Beyond the liquidity trap: ineffectiveness of monetary policy as an instrument to neutralize demand and supply shocks
Adam Koronowski *
Abstract:
The aim of this paper is to discuss limitations,other then nominal zero bound, to effectiveness of monetary policy as an instrument to neutralize a negative impact that demand or supply shocks may have on production and employment. Firstly, intent to boosteconomy with expansionary monetary policy may contradict central bank’s primary target of stable prices. The paper asserts this may be a case even when nominal and real interest rates are positive and the economy experiences a severe contraction. Secondly, counterbalancing negative shocks may require negative real interest rates. However, negative real interest rates would create costs of wasteful use of capital. A question arises whether monetary policy should set under any circumstances negative real interest rates. A conclusion comes that monetary policy cannot neutralize negative supply shocks without increase in inflation. “Negative” demand shocks – increase in propensity to save – allow for lower real interest rates in equilibrium and are welfare improving as long as there is no problem of the nominal zero bound. This inference allows for non-Keynesian effects of fiscal contraction. The paper presents also limitations to smooth functioning of a currency union when there are considerable differences in regional (national) savings patterns.
*National Bank of Poland, Faculty of Economics – University of Warsaw
Comments welcome:
Views expressed in this paper are those of the author and not necessarily represent those of the National Bank of Poland.
I Introduction
Ineffectiveness of monetary policy as an instrument to counteractrecessionary pressures is typically related to the concept of liquidity trap. This traditional concept asserts that monetary policy is ineffective in a situation when increase in money supply does not result in falling interest rate but merely in growth of idle balances; the interest rate elasticity of demand for money becomes infinite. According to Keynesian tradition this is the case when individuals believe that bond prices are too high and therefore will fall and, correspondingly, that interest rates are too low and must rise. This interpretation is awkward and obsolete when considered from the perspective of endogenous money supply. A central bank can generally cut the rate if it believes it is too high and it really wants to boost the economy via more expansionary monetary policy. Such an expansion is inefficient in promoting more spending only in a situation when nominal interest rate is zero; individuals could avoidnegative interest rates by holding more money. This is nominalzero bound problem which may be considered as a modern interpretation of the liquidity trap.[1]
The aim of this paper is to discuss other possible limitations to effectiveness of monetary policy as an instrument to neutralize a negative impact that demand or supply shocks may have on production and employment.
Firstly, intent to boosteconomywith expansionary monetary policy may contradict central bank’s primary target of stable prices. This may be a case even when nominal and real interest rates are positive and the economy experiences a severe contraction.[2]Such a situation is beyond the scope of these concepts and models which consider inflation (recession) as an effect of spending that is excessive (deficient) as compared to a certain potential level of production so that nominal and real developments have the same expansionary or recessionary direction. This attitude is clearly unrealistic as a general paradigm and it is enough to call the notion of stagflation to prove this opinion.
Secondly, counterbalancing negative shocks may require negative real interest rates. In case of sufficiently high nominal interest rates (and rates of inflation) the nominal zero bound may not constitute a problem. However, negative real interest rates would create costs of wasteful use of capital. A question arises whether monetary policy should set under any circumstances negative real interest rates. This is not a technical problem, as in the case of nominal zero bound, but a matter of policy optimization.[3]
In the second part of this paper a simple model of macroeconomic production and equilibrium is developed. The third and fourth parts present within the structure of the model the consequences of supply and demand shocks respectively and formulate conclusions on whether monetary policy can neutralize their impact on production or employment. The fifth part takes short notice of some outcomes of the model which shed light on functioning of a monetary union. The article ends up with a resume.
II Model
The first and central equation of the model is production function:
1. Y = aKL , where
In eq. 1 Y stands for the real domestic production (income) and K and L for capital and labor engaged in production.
The condition expresses decreasing returns to scale. This condition is crucial for the model. Only limited amounts of labor and capital can be engaged in production due to decreasing returns which must cover some given prices of the factors of production. As a consequence, domestic product may remain below full utilization of available resources. When the amount of capital is fairly flexible at least in a longer term – due to investment and capital depreciation – the factor of production which may remain idle is labor; the model allows for unemployment.
Since the condition has much importance for the model and it is not the conventional assumption of constant returns to scale, it needs some justification. It seems fairly uncontroversial that there should be at least one factor of production other then labor and capital (such as land, human capital, natural resources). Let’s assume its amount is constant. This means diminishing returns of capital and labor. If the factor(s) of production non-specified in the production function is Z then
2. Y = a*KLZ , where
Since Z is a constant eq. 2 is none other then eq. 1 with the condition of decreasing returns to scale given only two factors of production are considered openly (they define the scale).
From eq. 1 we get
3. MPK = aKLand
- MPL = aKL,
where MPL and MPK is marginal product of capital and labor respectively.
Marginal product of labor equals real wages (w) which for simplicity and to focus attention on monetary aspectsare assumed to be constant.[4]
- MPL = w.
Marginal product of capital must cover the real cost of capital which equals real interest rate (r) and the rate of capital depreciation (
- MPK = r + .
Eq.. 3 – 6 give
7. (r + / w = MPK / MPL = L / K and
- K = wL / (r + .
Eq. 8 and eq. 3-6 allow to determine the amounts of capital and labor used at its’ given costs:
9. L = [w / a] [(r + / w]and
- K = [(r + / a] [w(r + ].
Eq. 7, 9 and 10 have some important properties. Interest rate cut brings about – as eq. 10 predicts - more capital engaged in production. It also leads to higher employment according to eq. 9. However, technology becomes more capital-intensive (eq. 6). It is easy to show that product increases. Similarly, wage reduction leads to higher employment, more capital engaged, more labor-intensive technology and production increase.
Eq. 9 and 10 describe the amounts of labor and capital that can be effectively used in production. This need not be consistent with full employment. Moreover, none of the equations above touches upon the issue of equilibrium in the market for goods and services.This problem may be addressed as in eq. 11 which is a condition for equilibrium in the goods market. This resembles the respective condition in Solow’s model, however, it is not expressed in terms “per labor unit”.[5]
11. mY, where m is a constant and 0<m<1.
Eq. 11 – together with eq. 1, 8, 10 – determines a level of the interest rate consistent with goods market equilibrium. This interest rate level is described by eq. 12.
12. r / m) – 1]
The equilibriuminterest rate from eq. 12 (through eq. 1, 8, and 9) determines how much capital and labor is engaged in production and how big the productis.[6]
Let’s notice here that the equilibrium interest rate may take negative values. However, since m<1 it always holds that r> – 1). Moreover, extreme and improbable case when m=1(savings and investment in equilibrium is equal income) shows that sufficiently low negative equilibrium interest rate must reduce consumption which is price for wasteful use of capital in a steady state.
The model does not describe situations which emerge in case of interest rates below or above the equilibrium interest rates. It is only assumed here that interest rate above the level specified by eq. 12 implies that aggregate demand is insufficient to uphold current production and it results in recession (prices are inelastic downward), interest rate below the level fuels inflation. In this sense the interest rate from eq. 12 is natural.
Eq. 11 assumes a very simple form of consumption (and savings) function. Savings are not determined within a general equilibrium model based on micro-foundations. As a consequence, the model pertains only to steady-states when investment equals depreciation. It allows for comparative static only.[7]If the economy experiences macro disequilibrium, within the model a necessary condition to restore equilibrium is a change of the interest rate set by the central bank (changing fiscal stance not explicitly included in the model). The model does not supply a dynamic description of disequilibrium and in particular of a mechanism which could restore a recessionary steady-state in absence of monetary policy reaction. In case of overheating, a natural interpretation is that rising prices reduce real consumption and investment in relation to the levels planned in nominal terms of a previous period. As long as real interest rate is falling (nominal rate is constant) it results in ever growing inflation[8]. In case of recession a mechanism which should restore equilibriummight be looked for in dynamic feedbacks among savings, income and income expectations. Such analysis would require much more attention paid to the savings function than in the presented form of the model. However, it needs to be openly expressed that the model purposefully is not a general equilibrium model. In fact it is rather admitted here that recessionary pressure may lead to abandonment of first best procedures of micro-optimization: only its violation may lead to a new steady state of a recessionary macro-equilibrium. In particular, deficient demand may entail job reduction and - at constant capital in a short period – a fall in output combined with rising marginal productivity of capital. This would support investment and reduce savings (in parallel with income) and restore a macro-equilibrium. The conditions expressed by eq. 3 and 4 would, however, not be met in this case.[9] Of course, this a simple suggestion and it is not alleged here that during real life recession capital stock and investment would not fall. The example is only to show what role can be played by changing proportions of factors of production when disequilibrium is transferred from macro to micro level.
In this context it is worth noticing that savings levels (determined with respect to micro-optimization) would always be consistent with macro-equilibrium if they were dependent on endogenous interest rate that should match savings and investment. Any autonomous shift of the savings function would lead to a change of the interest rate consistent with macro-equilibrium. In such a case natural and actual interest rateswould be the same. However, in modern economiesnominal interest rates – and in consequence real ones - are exogenously set by central banks (and money supply is endogenous, not the opposite).
The formula for the natural interest rate given by eq. 12 does not contain real wages. However, more sophisticated consumption (and savings) function would make the natural interest rate depend on real wages. Then flexible wages could – formally – constitute a factor which allows for macro-equilibrium for any given level of interest rate set by the central bank. Even so, elastic wages considered as a subject to micro-optimization consistent with labor market equilibriumwould not help much with regard to macro-equilibrium. On the one hand, labor market equilibrium may notbe compatible with recessionary macro-equilibrium (or simply with the steady state balance between savings and investment) but at a particular (natural) level of interest rate; macro-equilibrium remains the task of the central bank (given fiscal stance).On the other hand, even though – due to changes in techniques and amount of capital employed – there may be a level of wages that guarantee macro-equilibrium for any given interest rate it would not be consistent either with micro-optimizationor full employment. There is no reason why wages should be a factor which restores macro-equilibrium.[10] Moreover, if wage changes were to restore macro-equilibriumat any interest rate there would be nothing like the natural rate of interest. It seems thus reasonable to approve of the feature of the model that wages are not considered as a factor which allows for macro-equilibrium; wages are assumed to be inelastic. Also a form of savings function which eliminates wages from the formula for the natural interest rate is accepted.
Things are even more complicated. National savings should contain public savings or deficits. As a consequence a mutual relation between balances of private and public sectors should also be included in the analysis – for example in the form of the Ricardian equivalence. This problem is not formally recognized within the model.
It is a useful extension of the model to lift the implicit assumption of closed economy. Eq. 11 and 12 may be easily transformed to forms which should comprehend current account balance. Current account developments due to changes in absorption and in relative price levels (real exchange rates) – for example as a result of excessive demand – could be a short and medium term adjustment mechanism which restores internal equilibrium. This in turn opens questions on long term current account sustainability. Since this is not a task and ambition of the paper to present a coherent theoretical general equilibrium model ranging from dynamic, short term micro-optimization on the one hand to long-term current account sustainability on the other hand, the issue of balance of payment is not formally recognized in the model. It is enough to say here that excessive demand results not only in inflationary pressures but also in widening of current account deficits at fixed exchange rates.
Summing up, the model is not a general equilibrium model rooted in micro-foundations but it could be subject to some interpretations (or extensions) featuring dynamic mechanisms which should allow for recessionary macro-equilibrium (steady state) or lead to inflationary explosion. However, what is important here is not these mechanisms but the role that monetary policy can play to counterbalance shocks and restore equilibrium before any medium- or long-term endogenous mechanisms find their way.
III Supply shocks
A negative supply shock is interpreted here as a decline in production achieved form given amounts of the factors of production (capital and labor). Marginal products of capital and labor diminish and they are above their cost. Since diminishing returns of scale are a feature of the model, quantities of both factors engaged in production shrink as a result of a negative supply shock so that marginal products should grow and meet the factor costs again. In this context there are two important questions; 1. whether monetary policy can offset the shock and allow to maintain the level of production and/or employment, 2. what monetary policy stance is needed to maintain equilibrium.
A negative supply shock in terms of the model means the parameter a of the production function has decreased; a2<a1. Let’s assume that the aim of monetary policy is a constant level of employment, L2 = L1. Using eq. 8 for two different values of the parameter a and two different interest rates under the condition of stable employment we obtain
- r2 + r1 + (a2/a1)1/.
Eq. 15 shows that lower interest rate (r2r1)is needed to compensate for the impact a negative supply shock exerts on employment. It is easy to prove that the condition of constant production gives the same solution (eq. 15).
A few conclusions are worthwhile. Firstly, even when r1>0 it is possible that r2<0; it may need negative real (and probably nominal) interest rate to counter the shock.Secondly, the shock has no impact on natural interest rate given by eg. 12. Should interest rate be cut with respect to employment and production considerations, it would give rise to excessive demand and – as a consequence - inflationary pressures and current account deficit widening. This is intuitively understandable; at a constant level of production and savings (given the assumed savings function),at lower interest rate, higher capital engaged and higher investment there must be excessive demand. Thirdly, in case of a modest monetary reaction to the shock and interest rate set between r1 and r2 there should be both economic contraction and inflation – the model admits stagflation. Fourthly, it is worth lifting the assumption of inelastic real wages to note that wage reduction could boost production and employment – it could offset the real effects of the shock – at a constant, natural interest rate.[11]
IV Demand shocks
A negative/positive demand shock can be expressed in terms of the model as decrease/increase of the propensity to save (m2>m1 or m2<m1 respectively). First of all let’s notice that in face of a demand shock monetary policy has to react to disequilibrium and not to fall in equilibrium level of output, which was the case of a negative supply shock.
In the case of decrease in propensity to save natural rate of interest is rising. To stop inflationary pressures central bank has to raise its interest rate. Until now interpretation of this situation is easy – excessive demand leads to inflationary pressures and monetary policy contraction. A part of such interpretation might be, however, inference that demand has grown above the level of potential product and interest rate hike has returned demand and production to its’ previous levels. This inference is false; as interest rate is raised to its new natural level, product declines. Anti-inflationary monetary policy as a reaction to consumption boom entails costs in real terms. This conclusion reflects the fact that the model encompasses both supply and demand aspects of economy.