The Revival of the Long Run Philips Curve (Part 2)

Today I want to take a closer look at several of the high income OECD countries following yesterday’s post on the short run and potentially long run trade-off between unemployment and price stability. For this I obtained country statistics on the rate of unemployment as percentage of the civilian labour force, the annual percentage change of consumer prices and the NAIRU (Unemployment rate with non-accelerating inflation rate) from the OECD’s Economic Outlook No 91 in June 2012. All measures are available online from OECD Statistics. In particular, I am looking for the standard short run trade-off between unemployment and inflation (Philips curve) as well as a long run trade-off between the NAIRU and inflation below a certain threshold level close to absolute price stability or deflation.

OECD inflation unemployment trade off 1

OECD estimates for the US NAIRU are available from 1970 onwards. The NAIRU is relatively constant over the complete period and ranges from 5.4 to 6.4 percent. Overall, the US has recorded positive levels of inflation (excluding 2009) often above the 2 percent level except for 1986, 1998, 2002 and 2013. The short run trade-off between unemployment and inflation seems to hold on average after 1986 except for a period in the 1990s. There might also be evidence for a long run trade-off as the US inflation rate dropped to -0.4 percent in 2009 with a spike in unemployment and a small increase in the NAIRU by 0.3 percent. Overall though the data remains inconclusive because the US has experienced mostly inflation large enough to ensure wage flexibility.

OECD inflation unemployment trade off 2

Data for Canada is available from 1970 onwards. The country experienced almost a decade of disinflation after the spike at 12.5 percent in 1981. Lately, inflation seems to have fluctuated around 2 percent and prices were almost perfectly stable in the years 1994 (0.2) and 2009 (0.3). However, this was accompanied by a small rise of 0.5 percent and 0.3 percent in the NAIRU, respectively. So there might be a long run trade-off but in this case at a very low threshold of close to zero. The short run trade-off between unemployment and inflation seems to hold on average after 1989.

OECD inflation unemployment trade off 3

NAIRU estimates for France are available from 1970 onwards. After the two inflationary spikes in 1974 and 1980-81 inflation has come down to levels of around 2 percent recently. Overall the short run Philips curve seems to hold after the second inflationary spike. What is more, the NAIRU has more than doubled in line with the unemployment rate since the 1970s which hints at a short run and long run trade-off. There seems to be an increasingly lower inflation threshold of circa 1 or 2 percent at which the NAIRU responds. For example, I would interpret the period from 1992 to 2000 and 2009 to 2013 as such a long run trade-off where France achieved price stability at the cost of a rising NAIRU.

OECD inflation unemployment trade off 4

NAIRU estimates for Ireland are available after 1990 and inflation rates are available from 1977 onwards. Similar to France, Ireland experienced inflationary spikes around the same time but of a greater magnitude. Since then inflation has dropped significantly and exceeded 5 percent only in 2000. Inflation reached its all-time low of -4.5 percent during the GFC in 2009 and has recently recovered to levels close to price stability.  The period of disinflation in the 1980s and 1990s was accompanied by high unemployment exceeding 15 percent for several years. From 2000 to 2007 inflation and unemployment have been moderate with a large fall in the NAIRU to 7.5 percent. However, deflation triggered a sharp rise both in the unemployment rate and NAIRU in 2009 providing evidence for both an adverse short run and long run impact of deflation in Ireland. In case that there is a threshold it would probably be around a zero or 1 percent level.

OECD inflation unemployment trade off 45png

NAIRU estimates for Greece are available from 1995 onwards. After a long period of stagflation and breakdown of the Philips curve until circa 1990 inflation came down to levels below 4 percent while employment rose by 3.9 percent from 1990 to 1999. In the 2000s inflation remained relatively stable and unemployment came down again to 7.3 percent in 2008. Meanwhile the NAIRU remained constant. However, after 2008 the NAIRU rose significantly from 9.9 to 12.3 percent in 2013. During the same period inflation dropped to all-time lows of 1.2 percent in 2009 and -0.9 percent in 2013. Unemployment soared at more than 27 percent. Hence Greece provides some evidence for an inflation threshold of circa 2 percent in order for it to affect both short run and long run unemployment.

 OECD inflation unemployment trade off 6

Data on Italy is available from 1970 onwards. After the inflationary spikes of 1974-77 and 1981 inflation has dropped to levels below 4 percent since 1996. However, the short run Philips curve does not seem to hold well on average for the period before 2008. Italy’s NAIRU steadily increased until 1998 with even higher actual unemployment rates. Unemployment fell to 1970-levels again in 2007 (6.1%), but increased back to 1980/90s-levels in the aftermath of the GFC. Overall, Italy does not provide compelling evidence for a long run trade-off between unemployment and inflation based on OECD NAIRU estimates. This should be treated with caution though because the sharp rise in the unemployment rate after 2008 might actually translate into a rise in the NAIRU like in the case of Greece.

OECD inflation unemployment trade off 7

Data on Portugal’s NAIRU is available from 1980 onwards. Portugal was hit by three massive inflationary spikes in 1974 (15.3), 1977 (31), and 1984 (28.4). Thereafter inflation fell rapidly and has been below 4 percent since 1996 on average. The all-time low was in 2008 with deflation of 0.8 percent and in 2013 inflation again dropped to almost zero (perfect price stability). The short run Philips curve seems to hold on average after 1992. The NAIRU fell by 1.2 percent over the period from 1980 to 2000. Since 2001, however, it has risen by 5 percent to 11 percent in 2013 and unemployment has soared even more sitting at over 30 percent in 2013. During the same period inflation was relatively low. However, the period from 2001 to 2008 is comparable to the late 1990s which did not cause such a massive rise in the NAIRU. Overall, there might be an inflation threshold of around 2 percent at which both unemployment in the short run and long run soar.

OECD inflation unemployment trade off 8

The last country I want to look at today completes the Southern European panel of Greece, Italy and Portugal. Spain’s NAIRU is available from 1978 onwards. Starting at 4.8 percent it rose to 15.8 percent in 1995. After a short recovery until 2005 the NAIRU is now back at an even higher level of 16.5 percent. Similar to other countries inflation spiked in 1977 at 24.5 percent. After a prolonged period of disinflation it is now down to levels below 4 percent. In 2009, Spain actually recorded deflation for the first time and in 2013 inflation was again down to only 1.4 percent. Overall the Philips curve does not seem to hold very well for the time until 2007. Since 2008, however, both the NAIRU and the unemployment rate are rising sharply with inflation below 2 percent from 2008-10 and 2013 providing evidence for both a short run and long run trade off below a certain level. This finding is comparable to Italy. What is more, the inflation trend line looks a little bit like the NAIRU mirrored along a slowly decreasing line through the point of intersection of the NAIRU and the inflation trend which would be another hint for a long run trade off.

In sum, Canada, France, Greece, Portugal and Spain might actually provide evidence for a long run trade-off between unemployment and inflation below a certain inflation threshold due to impaired wage flexibility as discussed yesterday. A main limitation is that the data is relatively noisy and for example the US data is more or less inconclusive due to inflation on overage high enough to allow for wage flexibility. Italy, however, does not fit into the model.

Thanks for reading!

Jasse


OECD (2012), “OECD Economic Outlook No. 91 (Edition 2012/1)”, OECD Economic Outlook: Statistics and Projections (database).
DOI: http://dx.doi.org/10.1787/data-00606-en
(Accessed on 14 April 2016)

OECD (2016), “Labour Force Statistics: Summary tables”, OECD Employment and Labour Market Statistics (database).
DOI: http://dx.doi.org/10.1787/data-00286-en
(Accessed on 14 April 2016)

OECD (2016), “Prices: Consumer prices”, Main Economic Indicators (database).
DOI: http://dx.doi.org/10.1787/data-00047-en
(Accessed on 14 April 2016)

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s