Advertising Economics Jobs / Economist Jobs Help

<!– /* Font Definitions */ @font-face {font-family:Wingdings; panose-1:5 0 0 0 0 0 0 0 0 0; mso-font-charset:2; mso-generic-font-family:auto; mso-font-pitch:variable; mso-font-signature:0 268435456 0 0 -2147483648 0;} @font-face {font-family:”Lucida Console”; panose-1:2 11 6 9 4 5 4 2 2 4; mso-font-charset:0; mso-generic-font-family:modern; mso-font-pitch:fixed; mso-font-signature:-2147482993 6144 0 0 31 0;} /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:”"; margin:0in; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:”Times New Roman”; mso-fareast-font-family:”Times New Roman”; mso-fareast-language:EN-US;} p.MsoFootnoteText, li.MsoFootnoteText, div.MsoFootnoteText {mso-style-noshow:yes; margin:0in; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:”Times New Roman”; mso-fareast-font-family:”Times New Roman”; mso-fareast-language:EN-US;} span.MsoFootnoteReference {mso-style-noshow:yes; vertical-align:super;} @page Section1 {size:8.5in 11.0in; margin:1.0in 1.25in 1.0in 1.25in; mso-header-margin:.5in; mso-footer-margin:.5in; mso-paper-source:0;} div.Section1 {page:Section1;} /* List Definitions */ @list l0 {mso-list-id:1034228641; mso-list-type:hybrid; mso-list-template-ids:-1324421944 -1137020624 67698691 67698693 67698689 67698691 67698693 67698689 67698691 67698693;} @list l0:level1 {mso-level-start-at:0; mso-level-number-format:bullet; mso-level-text:-; mso-level-tab-stop:.5in; mso-level-number-position:left; text-indent:-.25in; font-family:Arial; mso-fareast-font-family:”Times New Roman”;} ol {margin-bottom:0in;} ul {margin-bottom:0in;} –>

Section II

Data collection

The data for inflation and unemployment values were colleted from the Economist Intelligence Unit (EIU) for each individual country. For inflation, the percentage of consumer price inflation (CPI) was taken, since this is the headline value of inflation that central banks tend to target. For unemployment, the total percentage unemployment rate was used, since it was the most commonly used and an aggregate of the unemployment statistics. Other studies[1] have suggested using capacity utilisation to monitor the output or unemployment gap, but since this study primarily concerns an investigation into the Phillip’s curve relationship, unemployment was chosen.

Due to the lack of more frequent data for the 1980’s, yearly values were used in Section III.A for comparing the experience of the 1980’s and 1990’s for each individual country. The original idea was to carry out a cross section for all the OECD countries in order to maximise the number of observations for the regression, but the wide variance of unemployment and inflation values between countries rendered the test inaccurate. Testing for an aggregate Phillips curve of highly industrialised countries is not yet possible yet, given the lack of convergence of real variables – particularly in the 1980’s. Therefore, despite the lack of observations, OLS regression tests were carried out for the individual countries, utilising each decade as a subsample. Given that data was available for the period 1980-2004 it was decided to split this period into two, so that rather than studying each decade, the subsamples chosen are in fact 1980-1992 and 1993-2004.

For Section III.B, since quarterly data from the EIU dataset was available for the 1990’s, this was used to increase the number of observations and accuracy of the regression. Data for the NAIRU estimates were collected from the OECD dataset. Since values were only available for the years 1990,1995,1999, and 2003, the values in between had to be interpolated for a smooth transition between the years with estimates. Although this reduces the validity of the test, it was pursued because the natural rate is believed to move gradually, and thus interpolation seemed a fair method of replicating this. For this section, since the unemployment gap was being measured by regressing Unemployment minus NAIRU (U-NAIRU) on Inflation, it was possible to carry out both an aggregate cross-section regression and an individual country regression. This is because using the unemployment gap accounts for differing positions of NAIRU values for each respective country.

The countries chosen for the analyses had to meet three criteria:

- They are in the OECD, and thus highly industrialised

- They had to have their NAIRU value estimated

- Countries with the majority of inflation values above 20% were ignored since this study is concerned with observing what occurs under low inflation values.

As a result, the countries analysed were: Australia, Austria, Belgium, Canada, Denmark, Finland, Germany, Greece, Italy, Ireland, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom (UK), and United States of America (US).

Data Description

Tables II.1 and II.2 represent a summary of the data used in this study, and are complemented by Figures A.1 and A.2 in the appendix which are time series line graphs showing the unemployment and inflation experiences of each country.

	UNEMPLOYMENT MEANS			UNEMPLOYMENT VARIANCES
	All years	1980-1992	1993-2004	All years	1980-1992	1993-2004
AUSTRALIA	7.68	7.81	7.54	2.22	2.37	2.22
AUSTRIA	4.12		4.12	0.13		0.13
BELGIUM	8.60	8.78	8.45	1.99	3.09	1.22
CANADA	9.02	9.49	8.52	2.59	2.77	2.08
DENMARK	8.36	9.96	7.70	6.35	1.06	7.20
FINLAND	8.55	5.43	11.93	17.70	4.66	9.56
FRANCE	9.93	9.21	10.71	2.09	1.57	1.57
GERMANY	8.95	7.88	10.10	2.79	2.48	0.57
GREECE	8.59	6.91	10.40	4.94	2.92	0.65
IRELAND	11.36	14.20	8.29	22.12	8.21	19.49
ITALY	9.45	8.74	10.04	1.50	0.63	1.50
JAPAN	3.22	2.41	4.10	1.24	0.08	0.99
NETHERLANDS	6.22	7.06	5.32	3.34	3.02	2.27
NEW ZEALAND	6.59	7.07	6.31	3.87	6.88	2.34
NORWAY	3.82	3.40	4.28	1.66	2.19	0.78
PORTUGAL	6.19	6.65	5.79	2.70	4.04	1.49
SPAIN	17.35	17.59	17.08	17.12	8.39	28.06
SWEDEN	4.41	2.70	6.26	5.09	1.00	2.82
SWITZERLAND	2.18	0.83	3.63	2.83	0.36	1.36
UNITED KINGDOM	7.97	9.54	6.66	5.11	2.77	3.39
UNITED STATES	6.25	7.13	5.31	2.15	1.90	0.74
Total Average	7.56	7.64	7.74	5.22	3.02	4.31

Table II.1

	INFLATION MEANS			INFLATION VARIANCES
	All years	1980-1992	1993-2004	All years	1980-1992	1993-2004
AUSTRALIA	4.83	7.12	2.54	10.81	9.16	2.00
AUSTRIA	2.76	3.56	1.97	2.37	2.83	0.76
BELGIUM	3.09	4.29	1.89	5.12	7.25	0.34
CANADA	3.71	5.57	1.85	8.30	9.28	0.55
DENMARK	3.52	5.04	1.99	7.40	10.11	0.29
FINLAND	3.46	5.68	1.43	7.20	4.62	0.79
FRANCE	3.70	5.78	1.62	11.95	15.24	0.30
GERMANY	2.27	2.77	1.77	2.62	3.91	1.03
GREECE	12.42	18.84	5.99	55.64	12.04	14.27
IRELAND	4.96	7.07	2.85	22.91	36.15	2.08
ITALY	6.10	9.20	3.00	22.88	25.59	1.30
JAPAN	1.13	2.13	0.12	2.20	1.72	0.68
NETHERLANDS	2.54	2.66	2.42	2.39	4.42	0.55
NEW ZEALAND	5.61	9.27	1.95	29.12	30.68	0.98
NORWAY	4.47	6.87	2.07	11.46	10.81	0.63
PORTUGAL	9.88	16.16	3.61	63.97	45.98	1.85
SPAIN	6.07	8.79	3.35	14.04	12.27	0.97
SWEDEN	4.43	7.31	1.55	13.42	8.09	1.84
SWITZERLAND	2.36	3.67	1.06	3.75	3.46	0.69
UNITED KINGDOM	3.63	5.57	1.69	7.07	6.21	0.35
UNITED STATES	3.53	4.56	2.51	3.44	4.61	0.30
Total Average	4.50	6.76	2.25	14.67	12.59	1.55

Table II.2

The data has been split into the subsamples representing the two periods studied in order to analyse average changes. Of notable interest is the lower inflation mean for the period 1993-2004, falling from 6.76 in 1980-1992 to 2.25. This can be largely attributed to the increasing devotion that policy makers have paid toward maintaining low and stable inflation, especially with many of these countries switching to inflation targeting and/or central bank independence in the 1990’s, starting with New Zealand in 1989. In addition to a significantly reduced mean, the mean variance of inflation has also dropped for these countries. This solidifies the proposition that lower average inflation is accompanied by lower inflation volatility.

There has been less movement in the unemployment data between the two periods, with aggregate average unemployment rising very little from 7.64 to 7.74. Whilst this aggregate suggests little movement, it is slightly misleading given that some countries experienced significant falls and others experienced significant increases of unemployment. Australia, Belgium, Portugal, New Zealand, Canada, Denmark, Ireland, Netherlands, UK and US have experienced a fall in average unemployment, with the latter six experiencing a fall of more than 1%. Even small changes of average unemployment should be considered important because these subsamples cover twelve years each, and thus the differences between the periods are unlikely to be of a cyclical nature, but rather a change of structural unemployment. The other countries suffered increases of average unemployment, particularly dramatic are the increases of at least 3% for Finland, Greece, Sweden and Switzerland. Little can be said of the changing variance of unemployment (rising from 3.02 to 4.31), except that 1993-2004 experienced a slightly larger fluctuation of unemployment rates.

<!– /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:”"; margin:0in; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:”Times New Roman”; mso-fareast-font-family:”Times New Roman”; mso-fareast-language:EN-US;} @page Section1 {size:8.5in 11.0in; margin:1.0in 1.25in 1.0in 1.25in; mso-header-margin:.5in; mso-footer-margin:.5in; mso-paper-source:0;} div.Section1 {page:Section1;} –>

Section I

Literature Review

Given that the topic of the project has to my knowledge not been explicitly explored what follows is a brief summary of literature that have contributed ideas and formed the base. The cohesive theory is that near rational expectations result in suboptimal unemployment equilibrium.

First, consider the phenomenon of high unemployment, with special reference to the post 1979 oil shock experience of Europe vis-à-vis the United States. As Figure I.1 shows these regions suffered a reversal of fortune, with the unemployment of the United States decreasing after the shock and undercutting a previously successful Europe, where unemployment continued to rise and plateau at a high level. Labour market institutions have received the thick of accusations, but as Nickell (1997, p.65) notes “…roughly speaking, the labour market institutions were the same. So how can the labour market institutions have anything to do with unemployment?”. Clearly other explanations are required.

Figure I.1; Source: OECD Economic Outlook

Second, various economists point to the negative externalities of high unemployment. Layard, Nickell and Jackman (1991) suggest that persistent high unemployment decreases the work and job search skills of the unemployed and that the employed closely guard their wage at the expense of expanding employment. Additionally Eisner (1995) notes that the rise in wages that can accompany a fall in unemployment may cause a substitution of capital for labour which would curb inflation. As a result, Stiglitz (1997) suggests that reducing unemployment can cause a reduction of the NAIRU via a positive feedback effect and thus the economy faces a moving target, which is argued empirically by Semmler and Zhang (2004). Considering that these effects take time the economy may experience a little inflation along the way. Galbraith (1997) highlights the asymmetrical problem this presents for policymakers by suggesting that if a higher NAIRU accompanies higher unemployment in response to a negative shock to the economy, the obsession with the NAIRU as a long term target and fear of inflation prevent a major positive shock or policy move from correcting this; thus the economy may become stuck at a higher unemployment equilibria.

Third, various studies including Blanchard (2003) suggest that the NAIRU may vary over time and can be affected by the inflation rate and monetary policy. Additionally, the presence of imperfect competition, information asymmetries and price rigidity may give rise to multiple equilibria. Bomfim and Diebold (1996) demonstrate that the combination of bounded rationality and strategic complementarity can lead to strong policy effects and output persistence, whereby rational agents anticipate the behaviour of suboptimal agents and strategically imitate. Cooper and John (1988) and Pagano (1990) suggest that such coordination failures lead to multiple unemployment equilibria and that well constructed and strong policy intervention may be necessary to prevent getting stuck at high unemployment.

Fourth, the previous points give cause for policy experimentation. This is typically frowned upon because of the theorised convex and risk-averse nature of the short run Phillips curve produces unfavourable rising inflation with falls of unemployment. However, CEA empirical studies cited in Stiglitz (1997) suggest that the best fit for the US economy was concave during the 1990’s, while separate studies by Eisner (1996) and Galbraith (1997) point to the lack of correlation between low unemployment and accelerating inflation. Therefore even risk-averse policy makers should find reason to experiment, considering the economy can always reverse course without cost. This study hopes to investigate this relationship for the European countries.

Fifth, Galbraith (1997) contests that the US experience shows little evidence for a vertical long run Phillips curve, as demonstrated by Figure I.2. Akerlof, Dickens and Perry (2000) postulate that the long run Phillips curve is not vertical but bowed inward and then forward bending, suggesting that there is an optimum positive inflation rate, which greases the wheels of the economy. Consequently overly strict inflation targeting may bring efficiency losses for the economy

Figure I.2; Source Galbraith (1997)

Sixth, research in the field of behavioural economics is substantial and growing, Akerlof (2002) cites many important studies in the field; potentially relevant findings for near rational expectations are listed below. Conlisk (1996) and Kahnemann et al. (1982) have shown that the physiological limits of human cognition exhibit what has come to be known as bounded rationality, which has proven important in economics. Systematic errors are evident in decision experiments, consumer behaviour and expectations, suggesting that actors are not rational. Interestingly, studies so far have concentrated on financial issues such as asset prices and inflation expectations have not yet been considered. Geanakoplos (1992) highlights the salience of common knowledge between agents in decision making when more than one agent is involved, an example could be policy makers and the public. Brian Arthur (1994) notes that the inductive reasoning process of humans results in a built-in hysteresis; because expectational models are tested by their previous accuracy of tracking the economy, a switch to a better model will occur only once the previous model has accumulated a record of failure.

Seventh, Carlstrom and Fuerst (2001) have provided an interesting critique of forward-looking inflation targets. Because “policy depends on expected inflation and expected inflation, in turn, depends on policy” decision making leads to an “infinite regress” whereby the public and central bank continue to affect each other. This leaves the economy vulnerable to self-fulfilling expectations, and sunspots can occur where extraneous information can be self fulfilled by changes in behaviour by either agent. The Economist (2005) intuitively picks up on this suggesting that because the ECB constantly gives the impression that policy needs to be tight to protect from the ever-present threat of inflation, the public believe nothing can be done to improve growth, presenting a vicious circle. Mismatches between the public, firms and policymakers need not be particularly complicated to cause disruption; Ball and Mankiw (2002) cite the role of the aspiration wage and how wage setters perniciously sustain real wage increases when productivity falls.

Eighth, Staiger, Stock and Watson (1996) point to the lack of precision when calculating the NAIRU (for example the 95% confidence interval in the US was 5.1% to 7.7% for 1990). Not only does this suggest caution when utilising the NAIRU as a guide for policy, but it also serves as caution for any subsequent results this project obtains.

<!– /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {mso-style-parent:”"; margin:0in; margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:”Times New Roman”; mso-fareast-font-family:”Times New Roman”; mso-fareast-language:EN-US;} @page Section1 {size:8.5in 11.0in; margin:1.0in 1.25in 1.0in 1.25in; mso-header-margin:.5in; mso-footer-margin:.5in; mso-paper-source:0;} div.Section1 {page:Section1;} –>

Abstract

The relationship between unemployment and inflation has become rather elusive from the 1990’s onwards. The results in this paper show that the experience in the 1990’s for OECD countries has been significantly different from that in the 1980’s, specifically that during the 1990’s the trade off between unemployment and inflation has been severely dampened. Consequently the level of the Natural Rate of Unemployment for these countries has become less clear since inflation appears to be less affected by unemployment movements, particularly in an economic climate dominated by inflation targeting. The implications of this trend will be discussed, and a co-ordination failure model is proposed which suggests that countries that maintain an overly strict inflation targeting policy may experience efficiency losses as a result of near-rational expectations.

Introduction

The driving impetus behind this paper has been the unemployment experience of the United States during the 1990’s, whereby the Federal Bank pursued an expansionary policy that took unemployment below what at the time was the perceived Non Accelerating Inflation Rate of Unemployment (NAIRU hereafter) without an increase of inflation, defying conventional wisdom. This experience served to underline what appears so far to be a changing trade-off between unemployment and inflation in the industrialized countries, with many countries exhibiting a far smaller sacrifice ratio in the 1990’s than previous decades. Whilst many of the OECD countries have made a switch to formal or informal inflation targeting during the 1990’s, which has enabled them to keep inflation low and anchored, many countries, particularly European, have been left with a considerable level of unemployment, as shown by Figures 1 and 2. With inflation now largely under control the implications for monetary policy, given what seems to be an increasingly weak short run Phillips curve (SRPC hereafter) relationship, have been relatively unexplored.

Figure 1; Source: OECD

Figure 2; Source: OECD

The first test, comprising Section III.A, will be to analyze whether the experience of the OECD countries in the 1990’s and early 2000’s has been significantly different from the experience in the 1980’s. If true, what does this result mean, is the data sufficient to believe that the SRPC relationship has now changed, and what are the causality issues that may be influencing the result? All of these questions and policy implications will form the discussion in Section III.C. Whilst other studies have touched on this topic, none have covered it with such a large range of countries, or to my knowledge such recent datasets.

Second, using NAIRU data for the 1990’s period Section III.B will attempt to examine this new dynamic in closer detail by comparing inflation changes to the unemployment gap, specifically the Unemployment rate minus the NAIRU level (U-NAIRU), with the aim of disseminating the recently elusive relationship. Utilising the unemployment gap data will hopefully allow the analysis to take into account the changing dynamics of an economy, and examine the SRPC with more precision. Interestingly, since these NAIRU values are considered to be medium to long term, the mere fact that they change suggests the lack of a single long run vertical Phillips curve. Whilst this paper does agree with the theory of the long run Phillips curve – the idea that there is a floor to the level of unemployment with stable inflation is plausible given bargaining power and efficiency wage considerations – the experience of the 1990’s onwards for many countries suggests that few countries have settled on or discovered their long run levels.

This leads to the third part of the paper, making up Section IV, which proposes that some of the countries that are suffering from high unemployment may be doing so unnecessarily as a result of a co-ordination failure between agents in the economy. Without discarding any previous explanations, this theory hopes to show that in light of tight inflation targeting an economy may miss out on opportunities for expansion due to asymmetries of information and near rational expectations. This will be followed by discussion and policy implications.

The paper begins with a review of the relevant literature and ideas that contributed to this project, forming Section II. An explanation of the data and methodology form Section III.