Thread by @robfordmancs

So, I've learned a lot about how the term "progressive" is understood on the left in the last 12 hours or so, which has led me to rethink this. I still think there's a lot of partisan filtering going on here, but its the ambiguities in NI which facilitate it.

https://twitter.com/robfordmancs/status/1435674563489185795

I had assumed that "progressive" with regards fiscal policy was generally understood as "redistributive" (i.e. a tax whose net effect is rich pay more, poor pay less). At the household level, that's clearly true of NI as graph below illustrates

https://twitter.com/jdportes/status/1435649903145926666?s=20

However, from my replies it seems there are at least three other definitions of "progressive" taxation: 1. A tax whose rates rise like a staircase across the income distribution 2. A tax where richest pay most 3. A tax whose impacts fall least on the poorest

Also, its clear from the tone of the replies I've had that "progressive" is for most people a normative, emotive term not a technical, analytical term - more like "green" than "energy efficient". So people get v upset if something is called progressive which they think isn't.

On top of this, the nature of NI as a tax is ambiguous. The first, substantial, chunk of income is exempt, then its charged at a pretty flat rate, until at the v top when (weirdly) a lower rate is charged. So its correct to say it isn't a "progressive" tax wrt staircase rates

But in terms of *effective* rates it *is* progressive over most of the income distribution (as Portes' graph illustrates). Someone on 10k pays it on virtually none of their income. Someone on 20k pays a single rate on 10k. Someone on 40k pays a single rate on 30k.

Many then say "a flat tax can't be progressive". If your definition is about tax structure, sure. But in terms of resource distribution, surely not. Imagine a world where incomes below £100k weren't taxed at all, incomes above £100k were taxed 90%. Not progressive?

Another way to look at this is to consider different criteria of "progressive" for NI:
1. Do the poorest pay least? Yes, they pay nothing
2. Do richer households pay more on average than poorer households? Yes, in both absolute and % terms
3. Do the *richest* pay *most*? No

4. Who is hit hardest by the impact of the tax? Not the poorest, evidently, because they don't pay it. But I think its reasonable to say the income hit will be harder for those in (e.g.) the bottom 20% to absorb than further up.

So there are intuitive definitions of "progressive" which NI meets, and intuitive definitions of "progressive" which NI doesn't meet. I was incorrect, therefore, to say NI is "clearly" progressive. Its not "clearly" anything!

The point I was trying to make, however, was about partisan filtering of information. In a few days, it seems this policy has become a lot less popular. As @anthonyjwells shows, this is driven by Labour voters turning against it:

https://twitter.com/anthonyjwells/status/1435646250985082881?s=20

Have Labour people turned against a massive funding increase for the NHS because of the flaws in the NI rise used to finance it? Perhaps. Or perhaps Labour partisans saw a Con policy as inherently suspect, started looking for reasons to dislike it, and soon found one.

If only there was a counter-example of a Labour govt funding a massive NHS spending increase using a rise in NI, which we could use for comparison...

theguardian.com/politics/2002/…

If the partisan filtering hypothesis is correct, we wld expect a v similar policy to be much more popular with voters because Lab voters would focus on the bits they like (more money for NHS) rather than looking for things to oppose. That's what we find: ipsos.com/ipsos-mori/en-…

NI was no more progressive in 2002 when Brown raised it - given changes in personal allowances I'm guessing it was much *less* progressive. But Labour partisans were much keener on "NI rise for NHS" as a Labour policy in 2002 than as a Con one now.

All of which serves to remind us that the common belief that people judge parties based on their policies is often wrong. People are just as prone to judge policies based on the parties introducing them - and partisans even more so.

(this tweet somehow seems to have dropped off the beginning of the thread - quite important as sets out my original assumption re: "progressive" tax)

https://twitter.com/robfordmancs/status/1435879843347017729?s=20

(As a postscript, this interesting point from Duncan Weldon that contempory left understanding of "progressive" seems to focus much more on impacts on the very richest than was true a few decades ago):

https://twitter.com/DuncanWeldon/status/1435670666209026059?s=20