The Debate over Investment and Saving

A number of earlier posts discuss issues relating either to investment/saving or interest rates or both. Many other posts relate to fiscal or monetary policy within a flexible exchange-rate fiat-money system, and these posts also often end up involving discussions of investment, saving and interest rates. It may be worth spelling out the significance of this debate.

A basic point made in the previous post is that ex ante (or planned) saving is not required prior to ex ante investment occurring, and that logically investment precedes saving. The argument is about causation. Investment is an independent variable, saving is not. More generally, injections are independent variables and leakages are not.

A spontaneous attempt to increase saving in aggregate is self-defeating. It will temporarily create unintended investment in the form of an unanticipated buildup in inventories, but this will set in motion negative income adjustments that eliminate the additional saving. The only sustainable saving is that underpinned by intended investment (in a two-sector model) or intended injections (in an economy with government and/or external sectors). To the extent investment plans go unrealized, income adjustments eradicate the excess saving or generate the required saving.

For neoclassicals, in the long run interest rates are meant to move in such a way as to adjust investment plans to the level consistent with full-employment saving.

In Marx or Keynes influenced approaches, in contrast, there is an openness in the possible outcomes that can occur which is absent in the neoclassical framework. The openness comes from the independence of the planned injections, which seems to provide a better depiction of capitalism.

In terms of investment/saving, it is capitalists who initiate the process through their investment decisions. Once they decide how much to invest, this sets production into motion. Irrespective of saving and consumption propensities, the exogenous investment decision will result in enough additional income to generate the necessary saving.

This view of causation runs across the different injections and leakages, and also applies to government and its fiat money.

There is an openness, in this view, not only in the investment decisions of capitalists but in the spending decisions of government. The price mechanism cannot be relied upon to push long-run output toward the full-employment level or entice investment sufficient to sustain full-employment saving. The level of output is open to a social determination that depends on autonomous decisions of the government, capitalists, to a lesser extent consumers (insofar as consumption is unrelated to current income) and foreigners. Once these actors have made their spending decisions, leakages occur in the required amount, irrespective of the propensities to save, tax and import.

Clearly, a central controversy relates to the rate of interest. What determines interest rates? Are they ultimately market determined or a political choice? What role do interest rates play? Are they an index of scarcity and intertemporal allocator or merely a distributive claim on surplus income? Can they be relied upon to induce investment sufficient to generate full employment? Can they even be relied upon to affect investment in a predictable way? The heterodox view is that interest rates have a political determination and are merely a distributive claim. They cannot be relied upon to induce investment of an amount sufficient to sustain full employment.

The level of output and employment, the level of investment and the rate of growth are all open to a social determination that is not dictated by the price mechanism.

To suggest otherwise is to imply that markets are in some sense natural phenomena that dictate social outcomes. To the contrary, markets are a social construction. Leaving the outcome to the market is one form of social determination. But this social choice is not inevitable. Nor can a choice to rely on market mechanisms be shown to generate the deterministic outcome (supply-determined full-employment output) often claimed in the long run. Such a deterministic outcome would rely on interest rates playing a role that they have been shown, in the capital debates, incapable of playing.


8 thoughts on “The Debate over Investment and Saving

  1. “this will set in motion income adjustments that eliminate the additional intended saving”

    It strikes me in the mathematics presented that ‘saving’ is a really bad word. In real terms ‘saving’ is just the arithmetic opposite of ‘investment’ with ‘investment’ being an arbitrary classification of durability of goods/services. So it’s hardly surprising that if ‘investments’ are eliminated then ‘savings’ vanish as well. It just follows from the equations.

    In nominal terms ‘saving’ appears to be just ‘investment’ plus ‘net financial assets’ – again with ‘investment’ being an arbitrary classification and ‘net financial assets’ being a leakage in the nominal capital distribution system.

  2. Thanks for your comments.

    What is your view on the causation? I’m expecting you agree that it is income that adjusts leakages to planned injections rather than interest rates that equilibrate investment and saving in the long run? Neoclassical causation implies an automatic tendency to full employment, irrespective of government fiscal actions (and irrespective of the level of NFA), provided real wages and real interest rates are sufficiently flexible.

    I take your point about the fuzziness, in practice, of classifying investment and consumption items, although I’m not sure I’ve understood the implication you have in mind.

  3. I do think that the nirvana promised by all economists has precious little evidence behind it and there is too much ‘supply-side’ vs ‘demand-side’ rather than “whichever side needs addressing today ‘cos it changes”.

    When I’ve asked for Real World evidence I’ve seen little of it, or worse it is ‘curve fitted’ – much as the ‘Mediterranean diet’ conveniently ignores the French.

    There clearly isn’t a tendency to full employment and there is a lower limit to real incomes in a civilised society unless you subscribe to the Sweeney Todd method of dealing with the ‘excess livestock’.

    On the flip side there doesn’t appear to be a cast iron guarantee that floating exchange rates automatically relieve external sector pressure, or that businesses automatically quantity adjust rather than raising prices.

    What I do know is that the price of money is not the problem at the moment. Of all the propositions that have crossed my desk in recent months, the main issue has always been the reliability of the sales stream. There was more money to be made giving the cash back to the bank than by investing in mediocre propositions.

    Normally after a recession there is a lot of investing space because so much has been destroyed. So the gap between demand and supply is sufficient to make it worth putting your money on the table as to profit margin covers the risk. That helps get up a head of steam to drive through the pinch point before new demand comes on stream from the new investment.

    This time the demand has collapsed, but the supply has not. So there is simply insufficient profit to make it worth the candle. Investing at the moment is like sifting coal from a slag heap.

    So if you implement policies to hold supply up in a slump, then you have to similarly boost demand to prime the investment pump again. So I see causality flowing from demand that is sufficiently in excess of current supply to get the investment juices flowing.

    “I’m not sure I’ve understood the implication you have in mind.”

    I started to describe that and it went on a bit. I need to do a post on that point.

  4. A basic point I am trying to make in a number of previous posts is that prior saving is not required for investment to occur, and that logically investment precedes saving. The argument is about causation.

    But prior saving is not the issue. Saving in the current period is required if investment is to occur in the current period, because saving is not consuming output and investment is the only other use for it. If you want to raise the amount of output you can produce in the future, you need to invest output now, and you can only use that output for investment now if you do not consume it.

    Since there are only two uses for output, it is obvious that at the margin, increased consumption in the future requires decreased consumption in the present. This does not imply that saving somehow causes investment, or investment saving.

    “Logically investment precedes saving”—I beg to differ. Logically, one implies the other. It’s not a question of chronology.

  5. Economic processes occur through historical time. Unless analysis is restricted to comparisons of full-employment equilibrium situations, historical time matters. Disequilibrium behavior affects the path followed and so the equilibrium – if any – that the economy approaches.

    I completely disagree with the use of marginalism – especially in macro – but for the sake of argument, when there is unemployment, then “at the margin” both investment and consumption can be increased. It is not a tradeoff, below full employment.

    Unless there is an automatic tendency to full employment (which neoclassicals have tended to suppose), there can be an equilibrium (realized plans) with both low investment and low consumption in the current period. Such situations involve both foregone consumption now and foregone consumption later. It is possible for both current and future consumption to be increased in such scenarios.

    So I still think it is correct to say that the debate comes down to whether there is an automatic tendency to full employment (or even whether full employment is the norm, since it is the only circumstance in which there is a macro tradeoff between investment and consumption), and this in turn comes down to controversies over the nature and role of interest. For investment and consumption to be a tradeoff, there needs to be full employment, and there is no guarantee of this unless the price mechanism (or some other mechanism) can be relied upon to induce demand sufficient to sustain full-employment output.

  6. “It is not a tradeoff”

    At the margin, it must be a trade-off, since there are only two uses for output and they are mutually exclusive.

    In the neoclassical view, the problem is not whether saving causes investment or investment causes saving, since these are the result of decisions taken by individuals at the micro level, but by what mechanism the system equilibriates the desires of savers with investors.

    The idea that we can have a higher level of GDP and so higher C and I in absolute terms if there is spare capacity is not particularly controversial. I get the BoE’s inflation report every quarter and they have a whole section devoted to the degree of capacity utilisation. If you are familiar with the stylized facts of real business cycles then you will know that capacity utilisation is a coincident indicator of the cycle, and widely used as such.

  7. In the neoclassical view, the problem is not whether saving causes investment or investment causes saving, since these are the result of decisions taken by individuals at the micro level, but by what mechanism the system equilibriates the desires of savers with investors.

    The bolded passage is why I wrote that I think controversies over interest are central. I agree with you that in neoclassical theory – which is typically full-employment equilibrium analysis involving only logical (not historical) time – the view is of simultaneous equilibration (and involves interest rates). But in most Marx and Keynes influenced perspectives, historical time matters and interest rates do not do the equilibrating. Further, in these perspectives, the economy is seen as inescapably a monetary production economy. The equilibration of saving and investment occurs through income changes, viewed in monetary (price-deflated) terms. Investment, from these perspectives, helps determine income, which adjusts planned saving to planned investment.

Comments are closed.