o...

On the edge

My week in the NHS

open.substack.com

76 170 400

Policy Sketchbook @policysketch.bsky.social · 5d

It's almost like... this is the point of a narrow tax base and heavily means-tested benefits.

Chaminda Jayanetti @cjayanetti.bsky.social · 5d

And ofc governments themselves have significant influence over how much tax people pay

George Osborne bragged that his decision to raise the personal allowance meant loads of low-income people wouldn't pay income tax

Years later, that became a stick to beat them samf.substack.com/p/the-someth...

Article text:

“How nearly HALF British adults pay NO income tax: HMRC data reveals a record 23 million leave 31m to foot bill for running the country,” bellowed the Mail in response. “Increases in the personal tax allowance since 2010 mean 43% now live tax free.” The Sun joined in, albeit more circumspectly, perhaps realising this was the result of Budgets they’d cheered on.

“There are fewer adults than ever paying the main tax that props up every public service,” followed up Emma Barnett on her Radio 5Live call-in show, asking listeners: “If you don’t pay as much into these services like schools, should you get as big a say on how those services are run?”

2

Policy Sketchbook @policysketch.bsky.social · 10d

Is the issue not that things are linear when you look backwards, but that doesn't mean you can draw a straight line forwards and know where it is going?

1

Policy Sketchbook @policysketch.bsky.social · 29d

Along similar lines, this is an oldie but a goodie:
slatestarcodex.com/2014/12/12/b...

Beware The Man Of One Study

Aquinas famously said: beware the man of one book. I would add: beware the man of one study. For example, take medical research. Suppose a certain drug is weakly effective against a certain disease…

slatestarcodex.com

1 2

Reposted by Policy Sketchbook

Toby Nangle @tobyn.bsky.social · Sep 18

It's been a rough two decades for non-pensioner real income growth after housing costs, especially for the poorest households. www.resolutionfoundation.org/publications...

2 20 39

Policy Sketchbook @policysketch.bsky.social · Sep 1

I don't think so either, but there is a line of argument (from the usual suspects) that says that rising demand means we simply cannot continue.

And to be fair, this is also the mainstream narrative in health policy! "Our current model is simply unsustainable because of an ageing population" etc.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

If that social contract will never again be affordable then I guess we need to tear it up. If we are dealing with a temporary hump then, if we value it, we should protect it and work out how to get through the next 15 years.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

That's true, but how you deal with an issue will depend on how long we think it is going to last. For example, is it worth tearing up social contract around things like the NHS, which took a long time and a lot of political will to build, for a temporary problem?

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

Yeah it's quite a long time, but it takes a long time to reshape the state, and if we talk ourselves into thinking things can never again work because we are living longer, then by the time we've reshaped the state according to that a decent chunk of the problem may have passed us by.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

So we need to keep our heads with this a bit. It's not a terminal ageing problem for our health system, it's a temporary cohort effect we need to get through.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

You can see the levelling off of life expectancy in the 2010 there. But unless life expectancy is going to fall, an increased death rate in the population is only ever going to be a temporary cohort effect.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

This illustrates some of the points I was making yesterday. Mortality rates have been falling consitently for decades. Fewer people are in the last two years of life, where most healthcare usage happens. And yet... healthcare usage has been going up.

1

Policy Sketchbook @policysketch.bsky.social · Sep 1

I haven't looked at the actuarial tables behind this, but it is interesting because as I previously mentioned, in a population that lives longer, the death rate is lower by definition. So is this about life expectancy going into reverse, or is it the baby boomer cohort coming to the end of life?

1

Policy Sketchbook @policysketch.bsky.social · Aug 31

That said, the consensus in the field is (or was, with the caveat that I haven't been involved for a few years) pretty much as I describe.

1 2

Policy Sketchbook @policysketch.bsky.social · Aug 31

Obviously this is all hard, as we are talking about complex systems and an even more complex population, but I think it is fair to say we don't really have super convincing answers on what is really driving health care costs.

1 4

Policy Sketchbook @policysketch.bsky.social · Aug 31

But the questions you ask are the right ones and my personal view is that this area of study is a bit of a mess. It's hard to get clear answers, for example, when you as "why are emergency department attendances increasing around the world?"

1 3

Policy Sketchbook @policysketch.bsky.social · Aug 31

This tends to be a pretty basic analysis based on constant cost over time for any given age. I was pointing out that actually there is some evidence that could even overestimate pure demographic effects as some costs go with proximity to death rather than age.

1 1

Policy Sketchbook @policysketch.bsky.social · Aug 31

I'm not saying what's right, just what the consensus seems to be among people who work on this. As @cassiarowland.bsky.social was saying, the OECD and others say demographics increase costs, but it's only a small effect relative to observed cost rises.

1 1

Policy Sketchbook @policysketch.bsky.social · Aug 31

This part is true - again, as long as our ability and obligation to work is predicted by age rather than time to death. But that's about the dependency ratio in general and not health care costs per se.

1 1

Policy Sketchbook @policysketch.bsky.social · Aug 31

This generalises to other population distributions. I'm not saying that proximity to death predicts all healthcare costs, but there are papers arguing it is an important predictor, and insofar as it is, demographic pressure should in theory be negative.

1 1

Policy Sketchbook @policysketch.bsky.social · Aug 31

No I'm talking about stocks and flows. Think about a stylised case where everyone dies at the same age, so the population is a rectangular distribution. The older the age of death, the smaller the outflow relative to the stock and the last two years are a smaller proportion of the whole.

1 1