BREXIT - Is This Really Still Rumbling On? 😴

Stevo_666

TheBigBean wrote:

In most cases, there are too many variables and unknowns to do it accurately which is the point I was making. It's (mathematical) chaos which is the point I suspect Stevo was making.

A quick google reveals various studies into traffic forecasts being so bad.

Correct - that and the point mentioned by Robert88 above that there is never enough information.

rjsterry

TheBigBean wrote:

rjsterry wrote:

TheBigBean wrote:

Surrey Commuter wrote:

TheBigBean wrote:

Surrey Commuter wrote:

TheBigBean wrote:

They are tricky e.g. HS2. There is no reliable way of forecasting the impact, so how do they make a decision? I'd imagine mostly someone just takes punt.

No they don’t.

Without wishing to go all pantomime, yes they do. If the economic benefits of HS2 were certain, they wouldn't be reviewing the whole thing having already spent billions.

I think you are being disingenuous suggesting "mostly somebody just takes a punt" as I am sure you are well aware that there were enormous studies done to assess the costs and benefits

Sure. Lots of studies which identify possible benefits and possible risks. How do you combine them all together? For example, what is the probability that cheap autonomous cars driving 10m apart are available by the time HS2 starts? What impact would this have on demand for HS2? No one can answer those questions, they just have to accept that there is a risk.

If you look at major PFI failures, they are nearly all transport projects where demand wasn't what people thought it would be.

Is that poor modelling of the demand or politicised interpretation of that modelling?

In the case of a lot of PFIs it was a private sector risk, so that removes the political element. There would have been an independent forecaster on whose recommendations the private sector relied. Clearly, there are also many examples worldwide where no doubt politics has played a part.

Poor modelling implies negligence and a PI claim. I'm sure that has happened too.

In most cases, there are too many variables and unknowns to do it accurately which is the point I was making. It's (mathematical) chaos which is the point I suspect Stevo was making.

A quick google reveals various studies into traffic forecasts being so bad.

I was very much thinking small p politics - which ambitious senior manager doesn't fancy a legacy.

I would argue that it's not that the results of projections necessarily become inaccurate but that the differentiation between likely and unlikely outcomes reduces the further into the future you look. At some point the results don't provide a useful prediction, but 'anything can happen' could still be an accurate prediction. If done well, the projections can give an indication of the relative impact of different factors, which is useful even if there is no one specific outcome predicted.

TheBigBean

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

TheBigBean

Tony Connelly
‏Verified account @tconnellyRTE
17m17 minutes ago

3/ On consent, I understand Mr Barnier said the UK is signalling a willingness to accept a simple majority endorsement in NI, rather than a veto for one party. Whether this is via a Stormont Assembly vote or even a referendum on exiting the new arrangements is not clear

daniel_b

Already seeing more red tape to clear.

Sending some clothing back to Bike24 (Which they are kindly paying for which I did not expect) and have been sent a customs form to put on the outside incase we leave without a deal, and it's still in transit.
Sigh.

rjsterry

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

Alejandrosdog

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

rjsterry

Of course not. Nobody has suggested that

Stevo_666

rjsterry wrote:

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

Have a look at the number of variables in complex real world systems where chaos theory applies - e.g. weather systems, financial markets. The permutations are mind boggling. Also part of this is sensitivity to initial conditions, where a seemingly insignificant difference at the start is amplified over time. And in economic models we are also trying to predict future human behaviour.

Fascinating stuff, but my own experience of forecasting in a business environment bears this out quite nicely in the real world.

TheBigBean

rjsterry wrote:

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

I think few economists (or perhaps only one) would disagree that being part of the EU is likely to be economically better than signing lots of new trade deals. I'm just quite uneasy when people start to put numbers on this especially at some arbitrary date far in the future.

In terms of specific criticism of the various models these are available if you google "Brexit gravity model". One easily understood criticism is that some include GDP growth related to population growth, but don't compare GDP/capita.

I largely stick the each family will £4k worse off in the same bucket as £350m a week - noise to win a referendum. The only difference being that the voting population still has limited faith in economists since 2007/8 crash, so tends to ignore them.

tailwindhome

Alejandrosdog wrote:

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

In fairness they all agreed it was the moon they were going to before they took off.

Stevo_666

TailWindHome wrote:

Alejandrosdog wrote:

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

In fairness they all agreed it was the moon they were going to before they took off.

Maybe if they had agreed that their mission was simply to leave Earth, that would make the two more comparable?

surrey_commuter

TheBigBean wrote:

rjsterry wrote:

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

I think few economists (or perhaps only one) would disagree that being part of the EU is likely to be economically better than signing lots of new trade deals. I'm just quite uneasy when people start to put numbers on this especially at some arbitrary date far in the future.

In terms of specific criticism of the various models these are available if you google "Brexit gravity model". One easily understood criticism is that some include GDP growth related to population growth, but don't compare GDP/capita.

I largely stick the each family will £4k worse off in the same bucket as £350m a week - noise to win a referendum. The only difference being that the voting population still has limited faith in economists since 2007/8 crash, so tends to ignore them.

the £4k per family is a nonsense, if being kind to simplify the message.

On a simplistic level economic modelling assumes that all else remains equal barring the variable (s) that you want to measure. Of course London could be hit by a meteorite in year 3 and make all the forecasts "wrong"

The models suggested that leaving the EU would reduce GDP by 0.5%pa. This seems reasonable but politicians then put it into numbers. It should also be noted that NI farmers going bust and car factories scaling back are all small blocks in the wall that make up this 0.5%

I suspect people only don't have faith in economists when they say something they don't want to here.

kingstongraham

Stevo 666 wrote:

TailWindHome wrote:

Alejandrosdog wrote:

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

In fairness they all agreed it was the moon they were going to before they took off.

Maybe if they had agreed that their mission was simply to leave Earth, that would make the two more comparable?

So we're more like Laika?

Stevo_666

What's that?

surrey_commuter

Stevo 666 wrote:

TailWindHome wrote:

Alejandrosdog wrote:

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

In fairness they all agreed it was the moon they were going to before they took off.

Maybe if they had agreed that their mission was simply to leave Earth, that would make the two more comparable?

I am liking this analogy. Maybe we could clone everybody and have two concurrent projects, one with a clearly defined destination and the other containing a bunch of individuals pursuing their own vision. You can imagine the problems when the Americans find that their landing module won't fit as the Germans have designed it to carry a warhead.

TheBigBean

NI being in a different customs territory seems to be against the law, so that idea will require some repealing.

Stevo_666

Surrey Commuter wrote:

Stevo 666 wrote:

TailWindHome wrote:

Alejandrosdog wrote:

The Apollo mission began in 1963 and ran till 1972

in that time we learned how to go to the moon and back again.

A Brexit deal is not outside the wit of man.

In fairness they all agreed it was the moon they were going to before they took off.

Maybe if they had agreed that their mission was simply to leave Earth, that would make the two more comparable?

I am liking this analogy. Maybe we could clone everybody and have two concurrent projects, one with a clearly defined destination and the other containing a bunch of individuals pursuing their own vision. You can imagine the problems when the Americans find that their landing module won't fit as the Germans have designed it to carry a warhead.

It also fits in neatly with the question put to the electorate in the 2016 referendum.

rjsterry

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

kingstongraham

rjsterry wrote:

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

per wikipedia: "the technology to de-orbit had not yet been developed"

Stevo_666

rjsterry wrote:

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

An early example of socialist inefficiency?

kingstongraham

Stevo 666 wrote:

rjsterry wrote:

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

An early example of socialist inefficiency?

Surely would have been more inefficient to plan to return her to earth.

These days the lefties would say we need to build a robot with a poop bag am I right?

Stevo_666

KingstonGraham wrote:

Stevo 666 wrote:

rjsterry wrote:

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

An early example of socialist inefficiency?

Surely would have been more inefficient to plan to return her to earth.

These days the lefties would say we need to build a robot with a poop bag am I right?

Or maybe incompetence.

drhaggis

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

That's deterministic chaos. Which doesn't quite work, because reality is quantum, and quantum theory is probabilistic. So even in perfeclty defined, well-behaved systems, you never really know*** what you will measure next.

kingstongraham

Stevo 666 wrote:

KingstonGraham wrote:

Stevo 666 wrote:

rjsterry wrote:

Stevo 666 wrote:

What's that?

The dog. First animal sent into space. She didn't make it back to Earth.

An early example of socialist inefficiency?

Surely would have been more inefficient to plan to return her to earth.

These days the lefties would say we need to build a robot with a poop bag am I right?

Or maybe incompetence.

It was early days. The USA sent a monkey into space who survived the forces of lift off and the 10,000mph re-entry into the atmosphere. Then the parachute on his module failed and he fell into the sea and sank.

Can't shoehorn Brexit into that I'm afraid.

tailwindhome

Steven Swinford
@Steven_Swinford
· 4m
The DUP is *not* necessarily opposed to a post-Brexit customs partnership with the EU

'Provided it doesn't mean that NI is excluded from being able to participate in new trade deals we could be supportive,' DUP source says

TheBigBean

TailWindHome wrote:

Steven Swinford
@Steven_Swinford
· 4m
The DUP is *not* necessarily opposed to a post-Brexit customs partnership with the EU

'Provided it doesn't mean that NI is excluded from being able to participate in new trade deals we could be supportive,' DUP source says

All sources pointing in the same direction. Looks like a deal.

surrey_commuter

TailWindHome wrote:

Steven Swinford
@Steven_Swinford
· 4m
The DUP is *not* necessarily opposed to a post-Brexit customs partnership with the EU

'Provided it doesn't mean that NI is excluded from being able to participate in new trade deals we could be supportive,' DUP source says

ahhh regulatory alignment rears it's head

drhaggis

Stevo 666 wrote:

rjsterry wrote:

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

Have a look at the number of variables in complex real world systems where chaos theory applies - e.g. weather systems, financial markets. The permutations are mind boggling. Also part of this is sensitivity to initial conditions, where a seemingly insignificant difference at the start is amplified over time. And in economic models we are also trying to predict future human behaviour.

Fascinating stuff, but my own experience of forecasting in a business environment bears this out quite nicely in the real world.

A system being chaotic doesn't mean you can't do accurate forecasts. See weather vs. climate. And if you don't like climate because your political bias, then look at paleoclimate. Yes, small noise will cause exponential divergence, but that may only be important if you care about the details (e.g. next Sunday the hailstorm is in Windsor rather than Kingston). You can account for this variability, to a point, by analysing a set of nearby initial conditions, which would give you a better idea of what to expect, and what the odds really are. What's more, even in chaotic systems, viral theorem applies, so the envelopes of solutions have to be within a certain set of parameters. There are also basins of attraction, which you may be able to model. And of course, as time goes by, the new data can be fed back into the analysis to filter out the more unlikely paths in phase space.

Financial forecasts are a different beast, of course, as we now depend on the human factor.

rjsterry

Stevo 666 wrote:

rjsterry wrote:

TheBigBean wrote:

There is a nice chaos quote on wikipedia.

Chaos: When the present determines the future, but the approximate present does not approximately determine the future.

I think it's very easy to throw your hands up and say "there are too many variables; it can't be done". It may be the case, but I would guess that neither of us have looked at the mechanics of the modelling NIESR used. At the very least a set of figures showing an order of magnitude difference between the impact of signing new trade deals versus leaving an existing one demands further investigation.

If these figures have a low confidence of accuracy, then claims of windfalls from new trade deals should equally be treated as not reliable - the contrary argument falls away as well. Those promoting an independent trade policy have no idea either.

Have a look at the number of variables in complex real world systems where chaos theory applies - e.g. weather systems, financial markets. The permutations are mind boggling. Also part of this is sensitivity to initial conditions, where a seemingly insignificant difference at the start is amplified over time. And in economic models we are also trying to predict future human behaviour.

Fascinating stuff, but my own experience of forecasting in a business environment bears this out quite nicely in the real world.

Sure. I understand the issue, and have done enough of my own projections to understand the gap between prediction and reality. Taking the weather analogy, while we might not be able to give an exact forecast for January 3rd 2021, we can give probabilities for various options. There are patterns and their are more and less likely outcomes.