Duncan’s Economic Blog

Savings, Investment & IS-LM: An Answer for Tim

Posted in Uncategorized by duncanseconomicblog on July 7, 2009

As I noted last week Paul has a good post up on the need for a new  ‘common sense’ economic agenda.  The piece was cross posted over at Liberal Conspiracy.

 The infuriatingly readable Tim Worstall commented upon it.

 Where do you think investment comes from?

The traditional (neo-liberal if you like) answer is that investment comes from savings. So if you reduce savings by taking money that would have been saved (if held by the rich) and make sure that is is spent on current consumption (by giving it to the poor who do not save) then give that there will be fewer savings then there will inevitably be less investment.

Just so you don’t reject this as being only the neo-liberal answer, I think you’ll find that the investment / savings link is the IS in the IS/LM model which is pretty much the basic elucidation of Keynesian thought.

I fear that you’re a little confused in your economics here.

Paul asked yesterday for my thoughts on this.

Interesting that Tim talks about Paul’s (and implicitly my own) argument running against the ‘basic elucidation of Keynesian thought’. It might well do.  But then it’s not a Keynesian argument, it’s a Post Keynesian argument. I.e. it flows from an economic school that believes it follows the true path of Keynes’ ‘General Theory’ rather than from the school which grew up in his name and dominated macroeconomics for decades.

 I take the Post Keynesian line.

 Thus, Post Keynesians and neoclassicists can pretty much agree on definitions of saving and investment, for example. They can also agree that in a simple economy with no government and no foreign trade, savings must equal investment. This is, in fact, a basic national income accounting identity. Where these two schools differ is not here, but in how they see the causal nexus between the two. For neoclassicists, savings cause investment; it provides the funds needed to build new capital equipment. In contrast, Post Keynesians hold that investment determines savings. Investment can be financed by borrowing from banks, which does not require savings because banks create money by lending. Investment, in turn, generates jobs and incomes. Some of this income will be spent, and the rest saved; thus, at the end of the process, savings come to equal investment.

 Empirical Post Keynesian Economics, Holt & Pressman (Emphasis mine).

 

It’s obviously true that ‘savings = investment’ but that is a simple accounting definition. What matters is the causation and here myself and Tim disagree. Interestingly Paul intuitively ‘gets’ the Post Keynesian view in his own line that:

I’ve also written about how the now dominant narrative of neoliberalism and money supply control became so dominant; despite the fact that the fundamental assumption of a finite world money supply is flawed, the ‘good housekeeping’ / ‘you cannot spend what you have not got’ narrative has continued to hold sway over public opinion for a generation and more.

 To quote Holt & Pressman again:

 The neoclassical view sees savings as the answer to the problem of how we get more goods for tomorrow. Savings generates investment, which in turn, increases our productive capabilities. The paradox of thrift is forgotten or ignored on this traditional way of thinking. Post Keynesians see traditional analysis as putting the cart before the horse; they regard investments as the cause of savings,.

 A New Guide to Post Keynesian Economics, Holt & Pressman

 The Paradox of Thrift  is crucial. If savings are very high then spending will be lower and so demand is also smaller , in which case why would anyone invest?

On Tim’s more general point on the IS-LM model… don’t get me started. Remember the IS-LM model is not an invention of Keynes, it’s an attempt by others to formalise his work. I would mention that even John Hicks, its creator, later renounced it.

It has many problems – not least of which being that it is a model of a static economy, it allows no role for path dependency and that it places zero weight on expectations, which are crucial in the General Theory. The Journal of Post Keynesian Economics (not available for free online) is an excellent source for those interested in these issues.

About these ads

22 Responses

Subscribe to comments with RSS.

  1. Tim Worstall said, on July 7, 2009 at 8:25 am

    I see your point and accept that it’s held to be true by Post-Keynesians.

    I also don’t think it to be true. There’s an awful lot of weight being put there on the idea that banks can effortlessly create credit. Which they cannot.

    The banking system as a whole creates credit, yes, and there’s a limitation to it, the required reserve ratio. (That’s now under Basel I and II and not called that any more but functions similarly.) But an individual bank does not, it is still lending out its deposits (no, really, that’s what the scramble into the overnight market is, banks covering what they’ve lent during the day. Their books must indeed balance at 4.30 pm each day.)

    It also entirely ignores that part of investment which is not bank credit. And, as you will have noted, no bank lends 100% of the costs of a project or a firm. There always has to be a capital cushion.

    So, even if banks could and did in fact create credit as fast and as often as they wanted, we would still need real savings, proper cash somewhere (ie, saved potential consumption from an earlier period), to finance that required capital before the multiplier of the credit comes into effect. We still need those VC’s, personal savings, private equity etc etc. Which means that a reduction in those real savings will indeed lead to a reduction in investment.

    Just as an example, imagine the reaction to walking into a bank with a business plan. You’ve no collateral (ie, real savings from elsewhere) and no cash, no capital. You want the bank to fund the company 100% from that credit they create so freely.

    Do you get funded? No, you clearly and obviously don’t. Not a celluloid rat in Hell’s chance of getting funded.

    Thus savings are not (always!) the result of investment. They are a requirement, a necessity, before investment takes place.

    I would put this Post-Keynesian idea that savings are a product of investment up there with an awful lot of other macro ideas: fails, because they ignore micro.

  2. duncanseconomicblog said, on July 7, 2009 at 8:36 am

    Tim,

    I suspect that we are going to disagree on this issue. Although I also suspect we have common ground in both disliking the IS-LM model. Admittedly for very different reasons.

    First off, and taking your points in reverse order, – I do not like the trend to give solid micro foundations to macroeconomics. Mainly as I find much of micro economics to be simply wrong but also due to problems of aggregation. The paradox of thrift is surely an example of this – what is rational at a micro level can be irrational at a macro level. Prisoners’ dilemma type problem.

    Yes banks are constrained by reserve requirements and I am not arguing they can create as much credit as they want. But, two weeks the ECB lent 442bn Euros to European banks for one year at 1% in an ‘effort to boost bank lending’. Where are the savings coming from here?

    Equally when a company issues bonds – can it not be said to be creating ‘savings’, or at least an instrument to be used in savings?

    Of course internal finance is important, and there is a long stream of Post Keynesian thought on corporate cashflow. But in the end debt finance is still important to business in the real world.

    Surely our argument can be empirically tested.

    In your analysis any action that increases savings should lead to an increase in investment, and hence more capital formation and higher long term growth – correct? So a policy based on encouraging people to save more would be good for growth? I disagree.

  3. Tim Worstall said, on July 7, 2009 at 8:49 am

    “In your analysis any action that increases savings should lead to an increase in investment, and hence more capital formation and higher long term growth – correct?”

    I tend not to do macro for exactly these reasons. Not really my subject.

    However, I’m perfectly willing to agree that when growth is below trend, when there are unused assets, that an increase in savings “could” (not will) lead to a reduction in consumption and thus slower growth than might otherwise occur.

    However, I’d also want to insist that over the long term it’s productivity that matters (As Krugman has pointed out, productivity isn’t everything but in the long run it’s just about everything). Productivity is raised by either applying more capital to labour or by changing the technology we use. That latter simply is not funded by bank credit creation. That just isn’t how innovations or innovative companies get funded. Small companies is where I’ve spent most of my working life and you just don’t get to start up on the back of a bank.

    As I say, the Post-K thing may look good on the macro level but by ignoring how in the real world technological advances are really funded it fails because it’s ignoring micro.

  4. duncanseconomicblog said, on July 7, 2009 at 9:28 am

    Interesting debate we’re having.

    Productivity is indeed important.Howvere

    “Productivity is raised by either applying more capital to labour or by changing the technology we use. That latter simply is not funded by bank credit creation. That just isn’t how innovations or innovative companies get funded.”.

    I agree to an extent. I’m not arguing that bank credit alone is evrything. Post Keynesianism at least recognises the role of the entrepreneur, and the cruicial role of expectations and uncertainity. I’ve always found it strange that the entrepreneur is absent from so much micro and macro theory.

    I think you are actually attacking ‘traditional’ Keynesianism here, not Post Keynesianism. A traditional PK attack opn the IS-LM is that it ignores technological change. Going back to the static versus dynamic model point.

  5. Tim Worstall said, on July 7, 2009 at 10:46 am

    “I’ve always found it strange that the entrepreneur is absent from so much micro and macro theory.”

    Well, given that I’ve spent most of my working life a an entrepreneur (not all that successful a one it has to be admitted) that tend to be the lens through which I view much economics.

    And to make public what I mentioned in email, I’m not an economist and don’t pretend to be one. An interested amateur no more, so I most certainly don’t claim to have detailed knowledge of the whole shebang.

  6. Paul said, on July 7, 2009 at 12:14 pm

    Sorry I’m late.

    Yes, interesting discussion. Like Tim, I come to it as an entrepreneur (though I bet I’m worse at it than him), and I had no real understanding of anything even vaguely economics-like until I started reading books a year or two back (Tim may argue I still have no understanding, which is why I passed the debate from my ‘intuition’ over to Duncan).

    Surely, though, the micro/real world lies somewhere in the middle of all this. My original argument was that more equal incomes would lead to greater spending, but i didn’t think that automatically meant that there would be NO investment, which I accept is often at small enterprise level a question of personal equity (often hugely backed by ‘sweat equity’).

    And if the investment that continues to come, either through savings or through bank loans (or both) is somewhat more constrained/variegated because of that greater equality of income, then that’s not a bad thing because the investment willl be focused in more on the provision to the market of goods/services for which there is a sustainable demand (not based on rising lower income debt). Everything I’m now reading with a lay eye now is leading me to believe that sustained demand comes from a levelling out of capital and labour rewards, and that productivity is raised as much by the expectation of future demand (lessened risk of investment) as it is by increased availability of capital but higher risk demands. I run a childcare business (as a social enterprise) and I’m more interested in knowing that families wil be able to afford to use the serivce in three years and that I won’t be plunged into cashflow crisis than I am in knowing that I have the capital to invest in a somewhat risky expansion.

    In the end it’s sustainabilty of growth, in both senses, that’s important, and not speed, although speed is important enough in the initial post-recession surge.

  7. duncanseconomicblog said, on July 7, 2009 at 12:42 pm

    Paul,

    I think Tim, over at Lib Con, was arguing that in your (and my) prefered economic model of greater income equlaity there would be less investment.

    His argument was based on the idea that as there would be less saving there would necessarily be less investment.

    But I thinik that allow saving does indeed equal investment, the relationship is not that straight forward. Investment doesn’t rise simply becuase savings do. The world isn’t static.

    Investment happens because entrepneurs in small business, or managers at large business, decide to invest based on a view of the future. It’s not a mechanical response to more saving.

    I think in your, and my, model – a world of greater spending they are more likely to make this decision. Equally banks are more willing to lend as corporate cashflows are likely to be healthier and so loans are more likely to be repaid.

    So I totally agree with your view that:

    “Everything I’m now reading with a lay eye now is leading me to believe that sustained demand comes from a levelling out of capital and labour rewards, and that productivity is raised as much by the expectation of future demand (lessened risk of investment) as it is by increased availability of capital but higher risk demands.”

  8. Tim Worstall said, on July 7, 2009 at 12:56 pm

    “Investment happens because entrepneurs in small business, or managers at large business, decide to invest based on a view of the future. It’s not a mechanical response to more saving.”

    True to an extent. If more entrepreneurs want to try something then demand for capital will increase, calling savings into being as returns to savers rise.

    However, a rise in savings independent of entrepreneurs’ desires should lower the cost of such capital, encouraging more to start up (we’re not all that surprised when a fall in the price of something increases demand for it, are we?). Don’t forget, the real price of capital in a start up is not interest, it’s how much of the idea you have to sell to get the cash.

    But if we specifically reduce the amount of savings, by taking from those who would save and redirect to consumption, we’ve just raised the price of that capital….reducing those who start up and thus innovation.

    There’s another rather more cheeky argument as well. Most inventions (ie, the leaps, not innovation, the step by step improvements) start out as toys for the rich. Personal Computers, cars, mobile phones etc. In a society with a flat income structure who will buy such toys and thus subsidise their becoming mass market products?

    Take mobile phones: without rich poseurs in the 80s buying bricks that rarely worked how would Indian fishermen (I’m sure you know the Kerala story: reduced prices in the markets, higher profit margins for the fishermen and less catch rotting, very definitely a win all round, an increase in both wealth and productivity from what was a rich man’s toy) have them only 30 years later?

    This is of course another reason for my dislike of certain macro arguments: innovation, invention, technology and productivity are so often simply outside the models. But they are what drive things forward over anything more than the several year time frame.

  9. duncanseconomicblog said, on July 7, 2009 at 1:02 pm

    Tim,

    3 quick (ish) points.

    (i) “However, a rise in savings independent of entrepreneurs’ desires should lower the cost of such capital, encouraging more to start up (we’re not all that surprised when a fall in the price of something increases demand for it, are we?).”

    I think though, you’re still thinking in static not dynamic terms. As spending increases and investment occurs the stock of savings will rise. A lower marginal propensity to save but higher income.

    (ii) “There’s another rather more cheeky argument as well. Most inventions (ie, the leaps, not innovation, the step by step improvements) start out as toys for the rich. Personal Computers, cars, mobile phones etc. In a society with a flat income structure who will buy such toys and thus subsidise their becoming mass market products?”

    This is an interesting argument. One I’ll ponder.

    (iii) “This is of course another reason for my dislike of certain macro arguments: innovation, invention, technology and productivity are so often simply outside the models. But they are what drive things forward over anything more than the several year time frame.”

    Agreed. To simply talk of ‘technology shocks’ is to miss a lot of the complexity of the economy over time.

  10. newmania said, on July 7, 2009 at 1:26 pm

    Agreed. To simply talk of ‘technology shocks’ is to miss a lot of the complexity of the economy over time.

    I have a nasally intuitive feeling that the disincentive to start New small businesses over the last tem years most of which is regulatory is going to be a very long term drag because they are the source of innovation . Hard to model that as the demand for the products does not yet exist

  11. [...] that experience, and even more so by seeing my principal (and intelligent) critic Tim Worstall now engage in civilised and even half-though-reluctant agreement with My Post-Keynesian Economics Guru Duncan [...]

  12. chris said, on July 7, 2009 at 4:23 pm

    Tim – I think we can all accept – coz it’s true – that, for an individual company or entrepreneur, savings are a precondition for investment. Traditionally, internal funds (retained profits) have been a key determinant of investment, and wealth (redundancy pay or inheritance) has been necessary for entrepreneurs to get started.
    But this fact is awkward for basic neoclassical economics, which assumes free capital markets.
    And it’s wholly different from the argument that higher savings by government or households will stimulate investment.
    On this, I commend the late David M Gordon’s piece, “Putting the horse (back) before the cart” in Epstein and Gintis’s Macroeconomic Policy after the Conservative Era:
    http://books.google.co.uk/books?hl=en&lr=&id=6qTcuX3ufk8C&oi=fnd&pg=PA57&ots=sx2t1QZN3b&sig=TcxQafOY6TwfEiVFfSPa5Q1fMqA

  13. rommeldak said, on July 7, 2009 at 4:39 pm

    Tim, for someone who claims not to be an economist, you certainly have a solid grasp of many basic concepts! You have argued your positions well and have focused on the salient issues.

    My position on Investment-Saving is this. First off, I would agree with Duncan that the equality between the two is a mere accounting identity. These two values (in a closed economy with no government–gets a bit more complex if you include them, but still same basic story) must, ex post, be the same. Planned saving and planned investment need not be equal, however, and therein lie the interesting stories.

    Take, since it fits in with the discussion above, a situation in which planned investment is greater than planned saving. As you all have observed, what happens next depends on your view of the financial system. In the classical loanable-funds version, where savings is an absolute limit on the credit banks can extend, the planned investment cannot take place as it cannot be financed. Well, it goes up some, but only because savings rises, too. What happens is that the excess of investment-funding demand over savings causes a rise in the interest rate. That leads to both a decline in planned investment and a rise in saving. They will come to rest at a new, higher level of investment and saving. In this world, savers had to be goaded into saving a bit more or the investment could not take place and the rate of interest plays the role of setting S=I.

    Keynes (whose work is not represented by the “Keynesians” or the IS-LM model) disagreed, and Post Keynesians have extended his view thusly. Banks have a great deal of latitude, despite the reserve requirement, in terms of their ability to create credit. Banks are almost never “loaned up,” and the entrepreneurs in the banking sector are very creative in terms of coming up with new instruments that others will accept as collateral or payment. Is this ability endless? Some Post Keynesians think so (they are called “horizontalists,” for their view that the money-supply curve is horizontal), but I don’t agree. Surely it must become more and more expensive to squeeze another ounce of credit from the system. But, apparently, it’s a very long time before it becomes prohibitively expensive, such that in general banks do not have a terribly difficult time creating new credit. And one of Keynes’ major points is that savings is not the limit, liquidity preference is. Without going into great detail, we prefer to remain liquid when we are concerned about the future. Funds held in fairly liquid accounts are more difficult for banks to loan out (take a–and I’m not sure of the British equivalents here–checking account versus a savings account). So they will have to move to more expensive options. But they are not out of options.

    In terms of savings being necessary, that is absolutely true, but only in a relatively trivial sense. Banks must have some assets, to be sure, but only enough to cover the eventuality that they are forced to make good on the many promises they have made. While there are, as Tim points out, reserve requirements, there is still a lot of wiggle room–enough to make it absolutely possible for investment to rise without a prior increase in saving. In fact, that’s the typical pattern for, as Duncan points out, who wants to invest when consumers are saving more? I asked this of a Neoclassical colleague one day, incidentally, and he was flabbergasted. He’d never thought about it before! That’s something else that’s going on below the surface here, by the way. The traditional model used by my Neoclassical colleague is really set up to explain a different world, one in which full employment is either assumed to exist continuously or at least as the only equilibrium position. The salient issues in such a world are very different from Keynes’, where less-than-full employment is not only seen as possible, but likely. Why, indeed, bother about the effect on demand of rising savings if full employment is guaranteed? But, what if it is not. Now this requires consideration of some very different questions.

    To summarize, the traditional argument assumes the following sequence:

    rising S => falling interest rates => rising I (until S =I)

    Post Keynesians argue:

    rising S => falling demand => falling I => falling incomes => falling S (until S=I)

    I could go on and on (hazard of my profession!), but I’ll stop here. Very good discussion on both sides!

    John

    P.S. Duncan, you know there’s about to be a meeting of Post Keynesians (mostly) at Kingston U starting Friday? That’s why I am flying to London in a few hours.

    • duncanseconomicblog said, on July 7, 2009 at 4:45 pm

      Thanks for that.

      A very clear exposition.

      I only became aware of the conference this week and sadly will be at work.

  14. Tim Worstall said, on July 7, 2009 at 4:56 pm

    OK, I’m understanding the Post K view: and given my both small company and micro orientation, I’m still not buying it.

    “Banks have a great deal of latitude, despite the reserve requirement, in terms of their ability to create credit. ”

    I’ll even buy that. But this still doesn’t apply to that part of investment which is not funded by banks. There’s a whole sector of the economy which simply does not get bank loans, does not get funded by those who create credit nor have any ability to do so.

    That’s the small company and the inventive sector. Banks simply do not fund them. They must turn to what we’re all calling here the neo-classical version of savings. Real previously saved consumption.

    And as, in the longer term, I consider that the most important part of the economy in determining future wealth, even if the Post K view is true I think it an interesting oddity rather than an essential feature of the economy.

    (To consider how extreme my vews are on these sorts of subjects, I am very much of the view that large companies like regulation for it bars smaller competitors…..and corporatism, big business, big labour and big government in collusion I consider an abhorrence.)

    Chris, “Traditionally, internal funds (retained profits) have been a key determinant of investment, and wealth (redundancy pay or inheritance) has been necessary for entrepreneurs to get started.
    But this fact is awkward for basic neoclassical economics, which assumes free capital markets.”

    Or VC’s, private equity, angels, public subscription to share issues (I would love to see the provincial stock exchanges open again, with a sub-AIM, sub even OFEX, listing requirement)….at which point we might say that capital markets are a great deal freer now than they used to be.

  15. rommeldak said, on July 7, 2009 at 5:20 pm

    “But this still doesn’t apply to that part of investment which is not funded by banks.”

    As a matter of fact, Post Keynesians are among the few schools of thought that openly discuss this (i.e, that firms will finance investment from retained earnings–Duncan mentioned this above). However, and this still does not invalidate the other points made.

    The initial problem here is one of definitions. In the typical discussion of these issues (Post Keynesian or otherwise), the following assumptions are made:

    1. only households earn income and save;

    2. only firms borrow and invest.

    This is merely for convenience and simplicity. Everyone knows that firms save and households borrow. But, I strongly suspect that those of us with a formal background in this have been meaning “household saving” when we have said “saving.” When firms save, we would call that “retained earnings.” Hence, we may be arguing at cross purposes. From here on, I will use “household savings” and “retained earnings” rather than “saving” (and you should substitute “household savings” for “savings” everywhere in my previous post).

    I would argue that there is a critical difference between household saving and retained earnings in terms of their effect on the macroeconomy. The former represents a decision not to spend, while the latter is actually a means of financing spending. A rise in household spending is neither necessary to finance investment, nor it is conducive to investment. Meanwhile, retained earnings are most likely to increase when household saving declines.

    I really need to keep packing so I’ll cut it short here. Just one more quick thing in reference to an earlier discussion regarding micro foundations of macro analysis. I don’t agree that it’s necessary, but they do exist. The very first book I ever read out of graduate school was Alfred Eichner’s Megacorp and Oligopoly, which is a PK alternative to Neoclassical theory of the firm. It’s a bit dated now, but I still reference it on occasion AND the idea that firms actively seek to generate retained earnings for investment is a major premise of the work. I didn’t check Amazon UK, but here it is for $12.27 (or check your local library–but don’t hold your breath!):

    http://tinyurl.com/mc5lzj

  16. Mr. Mxyzptlk said, on July 7, 2009 at 6:49 pm

    From Duncan’s anorak squad

    “The very first book I ever read out of graduate school was Alfred Eichner’s Megacorp and Oligopoly, which is a PK alternative to Neoclassical theory of the firm.”

    Wow! what an interesting life you have led

  17. rommeldak said, on July 7, 2009 at 6:59 pm

    “Wow! what an interesting life you have led”

    INDEED! (he says, refusing to recognise the sarcasm that points to the sad, lonely existence Rommel has led…sad, lonely, desperate, sometimes funny but always and ultimately meaningless…*sniff*)

    • Mr. Mxyzptlk said, on July 7, 2009 at 8:02 pm

      I was only joking and anyway ole dunc has two pints and falls over how sad is that

      • rommeldak said, on July 7, 2009 at 8:07 pm

        (I assumed you were joking, hence my continuation of the joke!)

  18. rommeldak said, on July 7, 2009 at 7:45 pm

    Speaking of Eichner, here is the Amazon UK link (killing time before I head off to the airport!):

    http://tinyurl.com/nsjf57

  19. duncanseconomicblog said, on July 8, 2009 at 8:18 am

    On the micro foundations point:

    I often think that micro models lack a good macro foundation – firms and households don’t operate in isolation.

    Indeed I am prone to get annoyed at the whole macro/micro split.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 63 other followers

%d bloggers like this: