[aguahombre]

 dwar0123, on 2012-October-15, 11:40, said:

Does it involve time travel?

Either that or some grave digging and smelling salts to revive Ronnie...who also saw no need to interrupt or outshout his opponents.

[WellSpyder]

 phil_20686, on 2012-October-15, 12:04, said:

If its NGDP/increase in government spending then a multiplier of less than one is impossible in the absence of imports and (monetary) inflation.

Two reasons why you might get a multiplier less than one even without inflation:

a) an increase in government spending means the government will have to borrow more money than it otherwise would have done. An increase in the demand for money will increase the interest rate needed to borrow, and this will reduce the amount of private sector demand in the economy where that demand is financed by borrowing, either for consumption or investment

b) an increase in government spending (assuming it is not financed by a simultaneous increase in taxation) will require an increase in taxation at some point in the future to pay for it. Sensible taxpayers will realise they face an increased tax bill in the future and will therefore spend less now so that they have more money available to meet the higher future bills (this phenomenon is known to economists as "Ricardian equivalence").

Both these factors mean that an increase in government spending will tend to "crowd out" a certain amount of private sector spending even when the economy is not operating at full capacity.

[phil_20686]

 WellSpyder, on 2012-October-16, 02:36, said:

Two reasons why you might get a multiplier less than one even without inflation:

a) an increase in government spending means the government will have to borrow more money than it otherwise would have done. An increase in the demand for money will increase the interest rate needed to borrow, and this will reduce the amount of private sector demand in the economy where that demand is financed by borrowing, either for consumption or investment

b) an increase in government spending (assuming it is not financed by a simultaneous increase in taxation) will require an increase in taxation at some point in the future to pay for it. Sensible taxpayers will realise they face an increased tax bill in the future and will therefore spend less now so that they have more money available to meet the higher future bills (this phenomenon is known to economists as "Ricardian equivalence").

Both these factors mean that an increase in government spending will tend to "crowd out" a certain amount of private sector spending even when the economy is not operating at full capacity.

So point (a) is not really important. `Obviously' the central bank is capable of steering the economy. If it is committed to doing so, your fiscal policy should never have any real impact on production. You would only think in terms of fiscal multipliers if you believed that the central bank was planning to hold interest rates constant, or would be unable to gain traction in order to move interest rates. (a) is a classic consideration if we lived in a Gold Standard.

On (b), Ricardian equivalence doesn't really exist. . For one thing, it only makes sense if you expect your consumption to be constant. If you expect your consumption to rise due to rising productivity between now and when the tax is applied, it makes no sense to reduce my consumption now. At the more behavioural approach, no one budgets their finance more than twelve or eighteen months in advance. At least I don't know anyone who does. The payback is normally at least that far in the future.

Obviously, Ricardian equivalence, in some sense, exists in the cumulative (=long run) multipliers. However this is almost a non statement. In the long run growth is almost wholly determined by technological progress and population growth. If there is innovation going on in the background, there just becomes a backlog of useful growth potential that will eventually be used up when the economy goes back to functioning strongly.

[phil_20686]

 hrothgar, on 2012-October-15, 13:05, said:

Classic Keynsian arguments about the government "priming the pump" during a liquidity trap operate on very different principles.

This isnt actually true. There are various ways to look at the issue, but "priming the pump" arguments are really arguments about the price of money. For example, part of the `priming the pump argument' is about the paradox of thrift, where even well off people with steady jobs spend less and save more during a recession. However, a central Keynesian thesis is that "prices clear markets", If I have money, and I choose not to spend it, it means that the prices of things I am thinking about buying are too high. Since money buys more during a recession, it is sensible to save, and spend when your wages buy less. This also partly explains the behaviour of equity markets: if equities fall en masse, it is just a way of saying that cash has become more valuable. Each dollar now buys more of a company than it did before, so `saving' into equities is more attractive relative to consumption compared to pre-recession. Keynes made a hand wavy argument about uncertainty in the future, but to me, the price of money argument is much more believable, although, to a certain extent they say the same things.

This is clearer if you think of the value of a dollar as 1/NGDP. Since increased spending (priming the pump) increases NGDP, then with production held constant it lowers the price of money. This lowering makes both goods and wages cheaper, and prices clear markets, so since they are cheaper more people buy them and production rises.

To a certain extent, these are only matters of framing, but the price of money framework is much more unifying and applicable. Keynes grew up under the Gold Standard, he thought about money as having a fixed value, and when central banks had a much less important role. True fiat money/central banking, necessitates a change in approach. Central banks now see themselves in the role of maximising production. They should set the price of money so that it facilitates the maximum amount of economic activity. Given those definitions anytime fiscal policy can increase output, it means by definition that the central bank has not set the price of money to maximise output.

[hrothgar]

 phil_20686, on 2012-October-16, 07:24, said:

This isnt actually true. There are various ways to look at the issue, but "priming the pump" arguments are really arguments about the price of money. For example, part of the `priming the pump argument' is about the paradox of thrift, where even well off people with steady jobs spend less and save more during a recession. However, a central Keynesian thesis is that "prices clear markets", If I have money, and I choose not to spend it, it means that the prices of things I am thinking about buying are too high.

I used the words liquidity trap for a reason...

[WellSpyder]

 phil_20686, on 2012-October-16, 06:56, said:

So point (a) is not really important. `Obviously' the central bank is capable of steering the economy. If it is committed to doing so, your fiscal policy should never have any real impact on production.

Now I don't really know what you are trying to argue at all! I provided a reason why the multiplier might be less than one, and you seem to be disagreeing, saying it will normally be zero! That sounds to me like you are agreeing not disagreeing!

Quote

On (b), Ricardian equivalence doesn't really exist. . For one thing, it only makes sense if you expect your consumption to be constant. If you expect your consumption to rise due to rising productivity between now and when the tax is applied, it makes no sense to reduce my consumption now.

That just sounds like nonsense to me, I'm afraid. I'm not talking necessarily about consumption falling compared with what it was last year, but simply falling compared with what it would have been if the government hadn't in some sense signalled higher future taxes. That argument doesn't seem to me to be affected at all by whether the underlying pattern of consumption is rising, falling, or staying the same.

Quote

At the more behavioural approach, no one budgets their finance more than twelve or eighteen months in advance. At least I don't know anyone who does. The payback is normally at least that far in the future.

OK, finally a point I can accept may have some validity.

[phil_20686]

 hrothgar, on 2012-October-16, 07:49, said:

I used the words liquidity trap for a reason...

At best a liquidity trap means that the transmission mechanism is ineffective. There seems to be fairly broad agreement now that if you seek to target the forecast, then you can break out of a liquidity trap. You say, I will do x amount of easing until inflation hits y, rather than saying "I will do x amount of easing and see what happens". The Fed has now come on board with that argument, promising to ease "until the situation improves".

At its most basic, monetary policy can always gain traction if fiscal policy can, since the bank can always just print up some money to buy all the same goods and services that the government would have bought. Practically, there are a wide variety of financial products that the Fed could buy. I see they moved into the MBS market, for example.

[dwar0123]

 WellSpyder, on 2012-October-16, 07:50, said:

Now I don't really know what you are trying to argue at all! I provided a reason why the multiplier might be less than one, and you seem to be disagreeing, saying it will normally be zero! That sounds to me like you are agreeing not disagreeing!

Multiplying something by zero results in zero. Multiplying something by 1 results in no change.

[WellSpyder]

 dwar0123, on 2012-October-16, 10:54, said:

Multiplying something by zero results in zero. Multiplying something by 1 results in no change.

As it happens, I'm already aware of those rules of arithmetic (just as I assume you are aware that zero is less than one). Do you think they affect this discussion?

[hrothgar]

 phil_20686, on 2012-October-16, 09:16, said:

At its most basic, monetary policy can always gain traction if fiscal policy can, since the bank can always just print up some money to buy all the same goods and services that the government would have bought.

In the US, at least, the Fed does not get to purchase any (real quantity of) goods and services.
An order for new BIC pens does not a fiscal stimulus program make.

If the Bank were to do so, this would be an example of fiscal policy because the policy goal it operates through government spending rather than changing the money supply.

You are confusing the type of policy with the way the policy is funded...

[phil_20686]

 WellSpyder, on 2012-October-16, 07:50, said:

Now I don't really know what you are trying to argue at all! I provided a reason why the multiplier might be less than one, and you seem to be disagreeing, saying it will normally be zero! That sounds to me like you are agreeing not disagreeing!

I think we agree conceptually, but are having a technical disagreement over the appropriate way to define a budget multiplier. I think I can illustrate our technical disagreement more clearly. Suppose we draw one of those AD vs Y graphs that economists are so keen on, and I make the economic response function vertical - i.e. I fix real output to be a constant. Then an increase in government spending does not increase real output, and the multiplier is zero. Suppose instead that I am some forward thinking empiricist who realises that real output cannot be measured, so instead I decide that I will measure GDP, by measuring NGDP and stripping out inflation. So the government sells a bond for $100, and buys $100 of real output, so now NGDP has increased by $200, once for the government's spending, and once for the buying of the bond (Since saving = investment). Now if the $100 used to buy the bond, would otherwise have been used to buy the goods and services that the government is buying, then the change leads to a change in NGDP of 100, but since there is no extra demand for goods and services, there is no price change/inflation, and the increase in NGDP is exactly taken up by the increase in the supply of "goods". So its clear that in this case the `no effect' multiplier is one. This, at least, was the approach that I had in mind.

 WellSpyder, on 2012-October-16, 07:50, said:

That just sounds like nonsense to me, I'm afraid. I'm not talking necessarily about consumption falling compared with what it was last year, but simply falling compared with what it would have been if the government hadn't in some sense signalled higher future taxes. That argument doesn't seem to me to be affected at all by whether the underlying pattern of consumption is rising, falling, or staying the same.

Ok, so presume that the multiplier during a recession is higher than at full employment, then the cost of paying off my extra debt is lower, in consumption terms, than the gain I get. Thus the effect of higher taxes later is necessarily lower in total in any case where fiscal policy increases total output, since its self evident that when the economy comes up against its full employment limit, no increase in aggregate demand can increase output, so the multiplier must be small. If you look at the graphs in http://wps.prenhall...._Ch22_topic.pdf but imagine that the economy's response function to increased aggregate demand (the black line) is like a tan function. The explanation given there for the balanced budget multiplier has a nice property, if we assume the slope of the black line (for marginal changes) to be `m', then the multiplier for a balanced budget is (1-MPC)/(m-MPC), and it follows that the (instantaneous) multiplier ranges from infinity to zero. (if the Marginal Propensity Consume is greater than the slope, then each cycle keeps generating enough extra demand to move you along the slope in a self perpetuating cycle, hence an infinite multiplier, in practice, if the slope is a tan, its impossible to go further down than when the slope = MPC, since the cycle would exist for any spending).

Anyway, with this assumption about the response function, its clear that Ricardian equivalence will not (fully) offset spending, since the declining budget multipliers insure that paying it off will not reduce consumption as much as the initial spending increased consumption, even if economic actors act to smooth their consumption in an economically rational manner, which does not seem that likely in the first place. . If you include interest rates, then you can do an option pricing model, and the result will be that you need a change in fiscal multipliers at least sufficient to pay back the interest. But thats just obvious, and at current rates, practically impossible to avoid.

[phil_20686]

 hrothgar, on 2012-October-16, 11:12, said:

In the US, at least, the Fed does not get to purchase any (real quantity of) goods and services.
An order for new BIC pens does not a fiscal stimulus program make.

If the Bank were to do so, this would be an example of fiscal policy because the policy goal it operates through government spending rather than changing the money supply.

You are confusing the type of policy with the way the policy is funded...

So the BoE can purchase basically whatever through its Asset Purchase Facility. It has been purchasing corporate debt recently I believe. All central banks prefer financial instruments as they think that they are less distorting. Its fiscal policy it it involves the government. If it is done by the monetary authority it is monetary policy. . The Fed had a policy of Credit easing, which was buying up private sector assets. I don't think its problematic for the Fed to print money and buy, say, mortgage securities, or equities, etc. Its more problematic buying actual stuff since what would the Fed do with it all + retail markets are just much too small. There has been some talk from people like Ryan Avent of the Fed just buying underwater home owners out of their mortgages and reselling the house at fair value. This obviously has moral hazard problems, but in essence it is just a form of unsterilised debt purchase. I cannot imagine the Fed would ever go for that, but no one seems to think that it would be illegal.

[WellSpyder]

 phil_20686, on 2012-October-16, 11:42, said:

So the government sells a bond for $100, and buys $100 of real output, so now NGDP has increased by $200, once for the government's spending, and once for the buying of the bond (Since saving = investment). Now if the $100 used to buy the bond, would otherwise have been used to buy the goods and services that the government is buying, then the change leads to a change in NGDP of 100, but since there is no extra demand for goods and services, there is no price change/inflation, and the increase in NGDP is exactly taken up by the increase in the supply of "goods". So its clear that in this case the `no effect' multiplier is one. This, at least, was the approach that I had in mind.

Maybe I haven't grasped what you mean by NGDP - I thought you just meant the nominal (cash) value of GDP. If so, I don't think the purchase of the bond adds to NGDP at all. So the net effect in your example is $100 extra spending by the government minus $100 less spending by the purchaser of the bond = zero.

[phil_20686]

WellSpyder said:

1350458766[/url]' post='674971']
Maybe I haven't grasped what you mean by NGDP - I thought you just meant the nominal (cash) value of GDP. If so, I don't think the purchase of the bond adds to NGDP at all. So the net effect in your example is $100 extra spending by the government minus $100 less spending by the purchaser of the bond = zero.

So I thought that Y = G +C + I plus other term we don't care about. Since saving = investment, the effect of moving 100 dollars from personal consumption to saving/investment has no effect on Y. on the other hand the increase in government spending increases Y. As a understand it buying a bond is usually counted as savings.

[mike777]

 phil_20686, on 2012-October-17, 02:06, said:

So I thought that Y = G +C + I plus other term we don't care about. Since saving = investment, the effect of moving 100 dollars from personal consumption to saving/investment has no effect on Y. on the other hand the increase in government spending increases Y. As a understand it buying a bond is usually counted as savings.

You miss the main point:

1) 99.99% of world has no idea what you are talking about



2) Why is increasing "Y" the most important economic point you can make?


3) so unlimited goverment spending= unlimted good "y"

[WellSpyder]

 phil_20686, on 2012-October-17, 02:06, said:

So I thought that Y = G +C + I plus other term we don't care about. Since saving = investment, the effect of moving 100 dollars from personal consumption to saving/investment has no effect on Y. on the other hand the increase in government spending increases Y. As a understand it buying a bond is usually counted as savings.

With apologies to mike777, I'd like to continue this debate just a moment longer. The underlying problem, I think, is that the fact that S will equal I does not mean that it is helpful to substitute one for the other in an expression like the GDP expenditure identity. S and I will only be equal as a result of adjustments elsewhere in the economy, including for example adjustments in C. It is also true that the purchase of gov't debt does not actually increase savings - it merely turns an amount of savings from cash into bonds.

[phil_20686]

 WellSpyder, on 2012-October-17, 03:42, said:

With apologies to mike777, I'd like to continue this debate just a moment longer. The underlying problem, I think, is that the fact that S will equal I does not mean that it is helpful to substitute one for the other in an expression like the GDP expenditure identity. S and I will only be equal as a result of adjustments elsewhere in the economy, including for example adjustments in C. It is also true that the purchase of gov't debt does not actually increase savings - it merely turns an amount of savings from cash into bonds.

So I specifically mandated that the money to purchase bonds came from a reduction in C in my example, with a consequent increase in the savings rate. Obviously that is not very realistic, but I wanted to illustrate why I thought the default NGDP multiplier could be one even if real output is fixed by definition. If you assume that the bond was purchased by savings that were already existing in cash form, then we would have G=>+100, C unchanged, and S unchanged. Now the NGDP multiplier would be one in this case, despite no change in real output. Now its obvious in this case that this is just inflation - the extra competition for goods and services driven by the increase in G+C will simply raise the price level so that output is maintained.

If moving $100 from consumption to saving removes it from the calculation of Y, then I do not see how the expenditure and the income pictures can add up: Suppose that a worker has whose income has not changed, decides to save more, and if we assume that he `saves' by hiding cash under his mattress, its extremely hard to see how GNI=GDP can be true. I will try to read something on the use of S=I in analysis, accounting identities are always tricky in maths.

Ok so I read a little more this morning: S=I should be treated as an equilibrium condition. It can be violated temporarily. That is not a problem for this model as we can assume zero friction = instantaneous adjustment. A more serious issue is whether bonds should be counted as Savings/Investment. Basically, from a theoretical standpoint, apparently you should treat bonds as Savings only if they are spend on Investments, not if they are spent on consumption. In my example, they are spend on consumption, so you should not consider there to be a change in S/I. If the government bonds are instead spend on S/I we should consider that C falls, I rises and Y is unchanged.

It still leaves an empirical question of what you actually count. If you treat bonds as an investment and they are spend on consumption, you double count, if you treat them as consumption, but the government spends them on investment you under count. Either way affects the measured multiplier.

[Flem72]

 mike777, on 2012-October-17, 02:17, said:

You miss the main point:
1) 99.99% of world has no idea what you are talking about

+500

Expertise in economic 'science' is a wonderful thing. It provides so much to talk about, such a specialized, impressive, highly nuanced descriptive language, and so little true understanding of what makes things work. Always has seemed to me that, if it were really a science, rather than black magic, we wouldn't ever be in economic trouble. We'd just tweak a nut here and a bolt there and solve the problem. Or is it just me?