Swedish Institute for Social Research (SOFI)
Stockholm University
WORKING PAPER 5/2020
EXPLAINING INFLATION AND UNEMPLOYMENT
by Ante Farm
Explaining inflation and unemployment
Ante Farm
Swedish Institute for Social Research (SOFI), Stockholm University www.sofi.su.se
E-mail: ante.farm@sofi.su.se
Postal address: SOFI, Stockholm University, SE-106 91 Stockholm, Sweden
September 19, 2020
Abstract: This paper explains inflation and unemployment starting from updated models of price formation and labour demand. Inflation is always and everywhere a pricing
phenomenon. Unemployment is always determined as a residual, as people in the labour force without employment. This applies to both the short run (1-2 years) and the medium run (3-10 years). Employment is determined mainly by production and labour productivity, while production is determined by spending at prices set by firms.
Keywords: Inflation, employment, unemployment JEL-Code: E24, E31
1
This paper introduces an approach to macroeconomics that is based on an update of perfect competition with price leadership and an update of the traditional model of labour demand with price formation. It shows that employment every year, and not only in the short run, is determined by production and labour productivity, while production is determined by capacity or spending at prices set by firms. Unemployment is determined as a residual, as people in the labour force without employment, even in the medium run. The paper also shows how the effect of matching problems on employment and unemployment can be measured in business surveys.
Inflation is determined by changes in wages, labour productivity, import prices, mark-ups and capacities in the consumer industry, even in the long run. Changes in wages depend crucially on whether wages are set unilaterally by employers or in negotiations with unions. Real wages are determined partly by price formation in the consumer industry, partly by how relative wages are determined in other industries.
While Sections 1-5 deal with microeconomics, the corresponding macroeconomics is formulated in Section 6. The model first relates inflation to changes in wages. Secondly, it relates changes in production to changes in aggregate demand, as measured by changes in nominal GDP, at prices set by firms. Thirdly, it relates changes in employment to changes in production. Finally, it relates changes in unemployment to changes in employment and changes in the labour force.
1. Price formation and labour demand
A firm’s production in the market sector is, as a rule, restricted by its sales at the price (p) it sets. Having announced its price, it therefore adjusts its production to its sales D p and its
employment N to its production and labour productivity (a). In general, however, (1) N D p a
if D p
and otherwise N k ak ,where k denotes the firm’s capacity. And the price is, when production is restricted by sales, determined by indirect cost pq = w, so say
(2) p
1 m
1h w a
,2
where w is the wage level, h direct costs other than labour costs measured as a share of labour costs, and m is a mark-up that the firm chooses to cover indirect (fixed) costs and a normal rate of return on capital or to adjust to price competition as a price leader or price taker.1 Of course, a firm can be a price taker, that is, set the same price as a price leader, preferring a lower market price. Before prices change, due to cost changes, prices in the market are normally approximately the same, according do to the law of one price.2
This is a very general formal model. It assumes that prices are set by firms adjusting prices to changes in costs. In all cases, however, it is only the price revisions after cost changes that are entered as inputs in the market process. Firms decide on initial price and let a firm’s
customers decide on the output that will be sold – long before the output can be sold.
As emphasized in Farm (2017a) and Farm (2020a), the model above applies primarily to consumer markets, where we can assume that all buyers take prices as given and prices are set by sellers. But it may also be applicable to markets with sealed bidding and other
business-to-business markets, even if a powerful buyer may affect the mark-up in these cases.
Even in commodity markets, where the market price is determined by equality between offers to sell and offers to buy, firms submitting offers to sell a commodity will relate their price offers to their costs of producing the commodity.
2. Wage formation
According to eq. (2), the real wage for employees in firms with the wage level w is
(3) w pc
w wc
w pc c
,wherew denotes the wage level and c p the price level in the consumer industry (comprising c firms importing, producing, or distributing consumer goods). Assuming, to begin with, that production is restricted by sales at prices set by firms in the consumer industry,
(4) pc
1 mc
1h w ac
c c.1 Not all firms can take the market price as given, as in perfect competition. But all firms except one can, implying price leadership. Price leadership is motivated in more detail in Farm (2017a, b) and related to cost- plus pricing in Farm (2020a). That labour demand depends on product demand is proven in Farm (2020d).
2 In a market for differentiated goods prices cannot differ too much. In practice, individual prices differ from the
“price level” for a variety of reasons, including time of entry to the market, etc.
3
It follows, firstly, that the real wage in the consumer industry,
(5)
1
c1
c c
c c
w p a
m h
,
depends on the industry’s labour productivity (a ), mark-up (c m ) and input structure (c h ), c which depends on import prices and hence also on the exchange rate. Secondly, the real wage in other sectors of the economy is determined partly by the real wage in the consumer
industry, partly by a sector’s relative wage, according to (3).
Some basic results follow immediately. First, the real wage of all employees depends on the real wage level in the consumer industry or, more precisely, on labour productivity, import prices, and mark-ups in the consumer industry. Second, the nominal wage in the consumer industry affects inflation (as measured by a consumer price index) but not the real wage for the industry’s employees. Third, the real wage for employees outside the consumer industry depends on their nominal wages, more precisely their relative wages, compared to wages in the consumer industry.
Thus, workers in the consumer industry can raise their real wages by contributing to higher labour productivity (a) and more efficient use of inputs and capital (reducing h and m). Their nominal wages cannot affect their real wages, but they can affect inflation. If workers in the consumer industry prefer a low and stable inflation rate, say 2 percent, they can also adjust their wage claims to this target and expected labour productivity and import prices.
Moreover, employees in other sectors of the economy can only change their real wages by affecting their nominal wages or, more precisely, their relative wages according to (3). Thus, by adjusting their wage claims to the development of wages in the consumer industry, employees in other sectors can achieve the same real wage level as the consumer industry, suggesting that some sectors can be “wage takers”, accepting the consumer industry as a wage leader. However, since employees outside the consumer industry can obtain even higher real wages by higher nominal wages, they may, at some stage, attempt to do this and also succeed, giving rise to conflicts between unions and wage-wage inflation.
Now, this is a narrative which presupposes an organized labour market with strong unions, as in Sweden. Being a small open economy with a large export industry, the consumer industry partly overlaps with the export industry, which also indirectly supplies the consumer industry
4
with imports. In fact, wage leadership dominated the Swedish labour market after the Second World War, up to about 1970, with the export industry as the leader. Then a period of
confrontation between labour and capital followed, accompanied by wage pressure from new unions, organizing white color workers in the market sector and employees in the public sector. However, since 1997 the export industry, in agreement with its unions, has been a wage leader in Sweden, determining the growth rate of the wage level not only in the export industry but also in other sectors of the economy.3
Of course, there can be wage leaders even if wages are set unilaterally by employers, like a firm that dominates a region or an industry. In an economy dominated by its export industry, a few big firms in the export industry may even be a wage leader for the whole economy. But how will wages be revised if wages are set independently by firms?
When employers revise their wages in industries without unions they cannot normally do this without communicating with their employees. In a tight labour market, workers may also be tempted to look for employers offering higher wages, and firms suspecting that other firms will attract workers with higher wages may find it necessary to raise their wages. Thus, without coordination between firms in an industry, and between industries in an economy, wage competition between employers may raise the wage level when firms find it difficult to replace quitting workers or recruit new workers at current wages.
On the other hand, in a labour market with increasing unemployment, firms may be tempted to cut their wages and prices to combat decreasing sales with larger market shares. And when all firms do this, deflation and mass unemployment may follow, as in the 1930s.
But how will a wage leader revise the wage level? Will the norm be wage growth in another country (as today in Sweden) or a target for inflation? How will the wage level grow when there is no wage leader in an economy? How will strong but competing unions complicate the picture? How much will wage competition between firms increase the wage level when wages are set unilaterally by employers? There are no simple answers to these questions, and the answers differ between countries, depending on different histories and institutions.
3 See e.g. Elvander (2002) and Farm (2020b) for an introduction to industrial relations in Sweden.
5
In any case, employers and employees or their organizations never bargain over real wages but over the growth of nominal wages. The outcome of this process decides the evolution of the most important price in a market economy: the price of labour.
3. Inflation
Sellers will not accept money as a means of payment unless money also is a store of value, which presupposes an acceptable amount of inflation. But the purchasing power of money – the value of money in terms of goods – depends on the prices of what buyers intend to buy, so the concept is ambiguous, as emphasized already by Knut Wicksell.4 Different agents may be interested in different goods and hence also in different prices and in different summary measures, even if a price index summarizing the development of prices of basic consumer goods is important for most people.
There are two possible regimes for price formation in an industry, and which regime rules depends on the relation between effective demand (spending) and capacity at prices set by firms. Of course, the regime is not necessarily the same in all industries: while some industries may have production restricted by capacity and prices determined by market clearing and effective demand, other industries may have production restricted by sales at prices set by firms.
Price increases in an industry depend entirely on demand if all firms are producing at full capacity. On the other hand, if production is restricted by sales at prices set by firms, then inflation is determined by firms when they revise their prices according to equation (2). Thus, when production is restricted by sales, inflation depends on the evolution of labour costs (w), variable costs other than labour costs measured as a share of labour costs (h), mark-ups (m), and labour productivity (a). In fact, it follows from (2) that
(6)
1 1 1 1
dm m dh h
dp p dw w da a
m h
.
Production depends entirely on the demand for firms’ products at the prices they set, while employment depends on production and labour productivity. Note also that increasing demand for consumer goods will increase production without raising prices as long as production is restricted by sales at prices set by firms.
4 See, for example, Wicksell (2010 ch. IV).
6
Inflation is usually measured by a consumer price index (CPI). An increasing consumer price index presupposes that a weighted average of prices on consumer goods is increasing, which presupposes price increases in at least some consumer industries. The first step in explaining inflation is consequently to relate it to inflation in different industries. The second step is to explain inflation in the various industries. In some consumer industries inflation may depend on increasing demand and (endogenous) market clearing when firms are producing at full capacity, while inflation in other industries may depend on increases in import prices or wage increases exceeding increases in labour productivity.
The Phillips Curve
According to the Phillips curve, inflation is affected by unemployment. However, since 1970 the empirical relation is rather weak.5 And according to (6) inflation does not depend on unemployment, at least not directly.
Of course, unemployment can have an impact on unemployment indirectly, by affecting wage growth in the consumer industry when production is restricted by sales at prices set by firms.
And according to the traditional framework of supply and demand, wages should go up when there is shortage of labour and down when there is excess supply. But how does the market mechanism function in practice?
Suppose first that unemployment is so low that firms have recruitment problems, at least in some consumer industries. If these problems are sufficiently serious, it may be tempting for a firm to try to keep or recruit personnel by raising its wages, either because it is hoping that other firms will not come up with the same idea, or because it is afraid that they do. And if all firms are caught in this dilemma wages will grow.
Thus, wage competition between firms can contribute to wage growth. However, employers can be conscious of this dilemma and agree on a common “wage culture” that excludes wage competition.6 Employees in the consumer industry should also realize that higher wages will not raise real wages, only inflation.
On the other hand, high unemployment may moderate wage growth in the consumer industry by decreasing firms’ recruitment problems so much that it eliminates wage competition
5 See e.g. Blanchard et al. (2015 p. 209) or Romer (2019 p. 258).
6 According to Faxén et al. (1989 passim).
7
between firms. High unemployment can also reduce workers’ wage claims, even if wage competition between workers does not belong to their “wage culture”.
Inflation and money growth
Inflation is correlated with money growth. But money growth can never be the direct cause of inflation, as measured by a consumer price index,7 even if it sometimes may be associated with the cause of inflation, namely excessive effective demand caused either by commercial banks creating money by lending8 or excessive government spending financed by selling government bonds to commercial banks or the central bank. Normally, however, inflation is caused by price formation and accommodated by money growth.
More precisely, lending by commercial banks increases the stock of money and may, when borrowings are spent, also increase the effective demand for consumer goods. But this affects inflation only when production is restricted by capacity, since when production is restricted by sales, effective demand affects production and employment but not prices. Moreover, even if an increasing stock of money may precede inflation, inflation may also precede money growth, since higher prices may increase borrowings in monetary terms.
Controlling inflation
The basic idea of modern monetary policy is that a central bank can control inflation by controlling effective demand (spending) through its policy rate. Of course, a central bank can always restrict effective demand by a sufficiently high policy rate. But will this also reduce inflation? A central bank can also stimulate effective demand by lowering the policy rate, even if sometimes not even a policy rate equal to zero may be a sufficient stimulus. But will this also raise inflation? To answer these questions, we have to distinguish between the two possible pricing regimes.
When effective demand for consumer goods is sufficiently small at current capacity, then sales and production will be less than the capacity of the consumer-goods industry at the prices set by firms. Prices consequently depend on wage rates, input prices, labour
7 However, an increasing money stock may contribute to increasing house prices.
8 On the creation of money by banks, see e.g. McLeay et al. (2014).
8
productivity, and mark-ups. Unless mark-ups change, inflation must then be due to wages increasing more than labour productivity, or to increasing prices of imports.
Of course, a central bank cannot affect labour productivity. It can affect import prices by affecting the exchange rate, which it can influence in the short run by intervening in the foreign exchange market or setting a policy rate which differs from foreign policy rates.
However, in a system with flexible exchange rates, interventions are not necessarily carried out, even if they by stabilizing a volatile exchange rate can stabilize import prices. A policy rate above foreign rates may result in an inflow of capital which appreciates the currency and hence moderates price increases on commodities and other imports. However, in a world with free capital mobility interest rates cannot deviate much.
How can a central bank (CB) affect wages? It cannot affect wage formation directly. Hence a CB cannot control inflation in a mark-up regime unless it can control wage revisions
indirectly, by creating and using unemployment as a disciplinary device when inflation is above the target, and reducing unemployment in order to increase wages when inflation is below the target. In practice, however, a revision of the wage level in the consumer-goods industry is the result of a much more complicated process, usually involving negotiations between employers and employees or their organizations, or wage competition between firms, as discussed above. It may even be exogenously determined, for example by a revision of the wage level in the export industry, which is accepted as a norm by all parties in the economy (as in Sweden since 1997).
Note that a higher policy rate will reduce spending on consumption – and employment in the consumer industry – but not inflation, when production is restricted by sales at prices set by firms in the consumer industry. Note also that deflation can only be explained by decreasing import prices or wages increasing less than labour productivity in the consumer industry. And then a CB is powerless, since it cannot raise wages by reducing the policy rate, unless this rate is high to begin with and reducing it will increase sales, production and employment so much that the concomitant decrease in unemployment also raises wage claims by workers and wage concessions by firms.
On the other hand, when effective demand for consumer goods is so large that production is restricted by capacity and not sales in the consumer industry, then a central bank can control inflation by controlling spending on consumption through its policy rate. The effects are
9
transmitted through three channels: 1) new borrowings; 2) interest payments for existing debt; and 3) asset prices and hence the wealth of households.
A sufficiently high policy rate will reduce spending by households through all of these channels. First, it may even, if sufficiently large, completely eliminate borrowings needed for spending by households on new homes or cars. Second, by increasing payments for existing debt (with variable interest rates) a higher policy rate will reduce disposable incomes for households with debt. (But it will also increase disposable income for households with savings.) Third, increasing interest rates tend to reduce prices for homes and securities and hence the wealth and also the expenditures of the owners of such assets. Note finally that increasing the policy rate will not only reduce the effective demand for consumer goods but also discourage investment in increasing capacity in the consumer industry.
Restrictions on the policy rate
In an open economy the possibility of free capital flows between countries may equalize interest rates between countries, assuming that investors move their capital from markets with low yields to markets with higher yields. As also emphasized by Blanchard (2009 p. 456), if interest rates are equalized by perfect capital mobility, then a change in interest rates
presupposes either coordination among central banks or leadership by one of them.
Leadership by the Federal Reserve means, more precisely, that the target federal funds rate determines the level of policy rates set by other central banks. Even if some deviation is possible, this is, of course, an important restriction on a country’s policy rate.
Interest rates as objectives
The interest rate, or more precisely the policy rate of a central bank, is usually not thought of as an objective but as an instrument of stabilization policy. However, the policy rate is an objective in the sense that it is a target for the interbank rate of reserves, which a central bank (CB) cannot achieve merely by announcing it. In fact the interbank rate of reserves can be very volatile in a market without intervention by a CB, or with a CB focused on stabilizing the growth of money supply, as shown by the fluctuations of the federal funds rate in the U.S.
10
until 2002.9 Thus, stabilization policy is in practice very much concerned with stabilizing the interbank rate of reserves, particularly in systems with a “corridor” for the interbank rate, implying not only a ceiling for the rate but also a floor.
However, the policy rate is usually thought of as an instrument for stabilizing inflation and output, through its effect on all other interest rates and hence indirectly on effective demand.
With the policy rate perceived as a tool, it is only its effect on inflation and output which should matter. But the variability of the policy rate also matters for firms and households, and the policy rate has since the 1970s been very variable in many countries. Even if a CB often announces not only a policy rate but also a plan for its development over the near future, the variability of the policy rate makes planning of investment more difficult, both investment in housing by households and investment in productive capacity and new technology by firms.
In fact, “by keeping interest rates stable, policymakers can insulate the real economy from disturbances that arise in the financial system”, as noted by Cecchetti (2008 p. 503).
Of course, the level of the interest rate also matters for economic agents (as measured by its average over a business cycle), even if they cannot agree on what the level should be: while borrowers prefer low rates, lenders – including pension funds – prefer high rates. A
complication is that preferences on the (nominal) interest rate depend on inflation, even if an inflation rate which is stable at about two per cent should make this problem a marginal one.
Perhaps there exists a “critical value” of the level of the policy rate, such that both lenders and borrowers can accept it as “fair”, while a deviation from it will be opposed by either borrowers or lenders. In any case, we cannot a priori exclude preferences on the interest rate which depend on other things than effects on inflation and output. However, postulating a target for the level of interest rates, the policy rate can no longer be thought of as an instrument for controlling inflation. And then a government has to stabilize the value of money by other means than the policy rate.
4. Employment
According to equation (1), a firm’s employment (N) in the market sector is determined by either its capacity (k) or its sales D(p) at the price (p) it sets,
9 See, for example, Cecchetti (2008 p. 432).
11
(7) N D p a
if D p
and otherwise N k ak ,where a denotes the firm’s labour productivity.
In practice there are, of course, some complications. Production does not adjust perfectly to sales unless sales precede production (production to orders). In markets where production precedes sales, as in most consumer markets, production will in general differ from sales, even if the change in inventories during a year may be negligible. Adjustment costs (costs of hiring and firing) will stabilize employment when sales are variable or hard to predict.
Aggregation from a single firm to its industry is straightforward. Employment in an industry is determined as the sum of employment in its firms at a common market price. (In a market with differentiated goods prices may differ somewhat but the price level must be the same.) Mark-ups are not necessarily set independently by every firm. For example, in a market with price leadership all firms but one will take the market price as given. Then the mark-up for a price taker is determined by the market price p set by the price leader and the direct cost c of the price taker, m
p c c
, while the price leader sets that market price which maximizes its individual profits.Focusing on wages, employment is always sensitive to wages if they are sufficiently high, so that production and employment are restricted by sales and not capacity. Then higher wages will by raising prices reduce sales, production and employment and the effect depends on the price elasticity of the demand for the industry’s products, but also on labour’s share in total variable costs. However, the basic reason why there is a negative relation between wages and employment at the microeconomic level, is that product demand is independent of the firms’
wages, an assumption which is not necessarily valid at the macroeconomic level.
The effect on aggregate employment of a general increase of wages depends on how higher wages add to product demand. While higher wages will not increase demand for firms producing export goods, they may increase expenditures on consumer goods. But higher wages also imply higher prices on consumer goods, so the net effect on production and employment in the consumer industry is a priori ambiguous.
While a higher wage level does not raise employment in the export industry, it will not necessarily reduce employment – with two exceptions. Higher wages in exporting firms will raise prices and reduce sales, production and employment in firms which are price makers in
12
global markets, provided their production is restricted by sales and not capacity. However, in exporting firms that are price takers in global markets, higher wages will reduce profits but not employment – unless wage increases are so large that some firms go bankrupt.
Aggregate employment depends on the relation between demand and capacity in every industry. In general, an industry’s regime – production restricted by sales or capacity – depends on the industry’s product cycle: innovation, growth, maturity or decline. A new industry may find it possible to sell all it can produce at very high prices, while a declining industry may have excess capacity at the prices it sets. It can even happen that an industry’s firms deliberately install some excess capacity in order to be able to meet temporary increases in demand without having to ration customers by either price or quantity (or delivery times).
If production is restricted by sales in an industry, employment is determined by sales and labour productivity, while sales are determined by spending at prices set by firms. While labour productivity depends on technology, real capital, organisation and the use of imports as inputs in all industries, the determinants of spending differ substantially between
industries. Thus, while wages are crucial for consumption, “animal spirits” may be decisive for investment and “international competitiveness” important for exports, while “deficit policy” may guide government spending.
The effect of recruitment problems on employment10
Employment can be reduced by recruitment problems. Most hires are made more or less instantaneously by firms, for example by recalling workers previously laid off or by offering jobs to spontaneous job applicants. In other cases, there is no existing pool of job applicants which a firm can turn to. Instead the firm has to attract job applicants by advertising its demand for personnel in newspapers or other media, by placing job orders with a public or private employment agency, or by contacting potential candidates directly. And then vacancies understood as recruitment processes arise.
Moreover, firms start recruiting in anticipation of future needs. If, for instance, a separation can be anticipated and a replacement made before the separation, then replacement is
instantaneous even if recruitment is not. But otherwise vacancies understood as unfilled jobs (unmet labour demand) exist from the day the employer wants the worker to start to the day
10 The rest of this section is based on Farm (2020c).
13
the worker starts. An unfilled job can consequently be interpreted as an unplanned dip in employment or an “empty chair” or “idle machine”.
Thus, firms create “vacancies” in one sense (recruitment processes) in order to avoid
“vacancies” in another sense (unfilled jobs). Because of this ambiguity, both in the economics literature and in everyday language, I will refer to vacancies as recruitment processes as job openings (or openings) and vacancies as unmet labour demand as unfilled jobs. The distinction is important, since it is only unfilled jobs which reduce employment by making the number of employees less than the number of jobs.
In the Swedish business survey on vacancies, produced by Statistics Sweden since 2000, unfilled jobs are defined more precisely as unoccupied job openings which are available immediately. This is a definition which excludes occupied job openings and job openings to be filled later, in the same way as the definition of unemployed workers in labour force surveys excludes job seekers with a job and job seekers without a job who cannot start work until later.
Moreover, in the Swedish survey unmet labour demand (unfilled jobs) is measured indirectly, as unmet labour supply (unemployed workers) is measured in labour force surveys, by a succession of questions. More precisely, unfilled jobs are defined as a subset of job openings obtained by eliminating first “occupied job openings” and then “unoccupied job openings which are unoccupied because no work is wanted by the employer until later”. And
“occupied job openings” exist when, during recruitment of new workers, the corresponding jobs are occupied by retiring workers or substitutes until replacements or permanent
personnel have been hired.
Now, according to the Swedish vacancy survey, which measures both job openings and unfilled jobs, about 40 per cent of job openings are also unfilled jobs, and the rate of unfilled jobs has averaged 0.6 per cent of employment in the private sector in Sweden since 2000.
Thus, the time it takes to recruit workers reduces employment by creating a gap between jobs and employment of about 0.6 per cent of employment on average in Sweden, varying from 0.3 per cent in 2009 to 1.0 – 1.2 per cent in 2016 – 2019.11
The matching function
11 See ”Konjunkturstatistik över vakanser” at www.scb.se.
14
If hires are instantaneous, there are no job vacancies as measured in vacancy surveys. But not all hires are instantaneous. Thus, job vacancies (time-consuming recruitment) occur and a large outward shift of the UV (or Beveridge) curve – relating unemployment (U) to vacancies (V) over time – has often been interpreted as a decline in search effectiveness with large effects on unemployment.12 The stocks of vacancies and unemployment have even been interpreted as inputs in a production process described by a matching function with the flow of hires as output.13 And the hiring rate is assumed to affect unemployment by affecting the probability of finding a job for unemployed.
In fact, hiring is much simpler than the concept of a matching function implies. Hires are not
“produced” by job openings and unemployment, as suggested by the matching function.
Hires are produced or, more precisely, initiated by decisions to replace separations or change employment. Some – perhaps most – of these hires are realized more or less instantaneously (e.g. by recalls of former employees), while other recruitments take some time. However, a firm controls its employment at every point in time unless unfilled jobs arise.
5. Unemployment
Historically, large increases in unemployment have usually been caused by large drops in employment caused by a financial crisis, like the Great Depression in the 1930s or the Global Financial Crisis 2007-2009 – or the crises in the Nordic countries in 1991-93. Moreover, it usually takes several years after a financial crisis for employment and unemployment to return to normal levels. Otherwise changes in unemployment are usually associated with changes in employment caused by changes in incomes, credit, preferences, and foreign trade, changes which are associated with economic growth and structural change, often discussed iterms of “business cycles”. But sometimes high unemployment can only be explained by special factors, like a restrictive economic policy or an increase in labour supply.
Unemployed people are usually defined more precisely as persons without employment looking for jobs. Not all people without employment are searching for work and they may not even be able or willing to undertake employment. Ability may be restricted by illness or
12 For example: “In a world where economists have little certain knowledge, the shift of the U/V curve provides us with vital clues to the sources of the rise in unemployment. Large shifts indicate that a major part of the rise is due to changed behaviour of workers and employers in the filling of vacancies.” (Layard et al. 1991 p. 220.)
13 See Petrongolo and Pissarides (2001) for a precise definition of the matching function and a survey of its role in equilibrium models of unemployment since the 1970s.
15
disabilities or other activities, like education or homework. Willingness may depend on other sources of income. In fact, defining and measuring “unemployed” as distinct from other people (in working age) without employment is an intricate problem.
Defining and measuring unemployment
In labour force surveys unemployed persons are usually defined as people without
employment who are “looking for jobs”, and they are measured by a question on job search during the last month. But “discouraged workers”, who no longer find job search meaningful, are also registered, suggesting a broader definition of unemployment. As emphasized, for example, by Blanchard (2009 p. 138), unemployment as defined in traditional labour force surveys is not necessarily the best estimate of the number of non-employed people available for work.
Information on non-employed who are “available for work”, in the sense that they are both qualified for employment and willing to accept a job offer, would be important information on the “labour reserve”. In any case, when unemployment (U) is small relative to
employment (N), and the labour force (L = N + U) is approximately constant, then the change in the unemployment rate ( u U L ) between years is approximately determined by the growth rate of employment ( dN N ), since
(8) ut ut1 dU L dL L
dN N
N L
.Unemployment and its flows
It is tempting to think of the stock of unemployment at a particular point in time as
determined by past inflows and outflows. It is also obvious that for an individual the causes of unemployment can fruitfully be divided into, first, causes of becoming unemployed and, second, causes of remaining unemployed. But what is true for an individual is not necessarily true for a group of individuals.
In fact, even if policy measures can reduce the risks of becoming or remaining unemployed for certain groups of people, they can only reduce total unemployment by raising total
employment or reducing the number of non-employed looking for jobs. The second approach
16
has sometimes been implemented by facilitating early retirement, but otherwise unemployment can only be reduced by increasing employment.
Of course, employment can be increased by labour market policy measures for unemployed people, like retraining, employment subsidies, or intensified job search – even without
increasing effective demand at the same time. However, in order to show effects on aggregate employment one has to show that these measures will raise employment either directly, by reducing matching problems so much that the number of unfilled jobs also is reduced, or indirectly, by reducing recruitment costs and hence product prices for some firms so much that sales and hence also production and employment are increased (without decreasing production and employment for other firms).
Note also that even if reducing the generosity of the unemployment insurance system may increase the search intensity of the unemployed and also reduce the duration of
unemployment for some groups, aggregate unemployment will only be reduced if the rate of unfilled jobs is reduced or, perhaps, if lower unemployment benefits reduce wages and lower wages increase employment.
As emphasized by Devine and Kiefer (1991 pp. 307-308), search theory implies that policy measures can affect unemployment only if they affect the flows into or out of unemployment:
In the Markovian framework we are accustomed to, it is these flow rates that determine the equilibrium or natural unemployment rate. Policies affecting the flow rates are likely to have long-term effects; those focused on the stock at a particular period in time will have only short- term effects. This notion has been crucial in the discussion of unemployment insurance policy and its effect on the natural rate of unemployment.
However, Devine and Kiefer also note that: “the intuition leading to this generalization is based on Markovian (or semi-Markovian) models, and this specification is rarely checked.”
In fact, unemployment is not a Markov process. It is the stock of unemployment that is a restriction on the flow rates, not the other way around.
Relating unemployment to its flow rates is certainly an illuminating decomposition of the stock of unemployment. It explains, in particular, the distribution of unemployment between flow and duration and hence also between short-term and long-term unemployment. But it does not explain total unemployment – not even in the medium run.
17 Two key issues
According to Romer (2019 pp. 520-522), there are two key issues in the theory of unemployment. The first concerns the determinants of “average unemployment over extended periods”, while the second concerns the cyclical behavior of the labour market.
“First, what are the underlying reasons that labor markets differ from Walrasian markets in ways that cause significant unemployment?” (Romer p. 520) According to Romer there are two broad approaches to modeling the labour market as non-Walrasian: a traditional one and a modern one.
According to the traditional approach, unemployment reflects some force that prevent wages from falling to the level that equates supply and demand. For example, according to efficiency wage theories, firms may choose not to lower wages to the level that equates supply and demand because there is a cost as well as a benefit to them to paying lower wages.
According to the modern approach, “the process of matching up workers and jobs occurs not through markets but through a complex process of search and matching. Some
unemployment is inevitable in such settings, and various disturbances can change the level of unemployment.” (Romer p. 521) Moreover, according to Romer (pp. 550-551):
When workers and jobs are highly heterogeneous, the labor market may have little resemblance to a Walrasian market. Rather than meeting in centralized markets where employment and wages are determined by the intersections of supply and demand curves, workers and firms meet in a decentralized, one-to-one fashion, and engage in a costly process of trying to match up idiosyncratic preferences, skills and needs. Since this process is not instantaneous, it results in some unemployment.
Thus, for search theory a Walrasian market is not even a useful “benchmark” in labour markets. This is an important development in economics. However, a Walrasian market is a misleading “benchmark” in all markets except markets for commodities or securities. In fact, we have seen that when production is determined by capacities or sales at prices set by firms, as in most markets, this has important consequences for employment and unemployment.
Finally, the cyclical behaviour of the labour market differs in important ways from the implications of a supply and demand model with flexible wages (Romer p. 520). While the model suggests that increasing employment should be associated with decreasing real wages, empirics show that “if there is any correlation between aggregate employment and real wages, it may be positive” (Hamermesh 1993 p. 337).
18
Now, labour demand is not, after all, a function of the real wage. It is a derived demand, dependent on production (Section 1). Moreover, the average real wage depends on characteristics of the consumer sector and relative wages in the other sectors (Section 2).
Thus, there is no simple relation between employment and real wages, suggesting a weak correlation, including the possibility that a boom is accompanied by higher real wages.
6. A macroeconomic model
This paper explains inflation and unemployment starting from updated models of price formation and labour demand. Inflation is always and everywhere a pricing phenomenon (excluding acceptable cases in some countries). Unemployment is always determined as a residual, as people in the labour force without employment. This applies to both the short run and the medium run.
Note that in eq. (10) aggregate demand is measured by nominal GDP, denoted Y, that is, the money spent on services produced by domestic firms,
(9) Y C I G X Z .
Here C denotes spending by households, I spending by firms on investment, G spending by the public sector on goods and services from the market sector, X exports, Z imports and X – Z spending (net) by the foreign sector on goods and services from domestic firms.
The development of aggregate demand Y deals with all of its components according to (14) and their determinants, including wages, taxes, transfers and saving for consumption, profits, debt, and “animal spirits” for investment, and taxes and debt for the public sector. The components of aggregate demand are also affected by monetary policy (the policy rate) and fiscal policy (taxes and transfers).
On the neutrality of money
In classical economics, the aggregate supply relation and the aggregate demand relation were two relations between real output (y) and the price level (P), which together determined the price level and output, as in a supply-and-demand model for a single market. As emphasized, for example, by Knoop (2010 pp. 32-36), the aggregate supply curve was a vertical line,
19
while the classical aggregate demand curve was downward sloping, implying that output was determined by supply (Say’s Law), while aggregate demand determined the price level.
Aggregate supply of output was determined indirectly by real factors in the labour market, since both the demand for labour and the supply of labour – and hence also employment in equilibrium – was a function of the real wage, while output was determined by employment, capital and technology. Aggregate demand was derived from the quantity theory of money,
(15) Py VM ,
which also is a theory of aggregate demand if both the supply of money (M) and the velocity of money (V) are constant. Thus, in classical models, changing the money supply can affect the price level, P VM y , but not output, suggesting that money is “neutral”.
Most modern models have the same basic message in the medium run (“money neutrality”
with output determined indirectly in the labour market) but allow for effects of monetary or fiscal policy on output and unemployment in the short run. In the medium run, economic policy can only affect the price level. More precisely, monetary stimulation of effective demand may increase production and employment in the short run but not in the medium run, since, it is argued, it will sooner or later be “neutralized”.
Now, Say’s law is applicable even in the short run if production is restricted by capacity in every industry. On the other hand, in the medium run production at full capacity due to high effective demand may stimulate investment in capacity and consequently increase
production. Normally, however, production is restricted by sales in some (perhaps most) industries with sales restricted by spending at prices set by firms. Increased spending will increase production and employment at the same time as it increases the stock of money if – and only if – it is financed by money created either by commercial banks through lending or by a public budget deficit financed by borrowing from commercial banks or the central bank.
Note, however, that the correlation between output and money will be broken if money creation finances purchases of existing houses or securities instead of current output. Note also that the relation between the stock of money and output is indirect. The point is not that
“money matters” but that “spending matters”. Increased spending will always increase aggregate output when production is restricted by sales unless, of course, increased spending in some industries somehow – e.g. through taxes or credit allocation – is related to decreased
20
spending in other industries. Changes in effective demand as measured by nominal GDP always have real effects – unless its industry is producing at full capacity.
7. Conclusions
Changes of the price level, as measured by a consumer price index, are determined by changes in wages, labour productivity, import prices and mark-ups in firms producing consumer goods – even in the long run – if production is restricted by sales at prices set by firms. Inflation is affected by money growth – even in the short run – if production is restricted by capacity at prices set by firms and spending on consumer goods is partly financed by borrowings from banks creating money when they lend (commercial banks).
Only in this case is too much money chasing too few goods.
Unemployment is every year, and not only in the short run, determined as a residual, as people in the labour force without employment. Employment is determined by production and labour productivity and – to a degree – by friction in the matching of workers and jobs, while production is determined by capacity or spending at prices set by firms.
21 References
Blanchard, Olivier 2009. Macroeconomics, fifth edition. Pearson Education International.
Blanchard, Olivier, Lars Calmfors, Harry Flam, John Hassler and Per Krusell 2015. Makroekonomi [Macroeconomics]. Liber.
Cecchetti, Stephen G. 2008. Money, Banking, and Financial Markets, second edition. McGraw Hill.
Devine, Theresa J. and Nicholas M. Kiefer 1991. Empirical Labor Economics. The Search Approach.
Oxford University Press.
Elvander, Nils 2002. “The New Swedish Regime for Collective Bargaining and Conflict Resolution.
A Comparative Perspective”. European Journal of Industrial Relations, 8:197-216.
Farm, Ante 2017a. “Pricing and price competition in consumer markets”. Journal of Economics, 120:119-133.
Farm, Ante 2017b. “Pricing and production in consumer markets where sales depend on production”.
Economics Letters, 154:17-19.
Farm, Ante 2020a. “Pricing in practice in consumer markets”. Journal of Post Keynesian Economics, 43:61-75.
Farm, Ante 2020b. ”Om lönebildning” [On wage formation]. Ekonomisk Debatt, 48, 3:37-48.
Farm, Ante 2020c. “Measuring the effect of matching problems on unemployment”. International Labour Review, 159, 2:243-258.
Farm, Ante 2020d. “Labor demand and product demand”. Journal of Post Keynesian Economics, forthcoming,https://doi.org/10.1080/01603477.2020.1794905.
Faxén, Karl-Olof, Clas-Erik Odhner and Roland Spånt 1989. Lönebildningen i 90-talets samhällsekonomi [Wage formation in the economy of the 1990s]. Rabén & Sjögren.
Hamermesh, Daniel S. 1993. Labor Demand. Princeton University Press.
Knoop, Todd A. 2010. Recessions and Depressions. Understanding Business Cycles, Second Edition.
Praeger.
Layard, Richard, Stephen Nickell and Richard Jackman 1991. Unemployment. Macroeconomic Performance and the Labour Market. Oxford University Press.
McLeay, Michael, Amar Radia, and Ryland Thomas 2014. “Money Creation in the modern economy”. Quarterly Bulletin, 2014 Q1:14-27, Bank of England.
Petrongolo, Barbara, and Cristopher A. Pissarides 2001. “Looking into the Black Box: A Survey of the Matching Function”. Journal of Economic Literature, 39:390-431.
Romer, David 2019. Advanced Macroeconomics, Fifth Edition. McGrawHill.
Wicksell, Knut 2010. Lectures on Political Economy, Volume II: Money. Routledge Revivals.
