Click here to go see the bonus panel!

Hovertext:
I expect this comic to be amusing until roughly 2029.

Grant Negotiation and Authorization Protocol (GNAP)

The most interesting difference between GNAP and OAuth 2 is that GNAP no longer requires clients to be registered in advance. With OAuth the client_id and client_secret need to be configured for each application, which means applications need to register with their targets - creating a new application on GitHub or Twitter before implementing the authorization flow, for example.

With GNAP that's no longer necessary. The protocol allows a client to provide a key as part of the first request to the server which is then used in later stages of the interaction.

GNAP has been brewing for a long time. The IETF working group was chartered in 2020, and two of the example implementations (gnap-client-js and oauth-xyz-nodejs) last saw commits more than four years ago.

Via lobste.rs

Tags: rfc, oauth, security

[Apologies! Instead of the short Friday joke post, I found myself writing something that’s a bit more “from the headlines” than I usually like to do.  But I do think it gets back to the main thread of this ‘stack toward the end; it also updates my views since the “mean reversion pixies” post.  Normal service will be resumed presently, hopefully, to the extent that “normal” means anything round here]

I’ve written a bit already about how much I don’t envy our current Chancellor, Formerbankofenglandeconomist Reeves. (Rachel is her middle name).  As far as I can tell, she’s painted absolutely into a corner – there is no way to pass a non-disastrous budget which doesn’t breach at least one of her pre-election commitments.

Subscribe now

My sketch argument for this is that:

The criteria for “non-disastrous” are a) it has to do something about a backlog of deferred public sector investment which has now become critical, and b) it should not cause a jump up in interest rates which would give our opinion commentators a pretext to compare it to the Truss/Kwarteng mini-budget of 2022.

The relevant pre-election commitments consist of some form of debt rule, plus the promise not to raise any of the big four taxes (income tax, corporation tax, VAT and national insurance).

Of those commitments, all the current attention seems to be on the first, as the government is apparently planning on changing the basis of the rule so that it applies to some measure which recognises public sector assets and treats the Bank of England’s QE program unwind in a less stupid and annoying way.  As I’m the third most important accountancy influencer in England, I’m obviously in favour of better rather than worse accounting rules.  But I actually think this might be a bit of a red herring.

The problem is that gross or net, debt stocks don’t really matter much – that was true at the beginning of austerity in 2010 and it’s true now.  The flows of spending and taxation are the real purpose of fiscal policy and the balance of outstanding gilts to which the “rule” applies is just their cumulated residual.  And we’re close to full employment; this isn’t a situation like 2010 in which demand needs to be stimulated.

So, to put it in “modern monetary theory”-adjacent terms, to a first approximation, whatever money the Reeves budget puts into the economy through spending, it has to take out with taxation.  Which means that if she wants to carry out any investment program that people will notice, it’s going to need to have a counterpart[1] tax rise that people will also notice.

Makes sense to me.  Then the Director of the NIESR challenged me:

My initial response was something along the lines of “come off it, that might work in a DGSE New Keynesian model, but if you’re trying to do £20bn of deferred maintenance on school and hospital buildings, plus HS2 rail, plus more than doubling housebuilding, are the construction workers really going to look ahead to future productivity gains and moderate their wage demands?”.

Which I will still mostly stand behind, but when you’re disagreeing with an economist of the calibre of Jagjt Chadha, you had better bring a bit more than “come off it mate”.  And it was also pointed out to me that the precise thing I’m saying is impossible – “a big investment-driven spending program financed by borrowing in an economy close to full employment without inflation or bond yields rising too much” – is pretty much what happened in the USA with the Inflation Reduction Act.

So let’s reconsider.  The IRA experience was one of transitory (yep, that word) inflation, which resulted from bottlenecks in supply, and which was resolved as productivity recovered.  So my hydraulic-Keynesian accounting identity might win the battle but lose the war, particularly as the bond market, for the most part, correctly looked through the process and didn’t lose its mud.

Could this happen in the UK too? Hmmm.  It seems like the key to avoiding a Truss/Kwarteng debacle is to have a coherent story that the bond market can understand about how your budget is going to end up with the economy in aggregate supply/aggregate demand equilibrium at a higher level of output. Liz Truss didn’t have a coherent story about anything, so she really did have to balance the equation and her failure to do so reset inflation expectations, causing the system to try to rebalance itself in a really chaotic way, which was thankfully and quickly stabilised by the political system doing its job and getting rid of her.

Reeves might not be in the same place, though, as her story looks more like the Biden IRA and less like a randomly assembled bunch of libertarian cliches.  So … this boils down to a question that’s in my view much more like a cybernetic problem of handling variety.  Does the bond market, considered as a system, have enough bandwidth to absorb and process the Reeves plan, or is it in a state in which it’s likely to spook over the initial spike in inflation as the investment demand temporarily imbalances things and leads to wage increases?

As Chris Boardman says, maybe.[2]

I think this “cybernetic approach to stabilisation policy” might become a lot more important going forward.  As James Meadway keeps writing, the world we are going into is going to have a lot more temporary shocks and bottlenecks in it, caused by geopolitics, climate events and all that sort of thing.  The aggregate supply potential of the economy is never a fixed point, but it’s going to have a lot more volatility. 

So a big part of the task of fiscal and monetary policy is going to be the identification of what’s a temporary bottleneck to power through, what’s a structural change that requires structural response, and what’s just a sign that you ran the economy too hot.  I don’t know what the appropriate control framework is for that sort of problem, but I don’t think it’s going to look anything at all like the current Inflation Report model.

[1] Word chosen carefully here – I originally typed “be funded by”, but this is loose and bad language.  The relationship between spending and taxation isn’t well described by the word “funding”. 

[2] A wonderful proverb I found while researching The Brompton. On one hand, if you can’t keep your pace up in a time trial, you’re in trouble. On the other, nobody wins a Tour de France stage while in their comfort zone. So Boardman always used to try to make sure that when he asked himself “can I keep going like this?” the answer was “maybe”.

Kids today probably don’t know about punched cards, but when I was at school we all had to play around with them a bit as we learnt about state-of-the-art computing …. But the technology derives from weaving, and from the Jacquard loom of 1804.

Click here to go see the bonus panel!

Hovertext:
I am prepared for your call Hollywood. It's time.