Tải bản đầy đủ - 0 (trang)
Annex I. The Phillips curve in Japan: estimating the relationship between output and inflation at low inflation rates

Annex I. The Phillips curve in Japan: estimating the relationship between output and inflation at low inflation rates

Tải bản đầy đủ - 0trang

OECD Economic Surveys: Japan



204



Table A.1.



Co-efficients of the demand pressure indicator in standard Phillips curves

Core inflation (corrected for VAT hikes) is the dependent variable1



Capacity utilisation in manufacturing sector

Output gap

Change in output gap

Business conditions index4

Production capacity index4

Unemployment gap5

Change in the unemployment gap



3



Co-efficient2



t-statistics



Sacrifice ratio



0.06

0.04

0.12

0.02

0.04

0.06

0.34



3.14

2.17

1.73

1.61

1.65

1.77

2.30



1.52

2.30

4.62

2.61

1.47



1.

2.

3.

4.

5.



Estimation period is quarterly data from 1975 to 2002.

Variables of demand pressure indicators are re-scaled so that the coefficients are comparable.

Capacity utilisation in manufacturing is inverted so that its coefficient becomes positive.

From the Bank of Japan’s Tankan Survey.

Unemployment gap is defined as the difference between the unemployment rate and the NAIRU. This parameter

is inverted so that its coefficient becomes positive.

Source: OECD estimates.



Estimation results

In all of the estimated equations, the indicator of demand pressure is significant with the

expected sign (Table A.1). For the unemployment and the output gap, both the level and the

first difference were significant determinants of inflation, signalling the presence of speed

limit effects in the economy. The equations pass standard in-sample diagnostic tests. One

common way of assessing the relevance of the Phillips curve is to check the implied sacrifice

ratio, which is measured as the cumulative output gap that is required to permanently

decrease the inflation rate by one percentage point. The equation that includes the output

gap leads to a sacrifice ratio of between 2.3 and 2.6. The sacrifice ratio is somewhat lower for

the specification using the unemployment gap and capacity utilisation (around 1.5) and

higher when the business index is included (around 4.5).

Does the slope of the Phillips curve become flatter at low rates of inflation?

The next step checks whether the slope of the Phillips curve changes when inflation is

low. Phillips curves are re-estimated, allowing the coefficient associated with demand

pressure indicators to differ depending on whether inflation is above or below certain

thresholds. Two inflation thresholds were tested: 0 per cent and ½ per cent (quarter-onquarter, non-annualised growth rates). The estimations show several interesting results. First,

in most cases, the coefficient associated with the demand indicator is halved when inflation

falls below the threshold and the demand indicator looses significance (Table A.2). This

corresponds to an increase (in absolute terms) of the sacrifice ratio, meaning that the slope

of the short-run Phillips curve is flatter at low levels of inflation. This result is true for the two

examined thresholds. Second, the coefficient associated with the demand indicator is

statistically different whether inflation is above or below ½ per cent as indicated by a Wald

test (Table A.3). These results hold independently of the indicator used and whether or not

inflation is corrected. By contrast, a threshold of zero does not imply significant change in the

relationship except for the equation making use of capacity utilisation. This lack of evidence

might stem from the lower number of observations associated with negative inflation rates,

which renders the outcome of the test very sensitive to every single observation. The



© OECD 2004



Annex I



Table A.2.



205



Asymmetry in the relationship between the output gap and inflation

Core inflation (corrected for VAT hikes) is the dependent variable

Inflation

Co-efficient

(π)

threshold



t-statistic



Inflation

(π)

Co-efficient t-statistics

threshold



Capacity utilisation in

manufacturing sector



π<0

π>0



–0.10

0.08



–1.25

3.74



π<½

π>½



0.01

0.09



0.23

4.30



Output gap



π<0

π>0



0.02

0.04



0.17

2.17



π<½

π>½



0.01

0.06



0.23

2.51



Change in the output gap



π<0

π>0



0.02

0.14



0.09

1.95



π<½

π>½



–0.02

0.18



–0.16

2.29



Business conditions index1



π<0

π>0



0.01

0.02



0.43

1.70



π<½

π>½



0.00

0.07



0.01

3.33



Production capacity index1



π<0

π>0



0.03

0.05



0.75

1.76



π<½

π>½



0.02

0.10



0.62

2.94



Unemployment gap



π<0

π>0



0.02

0.12



0.45

2.52



π<½

π>½



0.02

0.18



0.58

3.36



Change in the unemployment gap



π<0

π>0



0.25

0.42



0.84

2.43



π<½

π>½



0.10

0.81



0.55

3.70



1. From the Bank of Japan’s Tankan Survey.

Source: OECD estimates.



existence of a sequential break is also tested in 1990 and 1998. The result of the reestimation allowing for different demand indicator coefficients before and after 1990

and 1998 suggests that there is a break in 1990 but not in 1998.

Does a low level of inflation rate or low variance of inflation matter?

As discussed in Box 1, the slope of the Phillips curve can become flatter as a result of

price rigidity, due to less frequent adjustments, or nominal wage rigidity. While nominal

rigidity in wages implies a non-linearity in the Phillips curve only at inflation rates near zero,

the possibility of less frequency in price adjustments suggests a stabilisation of inflation that

can change the slope of the Phillips curve at any rate of inflation. In order to check this

hypothesis, the coefficient on demand indicators is tested for asymmetry relative to the

variation in the rate of inflation. The results of the re-estimation of the Phillips curve show

that its slope becomes steeper if the absolute value of the variance of inflation changes

exceeds a certain threshold. A Wald test also confirms the asymmetry, though the evidence

is a bit weaker than in the case of the level of inflation. On the other hand, a test on the

evolution of the coefficient on demand indicators over time indicates that there appears to

be breaks at the beginning of both the 1980s and the 1990s. The econometric tests, together

with the observation of the development of inflation over time, imply that the flattening of

the Phillips curve in the 1980s can be explained by a lower variance in the inflation rate

compared with that in the 1970s, while the further flattening in the slope of Phillips curve in

the 1990s is likely to be associated with the low level of the inflation rate.



© OECD 2004



OECD Economic Surveys: Japan



206



Table A.3. Wald tests1 for asymmetry and breaks

Core inflation2 (corrected for VAT hikes) is the dependent variable

A. Asymmetry



Capacity utilisation in manufacturing sector

Output gap

Business conditions index3

Production capacity index3

Unemployment gap

B. Sequential breaks

Capacity utilisation in manufacturing sector

Output gap

Business conditions index3

Production capacity index3

Unemployment gap



If inflation2 exceeds ½%

dummy = 1



0.00*

0.02*

0.00*

0.01*

0.00*

Break in 1990



0.16

0.06*

0.02*

0.01*

0.01*



If inflation2 exceeds 0%

dummy = 1



0.04*

0.55

0.55

0.51

0.17

Break in 1998



0.01*

0.27

0.14

0.12

0.03*



1. The Wald tests check whether the co-efficients associated with the demand indicator differ depending on whether

they are above or below the inflation threshold. At a probability above 10 per cent, it cannot be concluded that the

co-efficients are significantly different. An asterix indicates significance at the 10 per cent level.

2. The rate shown is the quarter-on-quarter, non-annualised growth rate.

3. From the Bank of Japan’s Tankan Survey.

Source: OECD estimates.



© OECD 2004



Annex II



207



Annex II



Assessment of the government’s structural reform programmes

This annex reviews the government’s reform programmes in detail so as to supplement

the overall assessment in this Survey. The subject of the review includes the measures

announced in the Outline of Basic Policy for Macroeconomic Management and Structural Reform, which

was released in 2001, 2002 and 2003, and the 2002 October package. The focus, though, is on

more recent measures (see Annex I in the 2002 Economic Survey for a fuller assessment of

earlier measures).

What are the aims of structural reform?

Since the announcement of the first Outline in June 2001, the government has been

actively implementing structural reform. The goals are to promote economic growth (“no

growth without reforms”) and to support devolution of power from the government to private

agents and from central to local government. Consequently, reform should contribute to

smaller and more efficient government, while improving the allocation of resources. To

achieve these goals, the reform programme includes major initiatives in the following areas:

– Deregulation.

– Financial-sector reform (non-performing loans, rehabilitation, securities markets).

– Improved human capital and a more flexible labour market.

– Stronger competition, more openness and entrepreneurship.

– Social security reform.

– Changes in the budget process and better public management (including privatisation).

– Devolution of power from central to local government.

How are the reform measures assessed?

The reform programmes are assessed from the following three perspectives:

Policy design. Policies are rated on the extent to which their design maximises their

effectiveness, while minimising the associated costs that have negative side effects on other

policy goals. The rating ranges arbitrarily from 0 to 3.

Stage of implementation. This simply compares what has been done relative to the

announced programme. A rating of zero is given for reforms still in the discussion stage, 1 for

having a concrete plan, 2 for a preparatory stage of necessary legislation, and the maximum 3

for the passage of legislation. Partial implementation is rated 1 or 2 depending on the

degree.



© OECD 2004



OECD Economic Surveys: Japan



208



Effectiveness. This assesses to what degree reform measures are compatible with the

original policy goal. Since planned measures may change considerably during the process of

negotiation, implementation is not necessarily regarded as proof of successful reform. The

rating ranges from 0 for plans and legislation that are not compatible with original goals (all

reforms at a discussion stage are automatically rated zero) to the maximum 3 for those that

broadly meet original objectives.

To what extent has progress been made in each area of reform?

Deregulation

In the past few years, the government’s Council for Regulatory Reform (CRR) continued

to focus on promoting the entry of business into publicly regulated sectors, including

medical services, education, agriculture and employment services. However, the success in

these areas has been limited so far. The difficulty in opening up these markets on a nationwide basis, together with the aim of promoting devolution, led to the idea of establishing

“Special Zones for Structural Reform”, which allows local governments to ease specific regulations

in order to promote local business. As of August 2003, 164 plans for such zones, including

deregulation in public services, had already been approved. Another important focus of the

Council is deregulation in planning in urban areas, which has improved the efficiency of land

use in those areas. The overall initiative for promoting deregulation is currently led by the

CRR, which consists of representatives of business and academia supported by a

government secretariat. However, in the absence of strong political commitment, its power

has proven to be very weak, especially in dealing with politically sensitive areas. Another

drawback is that the scope of the CRR’s activities does not cover deregulation of network

industries. Instead, regulatory issues in these industries are discussed by their own

committees and the progress tends to be very slow, especially in energy.

Table A.4. Deregulation

Policy

design



Implementation Effectiveness



Average



Establish “special zones”



3



3



2



2.7



Allow entry into publicly regulated services



2



1



1



1.3



Ease planning and zoning in urban areas



3



3



3



3.0



Strengthen the framework for deregulation



1



2



1



1.3



2.3



2.3



1.8



2.1



Average



Financial-sector reform (non-performing loans, rehabilitation, securities markets)

The government is promoting the rehabilitation of the financial system in line with the

programme announced in October 2002 (see the postscript in the 2002 Economic Survey and

Chapter II). The government’s plan, which aims at halving the ratio of non-performing loans

to total loans by major banks by the end of FY 2004, calls for ensuring adequate provisioning

of non-performing loans, reinforcing banks’ capital base and strengthening the governance

of banks to improve their performance. Since last October, progress has been made in a

number of areas. To improve the assessment of loan quality, the Financial Services Agency

(FSA) has continued its special inspections of large borrowers, while major banks have



© OECD 2004



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Annex I. The Phillips curve in Japan: estimating the relationship between output and inflation at low inflation rates

Tải bản đầy đủ ngay(0 tr)

×