Prague Conference On Political Economy, April 2019

Third Annual Madrid Conference on Austrian Economics, November  2019

“The Austrian School of Economics in the 21st Century”, Vienna, November 2019



  1. Introduction

The standard Austrian economics view about what constitutes saving can be found in many scholarly texts. It begins with the statement that saving is the postponement/curtailing of consumption. The latter statement is beyond reproach but its continuation, i.e. the explanation how saving is transferred into investment is questionable. In particular we will argue that the standard Austrian view on the matter is very narrow, i.e. that it applies only to the very specific case of a barter economy that does not posses capital, i.e. to a basic, Robinson Crusoe type of economy. In a contemporary economy the processes of saving and investment do not require the prior or continuous production of consumer goods.

Once we have shown how saving and investment function from this perspective we will deal with the problem how savings are accumulated and used in a contemporary economy, i.e. an economy that possesses money and physical capital. We will show that every economy has an “investment fund” of capital which is used for capital substitution and investment. Based on this view we will also show how this fund can be used to compensate the capital depreciation, thus avoiding the necessity for net saving on the part of the economic agents. In other words, we intend to show that net saving is not a prerequisite for economic growth.


  1. Saving in a monetary, capitalistic economy

In order to reveal the real problem we will make use of the standard example used to introduce saving, namely, the case of Robinson picking-up berries to feed himself.  Robinson lives on a deserted island and the only way for him to survive is to pick berries from the trees. His problem is that he is not able to get many of them and must work most of his time, just to be able to survive. If Robinson had a suitable stick he could pick berries with less effort and therefore, either get more berries in a single day or work less time for the same number of berries he used to eat before. The problem is that in order to create/pick a suitable stick Robinson needs time just for this purpose but this will prevent him from picking berries and he will starve. The way out of this unfortunate situation is for Robinson to save some berries, i.e. to curtail his current consumption and to create over some days a stash of berries, which he could use to sustain himself during the period of stick production. After the necessary amount of berries has been accumulated he creates his stick and becomes much more productive.

We must note several obvious facts. First, the Robinson economy does not posses any capital and second, it does not use money (i.e. a barter economy). In economic terms what  Robinson does can be explained in the following way: He produces the usual amount of consumer goods (berries) for some days but sets some of them aside. In a later time period he consumes his saved consumer goods (berries, now considered capital goods) while producing fixed capital (a stick) in the meantime. He invests his effort and capital (stashed berries) into the production of a durable capital good.

This is the simplified view that determines the contemporary ideas what constitutes saving and investment. The problem with it is that it is derived for the most simple, basic case of economy and due to this fact alternative ways to implement saving/investment cannot be observed. In other words: the model used is oversimplified. In particular we note that in order to save Robinson must first produce the berries and just then, when the stashed berries are available can he use them. In a contemporary economy the process of saving includes money however. Thus Robinson would save not berries, but money. However, whereas berries constitute a consumption (or a capital) good, money does not. Money is not used up in consumption or production, it is the ultimate exchange good in fact. Thus when Robinson saves money he does not implicitly save consumption goods. In addition, since he has curtailed his consumption (does not spend the money for consumer goods) less consumer goods are being produced on the market as a whole. But what immediately follows is that if he lends his saved money to a capital good producer he does NOT implicitly transfer consumption goods. And the reason is very simple: by not consuming he has decreased the demand for consumer goods and in a short transitional time period the consumer goods produced by the market will decrease in order to comply with (match) this fact. In fact, if Robinson is the only one to save, then the rest of the people will have at their disposal the same amount of consumption goods. Robinson will be the only one in the whole economy who will get less.

What we explained above stays in a stark contrast to the contemporary economic views that Robinson hands out consumption goods to the capital producers for investment. The contemporary Austrian economics views assert that Robinson gives the capital producers consumption goods that they use for paying their workers and for sustaining themselves. Let us quote Ludwig von Mises (1990, ch.4): ”Capital goods come into existence by saving. A part of the goods produced is withheld from immediate consumption and employed for processes the fruits of which will only mature at a latter date”. Or we can quote De Soto (2012, pp.275) (more explicit): ”Saving always results in capital goods, even when initially these merely consist of the consumer goods (in our example the “berries”) which remain unsold (or are not consumed). Then gradually some capital goods (the berries) are replaced by others (the wooden stick) as the workers (Robinson Crusoe) combine their labor with natural resources through a process which takes time and which humans are able to go through due to their reliance on the unsold consumer goods (the saved berries). Hence saving produces capital goods first (the unsold consumer goods that remain in stock) which are gradually used up and replaced by another capital good (the wooden stick)”.

As we explained however, this is simply not the case in a contemporary economy. And the reason is simple: Robinson has decreased his consumption of consumer goods, but instead of these goods being produced first and handed out to the capital producers later (as in the basic Robinson economy) they are simply NOT produced. But something which is not produced (does not exist) cannot be given to anybody. Let us give an example to clarify the situation. In a particular economy the consumers are used to going to a restaurant let us say once per week. Suddenly they decide to eat out once every two weeks. The result will be that if the restaurants have produced X amount of services before they will have to produce half of that amount (X/2) after the change of the consumer preferences. The non-produced restaurant services however simply do not come in existence. But since they are  not created there is no way for them to be transferred to somebody else (i.e. to the capital producers). Note that we have purposefully used services in our example but the same situation can be described with material goods.


  1. How saving turns into investment in a monetary, capitalistic economy

Once we explained what the real problem with the contemporary views is we can try to clear up the confusion. Let us figure out what happens in a contemporary economy when Robinson saves money out of his monetary income. The first direct consequence will be that the demand for consumer goods will go down. After some (transitional) time the consumer good producers will realize that the demand has slackened. They will adjust to the new market situation by producing less, i.e. they will curtail the production of consumption goods. What that means is that resources will have to be freed.  In the case of the previous restaurant example this means the following: People will be laid off, buildings will be left free (the restaurants themselves), the capital equipment used for the restaurant business will not be used (bought). In addition intermediate capital goods, such as meat, flour, oil will not be used (bought). Note that the above represents all basic types of capital : original means of production (land and labor), fixed (houses, cooking stoves, tables,etc.) and circular (unprocessed meat, etc.). But let us now imagine what will happen to the suppliers of these restaurants. The suppliers of restaurant equipment ( of stoves for instance) will also realize that their products are not needed. Thus they will also curtail production, again releasing people, production facilities, etc. And their suppliers (of metal parts for stoves) will also have to do the same. And the suppliers of the suppliers (which create the metal for the metal parts) will also have to release resources. What we observe is the following: the initial act of saving causes the release of capital (original, i.e. land and labor, fixed and circular one) up the chain of production. A simple change of consumption habits affected all branches that produced intermediate capital goods geared towards the production of consumption goods and services (restaurant services in particular). To summarize what has happened: The act of saving led to a release of capital of all sorts in the industries which were dependent on the production of the particular good, whose consumption has decreased. And this is the capital that can be used for investment afterwards. Up to this moment it was geared towards the final production of consumer goods, but once released it can be geared into another direction, namely towards the production of fixed capital goods such as trucks, oil-tankers, etc. The latter will happen because the saved money will fill the bank coffers (to keep the situation simple). Banks will lower the interest rates and extend credit to producers of durable capital goods. For example, the building of a new ship yard will be started. Thus a part of the resources will flow towards ship building (i.e. labor, building materials, machines, etc.). The needs of the ship-builders will have to be matched by their suppliers however. Thus another part of the freed capital will flow towards the suppliers and later to the suppliers of the suppliers. Thus the capital released will find new uses, all of them with the final goal to produce durable capital goods. Note that no additional consumer goods for support during the transition period and during the time necessary for the production of the ship-yard are required. The same people who worked in the restaurants before will now work in the ship yard (for example) and since they are employed and paid they can provide for their own existence. No previously accumulated stash of consumer goods is needed. The overall production of consumer goods has decreased permanently and that is why the consumption of the workers from this moment on will be lower. They have in fact voluntarily agreed to consume less than before the transitional process took place. Thus the so called “subsistence fund” (Mises 1912), which supposedly consists of saved consumption goods or alternatively of saved intermediate products used for producing consumption goods, does not play any significant part in a contemporary monetary, capitalistic economy. Providing for the needs of the economy out of current production is the most efficient use of capital possible, since no capital is idle at any time, i.e. no consumer goods storages need to be built or supported.  

The above can be exemplified with the building of an industrial plant. Building a plant takes several years at least. During this time the people working on it must be fed and this happens from current production, not from some previously saved amount of food for consumption or from saved (delayed) intermediate products for food production (such as saved wheat) that are brought to completion. The same applies to the materials used for building the plant. Steel is being created on demand, i.e. iron ore is typically mined and processed to steel in the same year (time period) in which it is needed. No previous stash of steel or iron ore (an intermediate product) is being present.

Note that the above-given description is simply a description of a transition of an economy which moves from producing less to producing more durable capital goods. How secular growth happens and what the differences are will be handled in the subsequent part of the paper. Still, we must note some significant differences when comparing the discussed process to the one present in the simple Robinson Crusoe economy. In the Robinson case (no capital, no money) we need to produce consumption goods first and only then can we start the production of new capital goods. In a contemporary economy however this is simply not necessary. The intermediate step (producing consumption goods for sustaining Robinson) is simply passed over. What this shows is that the contemporary economy is much more effective in using the available capital. In particular less capital will be used (just the one necessary for the production of the final durable capital goods) and much less time will be needed (no time for the production of consumer goods is required). A contemporary economy has more degrees of freedom than the basic Robinson Crusoe economy. It has a capital structure and money, i.e. additional ways for capital redistribution exist. In other words a contemporary economy is much more flexible and efficient with respect to capital transfer and capital usage.


  1. Accumulation of savings, capital depreciation and growth in a monetary, capitalistic economy

Once we have dealt with the way how saving is converted into investment in a contemporary economy we could deal with the problem how these savings are accumulated and used.

In every economy the amount of capital available and in particular the fixed capital (machines, tools, buildings, etc.) constantly depreciates. In other words this capital wears out and needs substitution if the particular economy is expected to grow. At present there are two significant Austrian economics views how an economy can compensate and overcompensate this loss of capital so that economic growth becomes possible. In the first one, represented in Hayek (1967) and Rothbard (2009) the rate of interest must constantly go down for a sustained economic growth to take place. Thus the economic agents are supposed to not only save but save progressively more with the growth of the economy. In the second, more popular one (Garrison 2002) the rate of interest can stay the same but the economic agents must nevertheless save a constant part of their monetary income. What is common between these two views is that net saving is always necessary for an economy to grow ( increasing in the former case and constant in the latter). In fact both theories claim that net saving is necessary in order to offset the existing depreciation and to allow the economy to grow. What this means in practice is that both common workers in companies and the businessmen who run these companies must put a part of their monetary income/profits either in a bank account or in another saving option (for instance by buying company bonds or shares). The latter must happen on the average (i.e. not everybody needs to save) and must be continuous (i.e. saving never stops).

We intend to discuss what the consequences from the above-given view are for a banking system in an economy with invariable money. We will further accept that under invariable money the banking branch of the economy functions in the same way as any other economic branch (i.e. retail, ore-mining, etc.). In other words, we will presume that the conclusions we derive for the banking system will apply to the other branches of the economy as well. The reason why we have singled-out banking is that the processes we will describe are simpler to observe and clearer there from an economic standpoint.

Banks always have some amount of own capital, but most of the capital comes from outside, i.e. from savers who entrust the bank with the task not only to protect the amount of money they have put in it, but also to increase it (i.e. they expect to be paid interest on their bank account balances). Banks on their side lend the money they have obtained against interest. They are motivated to protect the monetary capital they have and therefore they lend it at such a rate of interest and choose their clients in such a way that they generate profit. Thus on the average the capital in the banks does not decrease. What this implies is that whatever capital is being lost ( e.g. bad loans and bank capital depreciation) will have to be compensated. And it is being compensated from the banking revenues. Let us however assume the contrary and see what follows from it. We will first presume that bank capital depreciation must be compensated by additional saving on the part of the banking clients. Savers however will expect that the additional money they put into their savings accounts is also protected, i.e. they will expect that their money does not decrease, i.e. that it is not lost. If this was not the case they would simply withdraw their money from the bank. But the latter would imply that the additional savings cannot be used by the banks for compensating the depreciated/lost capital. What is more, for steady growth in time we must presume that savers never stop saving. Still, reality shows that savers save for something, i.e. when a particular amount of saving is reached people stop saving on the average. But they should never stop saving if the depreciated capital is to be compensated constantly. All of the above leads us to a contradiction, which implies that our initial supposition is not correct, i.e. that the depreciated/lost capital in banking is not compensated by net saving on the part of their clients.

Up to this moment we showed that external savings (from clients) can never be used for compensating the depreciated/lost capital, which is a significant conclusion by itself. But what about the banks themselves? May be they themselves save and compensate the mentioned depreciation? The answer to this question is negative, since from the point of view of banks (and businessmen in general) the money used for compensating the depreciation is viewed/classified as explicit costs. In the same way as for ordinary workers net saving comes out of net income, in the case of businessmen net saving can come only from profits, which are the difference of revenues and costs. But since the money for capital maintenance represents explicit costs, it can never come out of net saving on the part of the businessmen.

What all of the above shows is that this branch of the economy can protect its capital without the need of net saving in general. But still depreciation must be compensated somehow and this is done by companies by passing their depreciation costs to their clients, i.e. by raising their selling prices, as a part of their revenues.

Note that the usage of net saving for capital depreciation contradicts the view that savings represent accumulated purchasing power to be exercised later. If net saving was used to compensate the capital depreciation it would be lost forever.

A corollary from the above discussion is that capital depreciation losses are shown to be identical to the general capital losses (due to bad management decisions, economic disasters, bad luck, etc.). Both types of losses are being covered from the revenues of the (successful) companies. Thus singling out depreciation as the only loss in an economy, which is to be compensated, is an erroneous approach. One should ask himself the question how all the losses inherent to an economy are to be compensated.

Up to this moment we were discussing only banking. The rest of the branches of the economy however are not supposed to work differently under invariable money. Thus in them capital depreciation will also be compensated from their revenues. From here we can generalize that since an economy is in fact all economic branches taken together and since no net saving is required in any branch, then net saving is not necessary for the existence of the economy in general. In other words, net saving is not necessary for compensating the economy – wide capital losses.

Once we have dealt with the net-saving issue we can discuss what the source of capital used for compensating the capital depreciation is. As we discussed above, depreciation losses are being compensated from the revenues. Thus we have to explain how on a macro level there is an excess of capital available for compensating losses (and depreciation in particular). In other words, how it is possible for the companies to pass their losses to their clients without decreasing the amount of gross saving in the economy as a whole. Here comes the contribution of George Reisman [Reisman 1990], who noticed that gross saving is able to reproduce itself by itself. In particular Reisman noticed that an economy produces not only consumer goods but fixed/final capital goods as well. The latter is not clearly visible in the contemporary views about the structure of an economy, i.e. in the Hayekian triangle [Hayek 1967]. In such a structure the fixed/final capital goods are supposedly just an intermediate station towards the production of consumer goods. They are not explicitly shown (and can not be shown in general), which makes them practically invisible, but still they exist. Note that it is the fixed capital goods that depreciate in an economy, not the circular capital goods or the original means of production (land and labor). But if we admit the fact that an economy produces final/fixed capital goods, then we must realize that these newly produced fixed capital goods can be used for compensating the depreciated/lost ones. Then it becomes a matter of quantitative difference between the produced and depreciated fixed capital goods to determine if an economy retrogresses, is in a stationary state or grows. If we manage to reach the state where the produced fixed capital goods compensate the depreciated/lost ones we will find ourselves in a situation in which the economy self-supports itself. Once we have reached it, we need not put any additional effort in order to create capital good, i.e. no net saving will be necessary. The process will become self-sustaining, i.e. it will feed on itself. In the exact sciences such systems are called positive feedback system. These systems do not require to be driven, they drive themselves.

Now the question we have to ask ourselves is how to reach such a state, in which an economy produces more fixed capital goods than it looses. Let us remind ourselves that the production of capital goods and consumption goods are complimentary ones. In other words, we can produce more of the one type only at the expense of the other type. But at any moment the overall productive ability of the economy is limited. The latter is best exemplified with the famous Guns and Butter graph [discussed in Garrison 2002]. What it shows is that for an economy at a particular state of development we can use the available resources for producing different goods, but we have a technologically determined limit, which we cannot surpass. Still, it is a matter of our choice where we decide to be on the particular graph. Thus every economy has a choice to produce more capital goods (fixed capital in particular) at the expense of consumer goods or vice-versa. Note however that the particular amount of capital that the economy possesses will depreciate in the same way, independently of what we produce. Thus the depreciation rate is fixed, but how much new fixed capital we produce depends on us. And if we decide to place ourselves in a situation (on the graph) where we produce enough capital goods to counter the depreciation we will find ourselves in a stationary economy. In the same way, if we decide to produce more capital goods than the ones being lost due to depreciation we will find ourselves in a growing, self-sustaining economy of the positive feedback type discussed above.

What we established so far is that an economy can grow by self-supporting itself, which in practice means that no net saving will be required on the part of the economic agents. Thus an economy can have zero net saving and still have positive net investment. Still, the question stays how we move the economy from producing less capital goods to producing more, so that we compensate the lost/depreciated capital. The answer to this question is simple: by saving, as described in section 3. Thus if we do not produce enough fixed capital goods to offset the existing depreciation we start saving, i.e. let net saving get positive. What will happen is that less consumption goods will be produced and the released economic resources will be re-targeted towards the production of fixed capital goods. But once we find ourselves in a situation in which we produce enough fixed capital goods to over-compensate the depreciation/losses, we can stop net saving and the economy will still grow.

From now on I will call the amount of capital, used for capital substitution and investment the “investment fund”. In contrast to the “subsistence fund” the investment fund consists of all types of economic resources, not only of consumer goods. The investment fund includes all three basic types of capital, namely, original means of production (land and labor), fixed and circular capital used in all the stages of production. The latter can be observed when one realizes that banks give loans for hiring labor ( original means of production), for buying machines and buildings (fixed capital) and for buying supplies for production (circular capital). The “investment fund” is self-sustaining and it is being constantly invested and reinvested. It is the extension of the concept of “subsistence fund” into an economy that in contrast to the one of Robinson possesses capital and money. The subsistence fund as such is meaningless for a contemporary economy and therefore cannot be a part of a modern theory of capital.

It is instructive to describe a simple economy which grows with zero net saving. A simplified example is an economy under invariable money in which the only way to save and invest is through the banking system. Thus banks have a constant amount of money, which they use for investment into the economy. However, there is no net saving, since on the average economic agents will not save. The saving of some (e.g. young people, saving for retirement) will be offset by the dissaving of others (i.e. retirees drawing down their savings). The latter will guarantee that the amount of bank money used for investment purposes will stay constant. What we observe is that there are funds, which will be invested, i.e. some companies will get loans to support their operations, thus being able to increase their capital. But at the same time all companies in the economy will lose capital due to the capital depreciation. And if the combined bank investment leads to more new capital than the one being lost the economy will grow, i.e. net investment will be positive.  If the bank investment is equal to the depreciation, the economy will stagnate, i.e. net investment will be zero.  And lastly, if the investment is less than the depreciation the economy will retrogress, i.e. net investment will be negative. Still, note that investment is present in every case, although net saving is zero. There are funds in the banks, which are used for this purpose. The only question is if these funds are enough for economic growth. If we find ourselves in a stationary economy, we can make it grow again if the economic agents increase the amount of money in the banks, i.e. if they save some additional money in the banks and later stop saving. In other words we let net saving be positive for some time, thus increasing the investable bank capital and lowering the rate of interest. But once the bank capital has been increased sufficiently, what actually means that the “investment fund” has increased, the economy starts growing with the already lowered rate of interest and zero net saving.  All of the above stays in full compliance with George Reisman’s views (Reisman 1990, ch. 15) but described from a different perspective.

In view of all of the above we can generalize that an economy can grow with positive net saving (well known), with zero net saving (just discussed) and even for a while with negative net saving. The latter may happen if the investment fund has initially been big enough to guarantee economic growth and its decrease due to the negative net saving has not brought it to equality with the depreciation yet.

We could contrast the process of secular growth we just described with the transitional process we described in section 3. During the transitional process net saving was present and it lead to re-directing resources from lines of production leading towards final consumer goods to lines of production leading towards durable (fixed) capital goods. A shift of capital towards the investment fund happened. During the process of secular growth described above no net saving is present but still net investment exists. No capital is shifted in or out the investment fund but still the economy grows because investment in all lines of production is present.


  1. Conclusions

In a contemporary economy the processes of saving and investment are based on money. Thus the availability of a previously accumulated amount of consumer goods or of intermediate products of the consumer goods in order to make investment possible is simply not necessary. In this way, in contrast to a Robinson Crusoe economy (no capital, no money), saving-induced investment takes place without producing consumer goods first.

We showed that every contemporary economy has an “investment fund”, i.e. an amount of free capital which is self-sustaining and is used for capital substitution and investment. When the size of this investment fund is big enough it can compensate the existing capital depreciation and let the economy grow without net saving being present. In this way we show that an economy supplies itself the necessary funds for compensating the capital depreciation and all economic losses in general.




De Soto, Jesus (2012), Money, Bank Credit, and Economic Cycles. Auburn, Alabama: Ludwig von Mises Institute.

Garrison, Roger (2002), Time and Money., New York, NY: Routlege

Hayek, Friedrich (1967), Prices and Production. Auburn, Alabama: Ludwig von Mises Institute.

Mises, Ludwig (1912), Theory of Money and Credit. Liberty Fund

Mises, Ludwig (1990), Economic Freedom and Interventionalism: An Anthology of Articles and Essays. Auburn, Alabama: Ludwig von Mises Institute.

Reisman, George (1990), Capitalism. A Treatise on Economics. Laguna Hills, California: TJS Books.

Rothbard, Murray (2009), Man, Economy, and State with Power and Market. Auburn, Alabama: Ludwig von Mises Institute.