ArticlePDF Available

Estimating the Natural Rate of Unemployment in Hong Kong

Authors:

Abstract and Figures

This paper uses unobserved components analysis to estimate the natural rate of unemployment in Hong Kong. We assume that the natural rate follows a random walk and is a determinant of the Beveridge curve. We find that the natural rate has increased since the 1990s but nevertheless lies clearly below the actual rate of unemployment. This indicates that unemployment is likely to decrease in the near future. for useful discussions.
Content may be subject to copyright.
Estimating the Natural Rate of
Unemployment in Hong Kong
Petra Gerlach-Kristen
Hong Kong Institute of Economics and Business Strategy
March 10, 2004
Abstract
This paper uses unobserved components analysis to estimate the natural rate of
unemployment in Hong Kong. We assume that the natural rate follows a random
walk and is a determinant of the Beveridge curve. We nd that the natural rate
has increased since the 1990s but nevertheless lies clearly below the actual rate of
unemployment. This indicates that unemployment is likely to decrease in the near
future.
HIEBS, University of Hong Kong, 919 KK Leung Building, Pokfulam Road, Hong Kong, ger-
lach@hku.hk. I thank Stefan Gerlach and Alan Siu for useful discussions.
1
1 Introduction
The Hong Kong economy has in recent years been exposed to a series of large negative
shocks. In 1998, the Asian crisis caused a decline of around ve percent in real GDP.
While the economy subsequently strengthened considerably in 1999 and 2000, the rst
recovery came to an end with the collapse of the global IT bubble and the terrorist attacks
of September 11, 2001. Real GDP declined by one percent in the last quarter of 2001,
but rebounded in 2002. In the spring of 2003, however, SARS stopped a second recovery,
causing another fall in real GDP of roughly half a percent. At the time of writing, the
economy once more seems to rebound.
Figure 1: Time plot of utand vt
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
1985 1990 1995 2000
unemployment rate (%)
vacancy rate (%)
Note: Sample period 1981:4 to 2003:3
Figure 1 shows the rate of unemployment and the vacancy rate over the period 1981:4
to 2003:3.1Unemployment increased over much of the 1990s and up to 2003, while the
vacancy rate declined. In 2003 the vacancy rate was roughly two percentage points below
1The data are from the database of the Census and Statistics Department and have been seasonally
adjusted.
2
its level in 1981, whereas unemployment had risen by four percent. This observation
suggeststhatthenaturalrateofunemployment—therateofunemploymentwhichwould
be observed if the economy were in equilibrium — might have increased. If so, we would
not expect unemployment to fall in the coming years to levels as low as in the 1980s in
spite of the economic recovery. The aim of this paper is to provide an estimate of the
natural rate of unemployment.
Our study is closely related to the Groenewold and Tang [3] who use a structural vector
autoregressive model of the rate of unemployment and real output to estimate over the
period 1982 to 2000 natural unemployment in Hong Kong (as well as in Korea, Singapore
and Taiwan). They focus on the impact of the Asian crisis and present an estimate of the
natural rate of unemployment that increases only slightly in 1998 and thus suggests that
the increase in unemployment at the time mainly was due to transitory factors.
Besides using a longer data sample, our paper complements and extends this analysis
in two important ways. First, while Groenewold and Tang provide estimates of the natural
rate of unemployment using an non-structural time series approach, the results presented
here rely on the Beveridge curve for the theoretical framework. The fact that two such
dierent approaches yield similar estimates of the evolution over time of the natural rate
of unemployment in Hong Kong provides some grounds for believing the results. Second,
our estimation technique allows us to quantify the uncertainty attached to the estimate of
the natural rate of unemployment. We nd that the condence bands are rather narrow,
which is further evidence suggesting that our estimate of the natural rate is reliable.
The rest of the paper is structured as follows. Section 2 discusses the concept of the
Beveridge curve. Section 3 shows how to use this theoretical framework to estimate the
naturalrateofunemployment. Sincethenaturalrateisavariablewhichisnotmeasured
directly, we apply the unobserved components estimation technique. Section 4 presents
the path of natural unemployment and documents that the estimate seems reasonable in
that it has high explanatory power in a simple Phillips curve and an equation for wage
ination in Hong Kong. Section 5 concludes.
3
2 The Beveridge curve
The relationship between unemployment and vacancies is commonly studied by means of
the Beveridge curve.2This curve relates the unemployment rate utto the vacancy rate vt.
Figure 2 shows the scatter plot for Hong Kong. The data points describe a hyperbola, so
that the functional form
utvα
t=ct(1)
captures the link between utand vtrather well (α=1would give rise to a rectangular
hyperbola; the larger α, the more curvature has the Beveridge curve). The variable ct
determines how far the hyperbola is from the origin. As illustrated in Figure 3, the higher
ct,thelargerisutfor any given vt.
Figure 2: Scatter plot of utand vt
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
0 1 2 3 4 5
v
u
Note: Sample period 1981:4 to 2003:3
2For an introduction on the Beveridge curve see Blanchard and Diamond [1]. See Shimer [5] for a
recent discussion.
4
Figure 3: The Beveridge curve
v
u(I)
u(II)
increasing c
Note: Rate of unemployment given vtfor dierent ct.
The natural rate of employment u
tis that level of unemployment which results if the
vacancy rate is at its mean, which we denote by v. Using equation (1), we dene
u
tct
(v)α.
If we divide both sides of equation (1) by (v)α, the Beveridge curve can be written as
ut³vt
v´α
=u
t
or alternatively
ut¡vt
v¢α
u
t
=1.
Taking logarithms we obtain
ln hut³vt
v´αiln(u
t)=0.(2)
Expression (2) can be thought of as an equilibrium condition that holds on average over
time but which can be disturbed by temporary shocks. We dene deviations from equi-
librium as
xt=lnhut³vt
v´αiln(u
t)+ex
t(3)
5
and assume that ex
tis white noise.
3 Estimation
We use the time series process of the actual rate of unemployment to estimate the path of
the natural rate. In particular, we assume that the rate of unemployment adjusts so as to
oset past deviations from the Beveridge curve. Allowing, moreover, for autocorrelation
in the changes of unemployment, we consider
ln(ut)=βln(ut1)γxt1+eu
t,(4)
where eu
tis white noise.3Equation (4) states that the rate of unemployment tends to rise
when it increased last period and if xt1waslow. Notethatweuseln(ut)rather than ut
to account for the fact that the unemployment rate cannot turn negative. Replacing xt1
with equation (3), we then estimate
ln(ut)=βln(ut1)γnln hut1³vt1
v´αiln(u
t1)o+et,(5)
where et=eu
tγex
tis distributed as N(0,σ2
e). Equation (5) thus implies that unem-
ployment tends to decrease if either ut1or vt1were high relative to the natural rate
of unemployment. In terms of Kalman ltering, equation (5) constitutes an observation
equation.4
The state equation is given by
ln(u
t)=ln(u
t1)+e
t,(6)
where e
tN(0,σ2
). Thus, we assume that the logarithm of the natural rate of unem-
ployment follows a random walk. We estimate the parameters of equations (5) and (6)
using Kalman ltering. Table 1 shows the regression output.
We nd that αis below unity, which implies that the Beveridge curve in Hong Kong
appears to have less curvature than a rectangular hyperbola.5The AR coecient βis
3We exclude the constant since it was insignicant in preliminary estimations.
4For a discussion of Kalman ltering, see e.g. Harvey [4].
5However, a Wald test does not reject the hypothesis that αequals unity (p-value =0.17).
6
Table 1: Unemployment model
α0.728∗∗
(0.201)
β0.229
(0.102)
γ0.403∗∗
(0.123)
σ2
e0.009
σ2
0.005
log likelihood 50.328
Note: Maximum likelihood estimates, 1982:2-2003:3. Standard errors in
parentheses. /∗∗ denotes signicance at the ve / one percent level.
estimated as 0.23 and the adjustment coecient γas 0.40. Correspondingly, the median
lag, which is given by ln(0.5)/ln(1 γ), equals 1.34. This indicates that, if condition
(2)wasnotmetinthelastperiod,theunemploymentratemovessuchthathalfofthe
misalignment is corrected after roughly four months. The estimate of the natural rate of
unemployment at the end of the sample is 2.99%. With 95% probability, u
tlay in the
third quarter of 2003 between 2.11% and 4.25%.
Figure4showstheestimateoftheBeveridgecurveattheendofthesampleperiod.
The actual data for 2003:3 lie within the 95% condence interval. However, the vacancy
rate of 0.59% is much below its equilibrium value, which is estimated to be 1.90%, while
utis with 8.30% above the estimate of u
t=2.99%.
4 The natural rate of unemployment
Next we construct an estimate of the natural rate of unemployment. Figure 5 shows the
smoothed u
ttogether with the actual rate of unemployment. Natural unemployment is
estimated as around 3.5 percent in the rst half of the 1980s. It then declines to two
7
Figure 4: Beveridge curve for 2003:3
0
4
8
12
16
20
0
4
8
12
16
20
0.5 1.0 1.5 2.0 2.5 3.0
v
estimated Beveridge curve
95% confidence band
actual data 2003:3
equilibrium
u
Note: Based on estimates in Table 1.
percent at the beginning of the 1990s. Thereafter, it rises over time and approaches three
percent in 2003.
Actual unemployment was slightly above trend from 1983 to 1986, but fell below the
estimated natural rate in 1987.6During the economic expansion at the beginning of the
1990s, utremained signicantly below u
t, but rose sharply during the Asian crisis. From
then onwards, unemployment in Hong Kong seems to have clearly exceeded its natural
rate. It thus appears likely that in the absence of further negative shocks, unemployment
in Hong Kong will return to a level of around three percent in the near future.
As noted in the introduction, our estimate of the natural rate is closely related to that
presented in Groenewold and Tang [3]. Their estimate also suggests that unemployment
was above its natural level during most of the 1980s, below it from the end of the 1980s
6For a detailed discussion of economic developments in Hong Kong over the period 1982 to 2000, the
interested reader is referred to Groenewold and Tang [3].
8
Figure 5: Actual and natural unemployment
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
1985 1990 1995 2000
u
u*
95% confidence band
Note: Sample period 1982:1 to 2003:3
to the onset of the Asian crisis and again above it from 1998 onwards. In contrast to our
estimate, the path of u
tin their paper displays slightly more time variation and is a little
higher throughout the sample.
As a last step of the analysis, we study the plausibility of the estimated natural rate
of unemployment. In particular, we assess whether the deviation of actual unemployment
from its natural rate is useful in explaining CPI and nominal wage inationinHong
Kong.7A plausible estimate of u
tshould make the unemployment gap, dened as u
tut,
impact both on πtand wt.
Figure 6 plots the unemployment gap, annual CPI ination and the output gap yt,
which is estimated using a Hodrick-Prescott lter. A clear link between the unemployment
gap and πtis evident. Using the data in the gure, we estimate two simple Phillips curves
for Hong Kong. We rst t
πt=a+bπt1+cyt+εt.(7)
7See Genberg and Pauwels [2] for a study on price formation in Hong Kong.
9
To account for potential simultaneity, we estimate all equations with GMM, using the
rst and second lag of the right-hand side variables as instruments. Table 2 holds the
regression output and shows that we do not reject the hypothesis that bequals unity,
which suggests that ination returns only slowly to equilibrium after a shock. Moreover,
the constant ais insignicantly dierent from zero, implying that the long-run mean of
ination is zero.
Figure 6: Phillips curve data
-10
-5
0
5
10
-10
-5
0
5
10
1985 1990 1995 2000
unemployment gap
output gap
CPI inflation
Note: Sample period 1982:1 to 2003:3
We next include the unemployment gap and estimate
πt=a+bπt1+cyt+d(u
tut)+εt.(8)
The impact of the unemployment gap is estimated as 0.58. Thus, when unemployment is
one percentage point below its natural level, CPI ination tends to increase by roughly
half a percentage point. The autoregressive coecient on ination is signicantly smaller
than unity, suggesting a comparatively fast return of ination to equilibrium. More im-
portantly, ais signicantly larger than zero. The long-run mean of ination is estimated
as a/(1 b)=6.75%. Thus, once the unemployment gap starts to close, we expect to
10
Table 2: Assessment of the plausibility of u
t
Phillips curves Wage equation
Equation (7) (8) (9)
a-0.142
(0.128)
1.148∗∗∗
(0.288)
1.050∗∗∗
(0.520)
b0.982∗∗∗
(0.020)
0.830∗∗∗
(0.042)
0.866∗∗∗
(0.060)
c0.263∗∗∗
(0.053)
0.061
(0.036)
d0.581∗∗∗
(0.190)
0.359∗∗
(0.169)
R20.959 0.970 0.970
Note: GMM estimates, instruments are two lags of the right-hand side variables. Sample
period 1982:4-2003:1 for equations (7) and (8) and 1983:3-2001:3 for equation (9). Standard
errors in parentheses. /∗∗ /∗∗∗ denotes signicance at the ten / ve / one percent level.
observe the emergence of a positive rate of CPI ination in Hong Kong.
To mo d el wage i nation, we consider the annual increase in nominal wages and the
unemployment gap. Figure 7 plots the data and suggests a close relationship between the
two variables. This is conrmed by a regression of the form
wt=a+bwt1+d(u
tut)+εt.(9)
The last column of Table 2 shows that the impact of the unemployment gap is signicant.8
If unemployment is low relative to its natural level, nominal wages tend to increase.
8Including CPI ination in equation (9) did not yield a signicant coecient.
11
Figure 7: Wage equation data
-8
-4
0
4
8
12
16
-8
-4
0
4
8
12
16
1985 1990 1995 2000
unemployment gap
nominal wage inflation
Note: Sample period 1982:1 to 2003:3
5Conclusions
We use the Beveridge curve as analytical framework to estimate the natural rate of unem-
ployment in Hong Kong. We nd that the natural rate declined in the 1980s and increased
from the beginning of the 1990s onwards. For the third quarter of 2003, when our data
end, the natural rate is estimated as three percent and thus lies clearly below the actual
unemployment rate of 8.3 percent. This suggests that unemployment is likely to decrease
in the near future.
To assess the plausibility of our estimate of the natural rate of unemployment, we t
equations for CPI and wage ination. We nd that the unemployment gap, dened as
the dierence between the natural and the actual rate of unemployment, is an important
explanatory variable.
12
References
[1] Blanchard, Olivier Jean and Peter Diamond (1989), The Beveridge curve, Brookings
papers on economic activity 1989:1, 1-60.
[2] Genberg, Hans and Laurent Pauwels (2003), Price formation and deation in Hong
Kong, mimeo, Graduate Institute of International Studies Geneva.
[3] Groenewold, Nicolaas and Sam Hak Kan Tang (2004), The Asian nancial crisis and
the natural rate of unemployment: Estimates from a structural VAR for the newly
industrializing economies of Asia, Pacic Economic Review 9(1), 45-64.
[4] Harvey, Andrew C. (1989), Forecasting, structural time series models and the Kalman
lter, Cambridge University Press, Cambridge.
[5] Shimer, Robert (2003), The cyclical behavior of equilibrium unemployment and va-
cancies: Evidence and theory, NBER working paper 9536.
13
... Table 5 reports our maximum likelihood parameter estimates and our estimates of the standard errors for these parameters based on the Hessian formula. 12 All parameters are precisely estimated. 13 Our estimate of the labor adjustment cost parameter is equal to 730. ...
Article
On average between 1981 and 2007, output per capita in Hong Kong grew at an annual rate of 3.2 percent. But economic performance over the period was far from even. Between 1987 and 1997, output per capita averaged 9.7 percent above its long run growth path. In 1997, the feast ended and the famine began. Over the next 11 years, output per capita averaged 5.2 percent below its long run path. This paper attempts to explain this big swing. We use two types of analysis. First, we use a nine-variable Bayesian Vector Autoregression and summarize the results as a set of stylized facts. Chief among these is the finding that economic activity in Hong Kong is highly sensitive to shocks of foreign origin. Second, we set our and estimate a dynamic, general equilibrium model which is able to account, at least qualitatively, for our stylized facts. We use the model to infer the relative importance of different shocks in accounting for feast and famine in Hong Kong.
Article
Full-text available
Concern and anxiety over the rising level of unemployment in Nigeria, especially among the youths, have mounted over the last decade. This concern will be misplaced, however, if we do not know the country's "natural" rate of unemployment. This paper attempts to identify Nigeria's "natural" unemployment rate and its evolution over time using the framework of the expectation-augmented Phillips relation and based on estimated parameters by the International monetary Fund (IMF) and historical data from the IMF and the National Bureau of Statistics (NBS). The results show that the rate may have risen significantly over time. This has implications for our understanding of the Nigerian labour market and the unemployment situation, the role and limit of macroeconomic policies, and the prospects of long-term growth in the country. While it is too early to make categorical statements on the reasons for the sharp increases in the "natural" unemployment rate, the paper draws attention to the possible influences of a strong social insurance around family networks (in particular, remittance inflows) and income from illicit "businesses" on the willingness to participate in the labour market at existing wage levels, and the likely impact of price increases arising from the unregulated nature of the country's product market. The paper concludes with the need for a more rigorous study of the labour market and the dynamics of change over the past twenty to thirty years.
Article
Full-text available
On average between 1981 and 2007, output per capita in Hong Kong grew at an annual rate of 3.2 percent. But economic performance over the period was far from even. Between 1987 and 1997, output per capita averaged 9.7 percent above its long run growth path. In 1997, the feast ended and the famine began. Over the next 11 years, output per capita averaged 5.2 percent below its long run path. To explain these big swings, we set out and estimate a dynamic, general equilibrium model. We show that the model tracks the data well and use it to infer the relative impor-tance of di¤erent shocks in accounting for feast and famine in Hong Kong. Our chief …nding is that disadvantageous shocks to the risk premium asso-ciated with foreign borrowing and world demand for Hong Kong exports were important causes of the transition from feast to famine.
Article
Prior to the late 1990s, low unemployment was a standard feature of macroeconomic life in Hong Kong. Between 1985 and 1997, the unemployment rate averaged 2.5 percent. But the picture changed dramatically thereafter with the unemployment rate rising to 6.2 percent by 1999 and remaining above 5 percent through 2005. What caused the large and sustained increase? This paper provides some answers with an analysis based on a dynamic, general equilibrium model of a small, open economy in which wage bargaining occurs. The model is calibrated using Hong Kong data for 1985 to 2005 and the calibrated model is analyzed in two ways. First, a set of comparative statics exercises investigates whether the natural rate of unemployment increased. Second, a dynamic analysis investigates whether the observed path of the unemployment rate might have been a temporary, although sustained, response to shocks. I conclude that the data favor the latter explanation.
Article
Full-text available
Over the past thirty years, macroeconomists thinking about aggregate labor market dynamics have organized their thought around two relations, the Phillips curve and the Beveridge curve. The Beveridge curve, the relation between unemployment and vacancies, has very much played second fiddle. We think that emphasis is wrong. The Beveridge relation comes conceptually first and contains essential information about the functioning of the labor market and the shocks that affect it. Labor markets in the United States are characterized by huge gross flows. Close to seven million workers move either into or out of employment every month. While that movement could be consistent with workers reallocating themselves across a given set of jobs, recent evidence by Steve Davis and John Haltiwanger suggests that these flows are associated with high rates of job creating and job destruction. Using a measure of job turnover, defined as the sum of employment increases in new or expanding establishments and employment decreases in shrinking or dying establishments, Davis and Haltiwanger find that during 1979-83, a period of shrinking employment, job turnover in manufacturing averaged some 10 percent per quarter. From a macroeconomic viewpoint, the labor market is highly effective in matching workers and jobs, yet those flows are so large that they imply the coexistence of unfilled jobs and unemployed workers. Examination of the joint movement of unemployment and vacancies can tell us a great deal about the effectiveness of the matching process, as well as about the nature of shocks affecting the labor market. In this paper, we first develop a conceptual frame in which to think about gross flows, about the matching process, and about the effects of shocks on unemployment and vacancies. We then turn to the empirical evidence, using data for the postwar United States. We focus first on the matching process, estimating the "matching function," the aggregate relation between unemployment, vacancies, and new hires. We then interpret the Beveridge relation. More precisely, we look at the joint behavior of unemployment, employment, and vacancies, and infer from it the sources and the dynamic effects of the shocks that have affected the labor market over the past 35 years.
Article
Full-text available
In this book, Andrew Harvey sets out to provide a unified and comprehensive theory of structural time series models. Unlike the traditional ARIMA models, structural time series models consist explicitly of unobserved components, such as trends and seasonals, which have a direct interpretation. As a result the model selection methodology associated with structural models is much closer to econometric methodology. The link with econometrics is made even closer by the natural way in which the models can be extended to include explanatory variables and to cope with multivariate time series. From the technical point of view, state space models and the Kalman filter play a key role in the statistical treatment of structural time series models. The book includes a detailed treatment of the Kalman filter. This technique was originally developed in control engineering, but is becoming increasingly important in fields such as economics and operations research. This book is concerned primarily with modelling economic and social time series, and with addressing the special problems which the treatment of such series poses. The properties of the models and the methodological techniques used to select them are illustrated with various applications. These range from the modellling of trends and cycles in US macroeconomic time series to to an evaluation of the effects of seat belt legislation in the UK.
Article
Full-text available
The turmoil that has rocked Asian foreign exchange and equity markets over the past eight months is the third major currency crisis of the 1990s. Its predecessors were the crisis in the European Monetary System in 1992-93, and the Mexican peso crisis of 1994-95.
Article
Full-text available
We use a structural vector autoregressive model to estimate the natural rate of unemployment for Hong Kong, Korea, Singapore and Taiwan for the period 1982-2000. Our results show that the dramatic rise in the unemployment rate observed in Hong Kong and Korea was mainly the result of demand shocks rather than structural changes, while in Singapore the unemployment rate rise reflected almost entirely a rise in the natural rate. Taiwan's natural rate has been relatively stable. We offer explanations for these different results in terms of the different economic characteristics (particularly labor market features) of the four countries. Copyright 2004 Blackwell Publishing Ltd
Article
This paper argues that the textbook search and matching model cannot generate the observed business-cycle-frequency fluctuations in unemployment and job vacancies in response to shocks of a plausible magnitude. In the United States, the standard deviation of the vacancy-unemployment ratio is almost 20 times as large as the standard deviation of average labor productivity, while the search model predicts that the two variables should have nearly the same volatility. A shock that changes average labor productivity primarily alters the present value of wages, generating only a small movement along a downward-sloping Beveridge curve (unemployment-vacancy locus). A shock to the separation rate generates a counterfactually positive correlation between unemployment and vacancies. In both cases, the model exhibits virtually no propagation.
Price formation and deflation in Hong Kong, mimeo, Graduate Institute of International Studies Geneva
  • Hans Genberg
  • Laurent Pauwels
Genberg, Hans and Laurent Pauwels (2003), Price formation and deflation in Hong Kong, mimeo, Graduate Institute of International Studies Geneva.