This Selected Issues paper of the United Kingdom analyzes the official projections of public pension spending and risks, as well as the strategy to increase private pension provision. It provides a comparison of stylized facts regarding business cycle developments in the three economies, and an analysis of how these cyclical differences reflect the way monetary policy changes impact the three economies. It analyzes the policy on Economic and Monetary Union membership; and also the interest rate changes in the United Kingdom and the United States compared with that in the euro area.

Abstract

This Selected Issues paper of the United Kingdom analyzes the official projections of public pension spending and risks, as well as the strategy to increase private pension provision. It provides a comparison of stylized facts regarding business cycle developments in the three economies, and an analysis of how these cyclical differences reflect the way monetary policy changes impact the three economies. It analyzes the policy on Economic and Monetary Union membership; and also the interest rate changes in the United Kingdom and the United States compared with that in the euro area.

II. The Interest Rate Sensitivity of UK Demand How Different Is It From the Euro Area and the United States?1

A. Introduction

1. Given its relatively greater openness, one would expect the UK economy to have been more vulnerable to the recent global slowdown. Yet, in response to this slowdown, the Bank of England cut its policy rate by n o more than the monetary authorities in the euro area and the United States and in fact maintained rates at higher levels. In the event, the UK economy weathered the downturn relatively well, with a more resilient growth performance than the other two economies. A possible explanation of these developments is that the UK economy is more responsive to interest rates.

uA02fig01

Monetary Policy Stimulus: Base Rates

(In percent)

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

2. The interest rate sensitivity of the UK economy was a central issue in the recent UK decision regarding whether to join the European Economic and Monetary Union (EMU).2 The UK Treasury’s background studies provide an extensive comparison of the UK economy vis-à-vis that of the euro area. The June 2003 assessment of the five tests draws the conclusion that cyclical and structural differences between the UK and euro area economies are the key risk factors posed by EMU entry in the near term. In part, these differences are attributable to relatively greater sensitivity of the UK economy, especially consumption, to changes in interest rates. One important source of this greater sensitivity is the influence on economic activity of the housing sector.

3. This paper attempts to gauge the interest sensitivity of UK aggregate demand and its key components relative to both the euro area and the United States. It is organized as follows. Section B provides a comparison of stylized facts regarding business cycle developments in the three economies. It suggests that the cyclical properties of the three are broadly similar, with a few notable exceptions, especially that consumption in the United Kingdom has been consistently more volatile relative to output than in the other two economies. Section C presents an analysis of how these cyclical differences reflect the way monetary policy changes impact the three economies. It concludes that, while the overall interest sensitivity of output is broadly similar, consumption is more sensitive to interest rate changes in both the United Kingdom and the United States than in the euro area. Section D analyzes consumption in the United Kingdom and the key role played by the housing sector. The final section summarizes the paper’s main findings.

B. Properties of the UK, Euro Area, and US Business Cycles

4. GDP growth in the United Kingdom over the period 1980-2003 was stronger on average than in the euro area, although it was more volatile. This result, however, reflects the performance of the UK economy in the 1980s and early 1990s. After the early 1990s, the United Kingdom continued to grow faster than the euro area, but the volatility of output declined to around the same level as in the euro area. This development partly owes to the strengthening of macroeconomic policy frameworks in the United Kingdom as well as structural factors, including the flexibility of labor, product, and financial markets which have been reformed over the past. In contrast, UK growth has been lower than that of the United States over the whole period, although it has been less volatile.

United Kingdom: Comparison of GDP Growth Rates 1/

article image

Annualized quarter-on-quarter growth rates.

5. The key features of business cycles in the United Kingdom, the euro area, and the United States are similar, but there are also important differences (Table 1).3 Cross-correlations of detrended GDP with its major components indicate that in all three countries consumption and investment are highly pro-cyclical. Investment also is more volatile than consumption. Government consumption in all three economies is counter-cyclical, while short-term interest rates are positively correlated with GDP at short leads and lags. Significant differences across the three economies are:

Table 1.

United Kingdom: Cross-Correlations with GDP (1980:1-2003:1)

article image

For period 1987:1-2003:1 due to data limitations.

  • Consumption tends to be more positively correlated with GDP in the United Kingdom and the United States than in the euro area;

  • Consumption in the United Kingdom is more volatile (both in absolute terms and relative to the standard deviation of GDP) than in either the euro area or the United States; and

  • Investment tends to be more highly correlated with GDP in the euro area than in the United Kingdom, and this correlation tends to be higher in both of these economies than in the United States.

Moreover, correlations between detrended GDP in the three economies indicate that UK business cycles have been more correlated with those in the United States than with those in the euro area. However, this relationship has changed over time; since 1993, the UK cycle has become more correlated with the euro area than with the United States.

Comparison of Business Cycles (1980:1-2003:1) 1/

article image

Sample moments were computed from detrended series using the Hodrick and Prescott filter. Sources: ONS, ECB, BEA and Fund staff estimates.

C. Comparison of Interest Rate Sensitivity

6. The differences in business cycles in the UK, euro area, and US economies in part reflect how they respond to monetary policy changes. Economic theory suggests that monetary policy has no long-run impact on output, which is determined largely by real variables. However, in the short and medium run, it can affect output primarily owing to price and wage rigidities in the economy. To identify the sensitivity of output to interest rate changes, this paper uses vector autoregressions (VARs), which have become a standard technique in the economic literature.4 Most of the studies on the euro area have focused on identifying compatibility and differences between individual countries. Peersman and Smets (2001) apply a VAR to “synthetic” euro area data5 to focus on the area-wide responses to interest rate changes. The authors show that, overall, the impacts of monetary policy shocks in the euro area and the United States are broadly similar. Bean, Larsen, and Nikolov (2002) extend the comparison to the United Kingdom and find that the impulse responses to a monetary shock in the United Kingdom are qualitatively similar to the euro area and the United States. However, the validity of this conclusion is open to some question because the specification of the VAR used for the United Kingdom, which included house prices as an endogenous variable, is different than that used in Peersman and Smets (2001).

7. To allow comparability, a stylized VAR model for the United Kingdom, euro area, and United States was estimated for the period 1980:1 to 2003:1 using a similar specification for all three economies. The model for each economy includes four variables in levels—real GDP, consumer prices, the short-term interest rate, and the real effective exchange rate, along the lines of the models estimated by Peersman and Smets (2001). In addition, the VAR for both the United Kingdom and the euro area includes a world commodity price index as an exogenous variable, while the VAR for the United States includes it as an endogenous variable.6 To control for changes in world demand and inflation, the VAR for the euro area and the United Kingdom also includes US real GDP and the US short-term interest rate. Given the relatively small sample period used for the estimation, no explicit assumption is imposed on the long-run cointegrating relationship among the variables in the models.7

8. The results show that interest rate sensitivity of output is roughly similar across the three economies (Figures 1 2-3 and text figure). In all three VAR models, a temporary rise in the nominal short-term interest rate is followed by a decline in output; the response is prolonged and hump-shaped, with the peak effect occurring in about six quarters. However, there is a difference in the duration of the response—the adverse impact on output lasts longer in the United Kingdom and the euro area than in the United States where the impact dissipates relatively quickly.

Figure 1.
Figure 1.

Impulse responses: Base VAR Model with U.K. Output

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

Figure 2.
Figure 2.

Impulse responses: Base VAR Model with Euro Output

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

Figure 3.
Figure 3.

Impulse responses - Base VAR Model with U.S. Output

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

uA02fig02

Effect of a Temporary 1% Point Increase in Short-term Interest Rate on Output

(In percentage points)

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

9. While the interest rate sensitivity of output is similar across the three economies, the composition of the output response is very different. The motivation for analyzing the composition is twofold. First, this could enhance our understanding of how the interest rate changes feed through the economy. Second, whether consumption or investment responds more, or more promptly, to an interest rate change is an issue of clear importance in the policy debate and in welfare analyses. For instance, in an economy with stronger response of consumption would suggest that consumer behavior is what needs to be monitored carefully in setting monetary policy.

10. The paper uses two approaches to examine the effects of a monetary policy shock on consumption and investment. The first simply involves re-estimating the VAR models substituting in turn real private consumption and real investment for real GDP. The second approach follows the methodology applied by Erceg and Levin (2002) and Angeloni et al. (2003), and replaces real GDP in the original VAR models with its main components—specifically, private consumption, investment8, and the rest of real GDP (government and net exports). The response of private domestic demand to a monetary policy shock is then obtained as the sum of the responses of real private consumption and investment, weighted by their shares in real GDP. The contributions from consumption and investment are then normalized to be comparable across the three economies.9

11. The results from the two approaches suggest that consumption is more sensitive to interest rates in the United Kingdom and the United States than in the euro area. In the first approach, consumption is not only more responsive to a unit interest rate shock in the United Kingdom and the United States, but the decline in consumption also leads a fall in GDP (see text figures). In contrast, the decline in investment following a unit increase in interest rate is larger and more persistent in the euro area. In the second approach, the contribution of consumption to the total response of private sector domestic demand is also significantly larger in the United Kingdom and the United States relative to the euro area. In the eight quarters following a monetary shock, the responses of consumption account for about three-quarters of the changes in the private domestic demand in the United Kingdom and the United States. In contrast, the contribution of consumption is substantially less in the euro area. In the euro area and to a lesser extent in the United Kingdom, investment maintains a significant contribution for a sustained period following the shock, while it disappears relatively more quickly in the United States.

uA02fig03

Effect of a Temporary 1 percent Increase in Short-term Interest Rate

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

uA02fig04

Comparison of Consumption and Output Responses 1/

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

1/ The responses of consumption are weighted by its share in GDP in each economy.

The Contribution of Consumption in the Total Response of Private Sector Domestic Demand 1/

(Share in the responses of private sector domestic demand)

article image

Cumulative effects. Contribution of consumption is calculated by obtaining the response of consumption and investment weighted by their shares in GDP, and then normalize the results so that they add up to one. The shares of each GDP components were calculated as the average shares from 1980:1 to 2003:1.

D. Housing and Consumption

12. These results raise a question as to what factors might explain why the structures of the UK and US economies are such that consumption tends to be more sensitive to interest rate changes. Differences in access to financial markets, regulations, flexibility of labor markets, legal frameworks, and other institutional and cultural influences are some of the key factors. However, in focusing on the housing sector, some important similarities emerge. In both the United Kingdom and the United States there is a competitive and well liberalized mortgage market; the housing sector plays an important role in the economy; interest rates have an important impact on housing; and housing exerts a significant influence on consumption. Moreover, particularly in the United Kingdom, the dominance of variable rate mortgages implies that the interest rate sensitivity of the household sector may be higher.10

13. Interest rates affect consumption via the housing sector directly and indirectly. Directly, the cash flow of households changes as debt service payments move in line with changes in mortgage rates (reflecting a pass-through from official policy rates). Indirectly, interest rate movements affect consumption through their impact on housing prices and subsequent effects on wealth and households’ access to credit.

14. Strong demand and a relatively slow supply response over time in the UK housing market have contributed to a significant rise in housing prices over the past 30 years. The United Kingdom has experienced three episodes of strong house price gains in the early 1970s, the late 1980s, and more recently since the mid-1990s (see text figure below). The recent sharp rise in housing prices reflects in part an increase in the demand for housing associated with a reduction in mortgage interest rates and greater access to credit. Although the debt burden of households has increased from around 95 percent of disposable income in the mid-1990s to about 125 percent by September 2003, the debt service as a proportion of disposable income has remained broadly unchanged.

uA02fig05

Real House Prices

(in log)

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

15. House prices are highly correlated with consumption (Figure 4). First, rising house prices reflect increasing demand for housing, and consumption of housing-related durable goods is closely correlated with this demand. But second, there is also an inverse relationship between household savings ratio and house prices, reflecting wealth effects driven by house prices. However, the impact of this wealth effect on consumption depends on households’ perception of its “permanence”. Given the high past volatility of house prices in the United Kingdom, there may be a some lag between a change in house prices (and the associated impact on wealth) and its effect on consumption. Moreover, housing equity withdrawal has provided an additional source of financing for consumption, generally at a substantially lower cost than other forms of credit available to households.

Figure 4.
Figure 4.

United Kingdom: Selected Key Housing Sector Variables

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

Source: Office of National Statistics, Bank of England and Fund staff estimates.1/ Detrended variables.2/ Household wealth deflated by consumption deflator.

16. To examine the channels through which interest rate changes impact consumption via the housing sector, another VAR model for the United Kingdom was estimated for the period 1980:1 to 2003:1.11 The model includes five variables: real consumption, consumer prices, real house prices, the ratio of outstanding loans to households secured on dwelling to total loans (bank and building societies), and the short-term interest rate.12 Real house prices are included primarily as a proxy for wealth effects on consumption. The ratio of loans secured on dwellings was included as a proxy for measuring households’ access to credit (although the house price variable might also pick up some of this effect). The ratio captures both changes in the availability of mortgage credit and of housing equity withdrawal.

17. As expected, results from the VAR model indicate that there is a significant interest rate effect on consumption coming through the housing sector.13

Impulse Responses of a 1 Percent Temporary Increase in the Short-term Interest Rate

article image

A temporary rise in the short-term interest rate is followed by a decline in house prices and the secured loan ratio. Consumption, in turn, is strongly correlated with changes in house prices and with the secured loan ratio. A variance decomposition analysis suggests that house prices explain a large share of UK consumption volatility. Although the initial impact is rather muted, after two years, changes in house prices explain about 45 percent of the fluctuations in UK consumption. Overall, consumption declines in response to a temporary one percentage point rise in the short-term interest rate with a short lag and reaches its maximum at nearly 0.4 percentage points below the baseline in 6 quarters. This impact is somewhat larger than results obtained from the other VAR models for the United Kingdom that exclude house prices.

Variance Decomposition of Private Consumption

article image
Sources: Fund staff estimates.

E. Conclusion

18. The evidence presented in this paper suggests that the interest sensitivity of output in the UK, euro area, and US economies are roughly similar, but that the main channels through which interest rates affect total demand differ significantly. In the United Kingdom, there is a high correlation between total output and private consumption. In turn, the analysis shows that consumption is highly sensitive to changes in interest rates. This sensitivity appears to reflect the importance of the housing sector and the development of UK financial markets, which provide greater access to credit to UK households. In this, the United Kingdom more closely resembles the United States than the euro area.

19. These results presented here, however, need to be interpreted with some caution. The estimates tend to be sensitive to the sample period reflecting the relatively short available data; and confidence bands around the impulse responses are generally wide, especially in the case of the United Kingdom. This reflects the fact that available data cover a period of significant structural changes in the UK economy and in its financial markets. In addition, with the recent increase in household debts combined with the dominance of variable rate mortgages, the economy’s interest rate sensitivity may have increased. In time, the interest sensitivity of euro area consumption to interest rates could rise, especially if structural reforms remove some of the barriers that prevent the euro area from developing mortgage markets more like those in the United Kingdom. In turn, the increase of long-term fixed rate mortgages in the United Kingdom may dampen the interest rate sensitivity of consumption. In the meantime, monetary policymakers in the United Kingdom need to monitor more carefully consumption and housing price developments in formulating policy than their euro area counterparts.

Figure 5.
Figure 5.

Impulse responses of the U.K. Consumption VAR

Citation: IMF Staff Country Reports 2004, 055; 10.5089/9781451814132.002.A002

APPENDIX Data Sources and Definitions (Source in parenthesis)

All data are in logarithmic form, seasonally adjusted excluding the short-term interest rates and secured loan ratio, which are in percentage points. Most euro area data are taken from the ECB Area Wide Model (AWM) database.

article image

References

  • Angeloni, I., A.K. Kashyap, and B. Mojon, 2003, Monetary Transmission Mechanism in the Euro Area, Cambridge University Press.

  • ______, and D. Terlizzese, 2003, “The Output Composition Puzzle: A Difference in the Monetary Transmission Mechanism in the Euro Area and US,” ECB Working Paper. No. 268

    • Search Google Scholar
    • Export Citation
  • Aoki, K., J. Proudman, and G. Vlieghe, 2002, “Houses as Collateral: Has the Link Between House Prices and Consumption in the UK Changed?,” Economic Policy Review, May, pp.163- 177.

    • Search Google Scholar
    • Export Citation
  • Bean, C., J. Larsen, and K. Nikolov, 2002, “Financial Frictions and the Monetary Transmission Mechanism: Theory, Evidence and Policy Implications”, ECB Working Paper No. 113

    • Search Google Scholar
    • Export Citation
  • Davey, M., 2001, “Mortgage Equity Withdrawal and Consumption,” Bank of England Quarterly Bulletin, Spring 2001.

  • Gerlach, S. and F. Smets, 1995, “The Monetary Transmission Mechanism: Evidence from the G-7 Countries”, Discussion paper No. 1219, Center for Economic Policy Research, July.

    • Search Google Scholar
    • Export Citation
  • HM Treasury, 2003a, “Housing, Consumption and EMU,” HM Treasury, June 2003.

  • ____, 2003b, “EMU and the Monetary Transmission Mechanism,” HM Treasury, June 2003.

  • Iacoviello, M. and R. Minetti, 2002, “The Credit Channel of Monetary Policy and the Housing Market: International Empirical Evidence,” Boston College Working Papers in Economics 541.

    • Search Google Scholar
    • Export Citation
  • IMF Country Report No. 02/46, Chapter II, “Why has U.K. Household Consumption Been so Strong?,” March 2002.

  • Kontolemis, Z., and H. Samiei, 2000, “The UK Business Cycle, Monetary Policy and EMU Entry,” IMF Working Paper 00/210.

  • Mojon, B., and G. Peersman, 2001, “A VAR Description of the Effects of Monetary Policy in the Individual Countries of the Euro Area,” ECB Working Paper No. 92

    • Search Google Scholar
    • Export Citation
  • Peersman, G., and F. Smets, 2001, “The Monetary Transmission Mechanism in the Euro Area: More Evidence from VAR Analysis”, ECB Working Paper. No. 91

    • Search Google Scholar
    • Export Citation
1

Prepared by Keiko Honjo.

2

See Chapter III for more details on the UK Treasury’s assessment of the five tests on the EMU entry.

3

To analyze business cycles, a Hodrick-Presscott (HP) filter was used to detrend seasonally adjusted quarterly data for GDP and its major components for the three economies over the period 1980:1-2003:1.

4

Other approaches include the use of large-scale country specific models to compare the overall strength of the interest rate sensitivity. See HM Treasury background study “EMU and the Monetary Transmission Mechanism” (2003b). For a comprehensive survey of the VAR for the United States, see Christiano et al (1999). For euro area, Angeloni, Kashyap and Mojon (2003) provide an overview of 16 recent VAR studies on euro area countries.

5

Weighted averages of data for the euro area countries for the period before the formation of the euro area in 1999; see Fagan et al (2001) for a description of these data.

6

The commodity price index is included in each model to reduce the impact of the price puzzle—a tendency in VAR models for the price level to rise in response to an increase in interest rate policy shock; see Sims (1992). It is treated as endogenous in the case of the United States because US demand is generally considered to have a significant impact on world commodity prices given the economy’s size.

7

All data are seasonally adjusted in logs, except for the short-term interest rates which are expressed in percentage points. Standard likelihood ratio tests are used to determine the lagorder of the VARs. The identification of monetary policy shocks is obtained using a Choleski decomposition. This involves the assumption that the monetary authorities observe the contemporaneous values of all variables in the model when setting the policy interest rate; however, all variables, excluding the exchange rate, respond with a lag to changes in the policy interest rate. The policy shocks have no contemporaneous impact on output and prices, but they may affect the exchange rate immediately.

8

The analysis uses private investment for the United Kingdom and the United States, and total investment for the euro area due to data availability.

9

As Angeloni et al. (2003) note, the relative contributions of private consumption and investment are evaluated in relation to real private domestic demand instead of real GDP because of the shortcut their approach takes in modeling the rest of GDP.

10

The UK Treasury’s EMU study “Housing, Consumption, and EMU” (2003) identifies the housing market to be one of the key structural differences between the United Kingdom and the euro area countries.

11

The model was only estimated for the United Kingdom due to lack of data availability on key housing variables for the euro area, particularly housing prices.

12

All variables are expressed in logs, except the secured lending ratio and the interest rate variables, which are expressed in percentage points.

13

The results are consistent with the household consumption function estimated in IMF Country Report 02/46, Ch. II.

United Kingdom: Selected Issues
Author: International Monetary Fund