A good way to think about the Phillips Curve is that it is a handy device for thinking about inflation and unemployment issues. It is like having a photograph of your mother-in-law with you. That photo is not your mother-in-law – it is just a piece of paper. It is handy because it helps remind you of her. But it is not the real thing – it is quite simple compared to the full story. The Phillips curve is similar to that photograph because it is a handy tool that reminds us about some key facts about inflation and unemployment that we already learned from the AS and AD analysis.
We have been drawing the Phillips Curve on a diagram with the Inflation rate on the vertical axis. The analysis is better explained if we make one important change to the drawing – let’s make the vertical axis be the CHANGE in the inflation rate. Notice that the diagrams below examine the relationship between the unemployment rate and the CHANGE in the inflation rate in a given year. Reading the diagram – any point above the horizontal axis is depicting a point of positive change in inflation. That means those points show a rising inflation rate –like when the rate goes from 1% to 2%. Any point below the horizontal axis is negative change in inflation rate – meaning that inflation is falling – like from 4% to 3%.
This alteration in the diagram helps us make a very important and specific point about the Phillips Curve – if the unemployment rate exactly equals NAIRU in a given year – then the inflation rate will not change in that year. That is, when the unemployment rate in a given year equals NAIRU, then there should be no change in the inflation rate. If the inflation rate was 5% then it will stay at 5%. If the inflation was 50%, then it will stay at 50%. Now we are ready to show how the Phillips Curve analysis is compatible with the AS and AD model developed in previous chapter.
The following graphs show the Phillips Curve for the USA from 1960 to 2009. The vertical axis charts the CHANGE in the inflation rate from the previous year to the next year. The horizontal axis is the unemployment rate for the current year.
- A rise in inflation is associated with a decline in unemployment
- A fall in inflation is associated with a rise in unemployment
Where does that slope come from? It comes from one thing – the short-run impact of changes in AD. Recall what happens when AD changes:
- An increase in AD raises output and prices – this generally translates into a reduction in unemployment and a rise in inflation
- A decrease in AD reduces output and prices – this generally translates into a rise in unemployment and a decline in inflation.
And now the same events are shown as movements along a given Phillip’s Curve:
Those of you who are still awake might ask, “does inflation really increase every time the price index rises?” And the answer would be no. In reality we know that many increases in price do not produce an increase in the inflation rate. For these discussions of the Phillips Curve here, we are analyzing changes in AD that are large enough and persistent enough that they would change the inflation rate in the way indicated.
To see this more clearly, begin with a situation in which today’s output is well below potential real GDP – implying that today’s unemployment rate is well above NAIRU. In such a situation workers have trouble finding jobs and firms have trouble selling their goods and services. Such a condition would be compatible with falling inflation. That is, in a dynamic economy prices might be rising but they might be rising at a slower rate compared to past years. See Zone A in the chart below. Now let AD increase somewhat – such an increase in AD could prevent the inflation rate from falling as much as it would have. Next consider an AD increase that was big enough to push output to a level that is above potential real GDP (and unemployment below NAIRU). In such a case demand is higher than potential output and labor is scarce. This situation is compatible with rising inflation. See Zone B in the chart below. The upshot is that as AD increases and as we move to the left on the above Phillips Curve, we go through a zone where prices are rising but as the inflation rate declines; and then finally into a zone where prices are rising and inflation is increasing.
Michael H. Moskow wrote about Phillips Curve tradeoffs in the Chicago Fed Letter (September 2006, “Reflections on monetary policy: Flexibility, transparency, and inflation.”) He is concerned about the Fed’s dual mandate to foster price stability AND maximum employment when faced with the Phillips Curve tradeoff. He begins by recognizing that the Fed has no real influence over the natural rate of unemployment or potential real output since those things are determined mostly by “demographics, productivity trends, and other factors.” He points out that if the economy is in Zone A of the above diagram, then monetary policy can improve both objectives. But once the economy moves into Zone B, then “there is a conflict in achieving both objectives. The inflation gap points to raising rates while the output gap suggests lowering them.” He explains that flexibility can be used to try to solve the problem as follows. Flexibility means that the central bank must balance the two deviations; therefore, it would take longer to close either gap….and the larger the policy dilemma, the longer it would take to close the gaps.” He doesn’t say, but we can infer that he means that the central bank will have to prioritize which gap to attack first and to do it measurably enough so as to not unnecessarily exacerbate the second gap. To retain this flexibility, it would likely be best not to have publicly announced explicit numerical targets for inflation or the unemployment rate.
Temporary Shifts in the Phillips Curve and Short-run AS Analysis: As the Phillips Curve charts in the first part of the note showed, the Phillips Curve in the 1980s was well above (or to the right) of the one in the 1970s. Then the one in the 1990s was to the left and down from the one in the 1980s. This Phillips Curve was shifting all over the map. Wazzup? These shifts were the result of AS factors – some temporary, some long-term. First, we discuss the temporary ones. That is, we discuss what might cause a temporary shift in the whole Phillips Curve.
That brings us back to the short-run AS and AD analysis. Recall that anything that causes costs to rise more than productivity would cause the AS curve to shift in an undesirable direction (left and up). A rise in inflationary expectations could do that. What is the result of such temporary AS shifts?
- A leftward shift (decrease) in AS causes output to fall and prices to rise – thus unemployment and inflation both rise.
- A rightward shift (increase) in AS causes output to rise and prices to fall – thus unemployment and inflation both fall.
- Viola – notice that this is not a movement along the curve – it is a shift of the whole Phillips Curve.
- Since the AS shift is temporary, then the Phillips Curve shift would also be temporary.
First the AD and AS model shows the undesirable shift in the AS curve:
Next, we show the same undesirable event as an outward shift of the Phillips Curve:
Durable Shifts in the Phillips Curve and Long-run AS analysis: These more persistent changes can be the result of either AD or AS changes – but to have persistent or durable impacts – these have to come from pretty fundamental and significant trends.
Above we reviewed the impact of a short-term AD change. Now we discuss what happens if the AD change is such that it keeps output above or below potential real GDP. Suppose AD increases and produces a short-run situation where:
Real GDP > Potential GDP
If this situation lasts for a while, then we believe that inflation will increase noticeably and inflationary expectations will begin to increase. This should make workers and other suppliers want higher wages and prices. Thus, if AD is such that it maintains high pressure on output for long enough, we should see a time period when business costs start rising faster than productivity. The result is that the AS curve shifts leftward and:
- Output falls and prices rise
- Unemployment and inflation both rise
- The Phillips Curve shifts right and up.
Unlike the previous shift in the Phillips Curve, this one would be permanent.
The conclusion here is that extended time periods of rapid AD growth can lead to undesirable shifts in the AS and Phillips Curves – and cause stagflation (the misery index climbs).
We can generalize this same point by saying that any AS trend that increases business costs faster than business productivity would have the same long-term effect of stagflation. One-time shocks would shift the Phillips Curve temporarily while long-term undesirable supply trends would lead to more permanent shifts.
What happens to NAIRU as a result of these Phillips Curve shifts? The fact that the Phillips Curve shifts to the right (and up) is telling you that the country has moved to a time period of generally higher unemployment rates. NAIRU increased.
The shifts of the U.S. Phillips Curve in the 1970s and then again in the 1980s, therefore has been explained by a combination of factors that ultimately impact the AS curve –
- too much government spending in the late 1960s (wars on poverty and Vietnam), rising inflationary expectations,
- increases in government regulations,
- an inexperienced baby boom generation entering the labor force.
- Social regulations to reduce discrimination in labor markets also gave more opportunities to minorities but contributed to the numbers of less-experienced workers.
The Phillips Curve began shifting in a more desirable direction sometime in the late 1980s. Why the turnaround? The answer must come from one or two directions – either sustained reductions in AD that resulted in lower inflationary expectations or from supply events and supply policies that raised productivity growth relative to business costs. These benign events would shift the AS curve rightward causing output to rise and prices to fall. This would be translated into both lower unemployment and inflation. This would be shown by a desirable shift (inward) of the Phillips Curve. This would imply also a lower NAIRU. This underscores why Michael Moskow in the article cited above concluded, “…..their actual approach to policy must be aimed at keeping inflation expectations anchored at a low level. They see this as a prerequisite to achieving maximum sustainable growth over the long-run.”
What might have contributed to the desirable shifts in the Phillips curve in the late 1980s?
- The residual impacts of very tight monetary and fiscal policy with the resulting decrease in inflationary expectations in the 1980s
- Impacts stemming from the supply-side policies of President Reagan
- Reductions in oil prices
- Global competition and other factors that might have decreased inflationary expectations
- Increased labor force experiences of the baby boom generation and minority workers
- The IT revolution and its impacts on business productivity
Janet Yellen writing in the FRBSF Economic Letter (“Monetary Policy in a Global Environment,” June 2, 2006) concedes that globalization might have impacted NAIRU for many reasons but she concludes that there is no real evidence to support this conjecture. This is an important issue for the Fed if lower import prices, lower wages, or other aspects of globalization have caused US NAIRU to decline, then the Fed could administer a looser monetary policy without fear of higher inflation. But if NAIRU is not lower, such a policy could have very undesirable consequences.
Industry application – Can I use the Phillips Curve or not?
You decide that you want to forecast and analyze inflation. You also read these articles about breakdowns of the Phillips Curve. (See the articles by Kliesen and Lansing cited below.) But don’t fret. Economists are a contentious lot and you have to know how to read between the lines. What these two economists are saying really is that you can’t JUST use the SLOPE (OR TRADEOFF) of the Phillips Curve to predict inflation. They underscore how things like changes in NAIRU, or AS shocks, or inflationary expectations need to be added to any inflation forecast. Of course, changes in NAIRU, shifts in the Phillips Curve, AS shocks and changes in inflationary expectations are part of the Phillips Curve analysis. If they are part of the AS and AD analysis, then they are part of the Phillips analysis. The upshot of this section is that we may never forecast inflation and unemployment perfectly, but hopefully by knowing as much as possible about short-term and long-term AD and AS shocks and trends, you may do a little better at this process.
Kevin Kliesen,The NAIRU: Tailor-Made for the Fed (5 pages),
The Regional Economist (Federal Reserve Bank of St. Louis) October 1999 (5 pages)
Kevin J. Lansing, Can the Phillips Curve Help forecast Inflation,
FRBSF Economic Letter, October 4, 2002 (3 pages)
Unemployment and Inflation Policy
Most countries want to have low unemployment rates, probably for two reasons.
- First, there is personal loss of income and self-esteem when a person is unemployed for too long.
- Second, from a macroeconomic standpoint it is wasteful to have people not producing when they could be.
As a result, most countries have policies designed to reach a good unemployment rate. We can think of this policy issue as having two dimensions:
(1) eliminating the output gap so as to bring the actual unemployment rate down to NAIRU and (2) reducing NAIRU.
Most of the unemployment policies we see in the real world are attempts to reduce the output gap – that is, they are expansionary monetary or fiscal policies designed to increase AD, output, and employment. History is full of examples. Pick-up today’s business paper and there should be at least one story about one country that is trying to reduce its output gap so as to reduce the unemployment rate. If not, today was a bad day. Try again tomorrow!
Common sense tells us that if the output gap is large enough, these policies have a good chance of working without causing higher inflation. It is only when the output gap starts to get smaller (real GDP gets closer to potential GDP) that these policies raise the possibility of higher inflation. When they do, the Phillips Curve tradeoff becomes a real policy dilemma – should we reduce the unemployment rate at the risk of higher inflation? So inflation worry becomes a constraint on unemployment policy. At such a juncture, policy makers have to decide if the benefit of lower unemployment is worth the cost of the higher inflation. This is illustrated in the below chart. Policymakers start at Point A, desiring to move to lower unemployment, yet risk moving to Point B or Point C. The policymaker sacrifices inflation for unemployment.
But of course, the reverse is true as well. If Macro Public Enemy #1 is inflation and a country attacks this problem by using monetary and/or fiscal policy to contract AD, then again we find ourselves on the trade-off curve. An AD policy to lower inflation would mean lower output and employment – a higher unemployment rate. People sometimes use the term “sacrifice ratio” to refer to this tradeoff. The ratio calculates the number of unemployment points you have to give up in order to gain an inflation point. So if you want to reduce the inflation rate from 3% to 2% and you expect the unemployment rate to go from 4% to 6%, then the sacrifice ratio would be 2. The policy maker at this point needs to evaluate if the benefit in inflation reduction is worth the cost of the higher unemployment. The unemployment rate in this case becomes the constraint on attempts to reduce inflation.
The slope of the Phillips curve – the so-called tradeoff – creates a real stalemate for macro policy:
- Worry about inflation constrains unemployment policy
- Worry about unemployment constrains inflation policy
Industry application – What should I forecast for Inflation and Fed Policy in mid-2006?
Page 11 of the Financial Times on June 8, 2006 offered several articles about inflation and central bank policy. The lead article was titled, “Is it back to the 1970s? Why inflation is again a spectre”. With headline inflation rising in the US and Europe many people were betting that the Fed and the ECB would raise interest rates as a clear signal to defeat rising inflation expectations (and the onset of stagflation). In 2001 and thereafter, much of the world was focused on exiting a recession. The policies worked and the world economy pulled out of recession and grew. By 2006 growth had resumed, output was near or above potential, oil and other commodity prices were rising, and it was clear that inflation was again public enemy #1. But was it? A clear case of uncertainty was evident. The uncertainty differed by region. How soft was the economy in 2006 in the US, Europe, and Japan? With respect to the US most people were willing to declare inflation as the main evil. But not everyone agreed. While inflation was rising in the US, many private forecasters saw the inflation rate falling toward the end of the year and into 2007. Would/should the Fed fight inflation and risk a tradeoff? But in Europe and Japan where the economy was not quite so strong the consensus seemed to be that the sacrifice ratio was too high. Fighting inflation too hard could mean a central bank-induced recession.
Is there a way to break these constraints? So what’s a good country to do? That brings us to part (2) above. Using AD policy to correct an output gap is not the only approach to unemployment and inflation problems. A second option is AS policy – or NAIRU policy. These options focus on productivity and costs. They focus on competition (see the note on AS for more information about AS policies). The upshot is since they do not work via changing AD (or moving along a given Phillips Curve), they end up without any trade-off at all. Increasing competition and productivity while reducing business costs has the effect of increasing output, and reducing inflation AND unemployment. The sacrifice ratio is zero. This is illustrated in the chart below.
As we stated in the note on AS, the politics of AS policy are difficult and may create a barrier to its implementation. But the truth is that many countries and their leaders have gone down this road and will continue to do so. Most developing or transforming nations are heavily invested in AS policy – though we might not name it as such. Most of these countries are moving away from communist or heavily state-run economic systems towards ones with more competition, freer markets, deregulated firms, and liberalized prices. These are policies designed to build strong AS sectors. Many of these countries have lowered their trade barriers and allowed foreign ownership of their assets? Why? Because they need more and better investment and they hope to import it from other countries. The new investment and imported technologies and the increased competition strengthens the AS sector.
Kenneth Rogoff (“The myth of how central banks slayed the hydra of inflation”, Financial Times, August 30, 2006, page 11) underscores how supply side factors make it possible to increase output growth while reducing inflation. He acknowledges that central bankers deserve some credit for bringing inflation down but they cannot take credit for avoiding the Phillips Curve tradeoff, “Instead one should think of the modern era of rapidly expanding trade and technological progress as providing a spectacularly favorable milieu for monetary policy…Rather than face the usual historical tradeoff, central banks have let citizens have their cake and eat it.” He worries that globalization will reverse itself. People might not understand that without globalization, central bankers have no way to deliver lower inflation without sacrificing economic growth.
International Application: Additional Evidence of AS Policy from Europe
Germany in 2003 and 2004 was in the throes of a debate about how to best change their social policy so as to make their labor markets more flexible and competitive. Agenda 2010 was an overall reform program designed to promote innovation and faster long-term economic growth and lower unemployment. Germany’s Agenda 2010 fit in perfectly with the European-wide Lisbon Strategy. The following words were taken from the site of the European Commission http://europa.eu.int/comm/lisbon_strategy/intro_en.html :
“The Lisbon Strategy is a commitment to bring about economic, social and environmental renewal in the EU. In March 2000, the European Council in Lisbon set out a ten-year strategy to make the EU the world’s most dynamic and competitive economy. Under the strategy, a stronger economy will drive job creation alongside social and environmental policies that ensure sustainable development and social inclusion.”
The following quote is from a 2004 report of the European Commission on how to strengthen their commitment to increasing employment. It makes the AS approach very clear as it relates to the Lisbon Strategy, http://europa.eu.int/eur-lex/pri/en/lip/latest/doc/2004/com2004_0239en01.doc
“If Europe is to meet its employment objectives in 2010 and increase its competitiveness and growth potential in the global economy, a trend break will be needed: both employment and productivity growth must accelerate strongly. A narrow approach to labour market reforms will not suffice. Sound macro economic policies are necessary to secure confidence and stability. Structural reforms are needed in the goods, services and capital markets to support competitiveness and job creation. Progress on all fronts of the Lisbon agenda, as completed by the environmental dimension at Göteborg, notably in terms of research and innovation, education and training, the development of the employment potential of environmental policies, in particular in the environmental goods and services sector, and the reform of social protection systems including pension arrangements, must go hand in hand. Policies in these areas, including the recent Growth Initiative, should boost business investment both in human and physical capital, and create better conditions for job creation and productivity growth by strengthening Europe’s capacity to manage change.”
Difference of Opinion
Inflation and unemployment are at the core of many important macro variables – like interest rates, wages, investment spending, and many others. They are also part and parcel of forecasts about the overall well-being of the economy. Your own personal and company forecasts will depend on many things. What AD and AS shocks have been impacting and threaten to continue to impact the economy? Will these shocks lead to tradeoffs of inflation versus unemployment? What problems would be associated with the change in inflation? What problems would be associated with the change in unemployment? How will households and business firms be impacted by these changes? Will the national policy makers perceive the new inflation or unemployment to be a big enough problem to require addressing with policy? Which problem is public enemy # 1? Will policy makers trade-off one problem for the other? Will policy makers find means to avoid a tradeoff? How will these policies impact your own decisions? I know that it is a lot of questions to ask but we had a lot of extra questions marks and didn’t know what else to do with them. These questions illustrate the kinds of things business planners should understand as they make their decisions in a world of uncertainty.