Country Specific Effects of Euro Area Monetary Policy: The Role of Sectoral Differences

November 12, 2024

Country-Specific Effects of Euro-Area Monetary Policy: The Role of Sectoral Differences¹

Ruslana Datsenko² and Johannes Fleck³

Economic growth in some euro area countries has been lackluster since the COVID-19 pandemic. Concurrently, the ECB hiked its policy rate to fight inflation. In this note, we show that high interest rates have depressed economic activity more in those euro-area countries with large manufacturing sectors. Our analysis also suggests that activity in those countries stands to gain more from lower interest rates.

Introduction

In response to a burst of inflation, the European Central Bank (ECB) rapidly hiked its deposit interest rate from -0.5 to 4 percent between July 2022 and September 2023, as illustrated by the left panel of Figure 1. The bank then maintained this peak rate for about ten months until it cut the rate by 25 basis points in June, September and October 2024. Currently, market participants expect the ECB to cut further, for two reasons. First, euro-area headline inflation has declined from its record high of 10.6 percent in October 2022 to 1.8 percent in September 2024 and the latest ECB projections see inflation at or below 2 percent from 2025Q4.⁴

Second, economic activity in the euro area has been stagnant since mid 2022, as shown in the right panel of Figure 1, and the ECB expects growth to pick up notably only from mid- 2025. Hence, as inflationary pressures have subsided, the bank's policy focus has now shifted towards stimulating a faster rate of economic expansion. However, the figure also points to remarkable growth differences within the euro area. Germany, the currency bloc's largest economy, performed comparatively well during the COVID-19 pandemic but has been stagnant since then. Spain, on the other hand, contracted sharply during the pandemic but has recorded strong growth ever since - despite the ECB's restrictive stance.

While some of the early post-pandemic growth of the Spanish economy presumably reflected a catch up from its contraction during the pandemic, its growth rate has remained above pre-pandemic prints. The German economy, on the other hand, has been hit especially hard by several post-pandemic adverse events.⁵ Furthermore, as we argue in this note, due to its sectoral composition, high ECB interest rates appear to have affected Germany's economy more than other euro-area economies; as Figure 2 shows, measured in value added terms just before the COVID-19 pandemic, the manufacturing sector in Germany is larger than the euro-area average while the service sector is comparably small. Spain, on the other hand, has a much smaller manufacturing and a somewhat larger service sector.

Since the seminal contribution of Bernanke and Gertler (1995), it is well-understood that restrictive monetary policy tends to have a stronger impact on activity in the manufacturing than in the service sector.⁶ The main reason is that demand for manufactured goods, especially for durable and business equipment goods, is more interest rate sensitive than demand for services. Another reason is that manufacturing firms typically have larger capital expenses than service firms, making their supply decisions more sensitive to interest rate changes.

For the U.S., we demonstrate in Datsenko and Fleck (2024) that these different sensitivities are reflected at the level of various geographical units; counties, core based statistical areas and metropolitan statistical areas with high shares of manufacturing workers experience stronger employment declines following increases in the Federal Reserve's policy rate. Other research on geographic variation in the transmission of monetary policy, for example, Bellifemine, Couturier and Jamilov (2023), studies county-level employment in sectors producing tradeable vs. non-tradeable output and arrives at conclusions consistent with our analysis.

Estimation and data

To investigate whether differences in sectoral composition contribute to differential impacts of restrictive monetary policy across euro-area economies, we use a local projection framework (Jorda, 2005).⁷ This framework allows us to employ a widely used measure characterizing changes in the ECB's monetary policy stance. Specifically, we use the monetary policy shock series of Jarocinski and Karadi (2020). This series aims to measure surprises in the central bank's policy decisions, excluding anticipated changes, which are already reflected in market conditions prior to the ECB's official announcements.

Our local projection specification is as follows:

$$$$\Delta y_{t+h,t} = \frac{y_{t+h}-y_{t-1}}{y_{t-1}} = \alpha^h + \nu_q^h + \eta^h d_t +\sum_{j=0}^{8} \beta^{h,j}e_{t-j}^{MP} + \sum_{j=1}^{2} \phi^{h,j} \Delta y_{t-j} + \gamma^hX_{t-1} + \varepsilon_t$$$$

$$\Delta y_{t+h,t}$$, is the cumulative percentage change in an outcome variable between time $$t-1$$ and $$t+h$$, where $$h$$ refers to the estimation horizon. As our outcome data are quarterly, we choose an estimation horizon of $$h$$ = 0, 1, ..., 12 which allows us to estimate effects until about three years after the change. $$\alpha$$ is an estimation horizon specific intercept and we also include quarter fixed effects, $$ν$$, and an indicator $$d$$, which is equal to 1 during the Great Financial Crisis of 2007 and 0 otherwise. $$e^MP_t-j$$ is the monetary policy shock series which measures changes in the ECB's main policy rate (we use shocks at $$t$$ and add lags $$j$$ to capture delayed effects). Finally, $$\Delta y_{t-j}$$ are lags in the outcome variable and $$X$$ denotes a vector of control variables. Throughout our estimations, we use Newey and West (1987) standard errors which account for serial correlation in the error term, $$\varepsilon$$.

In our specification, the beta coefficient estimates capture the effect of ECB monetary policy shocks on the outcome variable at each horizon, comprising an impulse response function. To estimate it, we use quarterly data on euro-area aggregate and country-specific variables from Eurostat. We include years from 1999 (when the Euro was introduced) to 2019 in order to ensure our results are not confounded by the highly unusual behavior of variables during the recent pandemic and the subsequent large energy supply shocks. Finally, we include all twenty countries which are currently euro area members.⁸

Results

Our baseline estimates, shown in Figure 3, illustrate that making euro area monetary policy more restrictive reduces aggregate economic activity, but the response differs between sectors. We arrive at this finding by estimating the above equation separately for real GDP and for manufacturing and service value added at the euro area level. In the estimation, we include lags of euro-area inflation and unemployment rates as controls.

The panels of Figure 3 show the effect of an unexpected one percentage point increase in the ECB's policy rate. The left panel indicates that this increase lowers euro-area real GDP by around one percent after three to four quarters. The right panel shows that it also reduces service and manufacturing sector value added for about eight quarters. Yet, the estimated effect is not statistically significant after four quarters for the service sector and after seven for the manufacturing sector. However, the effect is notably larger for the manufacturing sector; the peak impact is nearly three percent, around three times larger than the peak impact on service sector value added.

Next, we implement the estimation for each euro-area country separately. In addition to the aggregate controls for inflation and unemployment, we also include country specific variables (mean income and lags of changes in real GDP), which, together with the country specific term alpha, help control for characteristics other than sectoral differences. For each country, we report on the vertical axis of Figure 4 the effect of a one percentage point policy rate surprise on GDP (averaged over the first ten quarters following the surprise). On the horizontal axis, we plot each country's service (left panel) and manufacturing share (right panel).

We find that countries with a larger manufacturing share experienced a stronger GDP decline following an increase in the ECB's policy rate. For instance, the effect on Germany is about twice as large as on the next three largest euro-area economies. The opposite finding applies to the service sector; GDP in countries with larger service sectors tends to be less affected by increases in policy rates.⁹

Asymmetries: Restrictive vs. accommodative monetary policy changes

Up to this point, our estimation results have been produced under the assumption that changes in restrictive and accommodative monetary policies have symmetrical effects on GDP and sectoral value added. However, it has been argued that monetary tightening can have a stronger impact than easing, as the two affect prices, wages, and credit conditions differently. For example, Debortoli, Forni, Gambetti, and Sala (2020) provide evidence in support of this view.

As the ECB is currently expected to continue moderating its restrictive stance by lowering its policy rate, we also explore asymmetries between the effect of interest rate hikes and cuts. Accordingly, we now estimate the GDP impulse responses separately for negative and positive shocks; differences between the absolute values of these responses may suggest asymmetrical effects of restrictive and accommodative monetary policy changes.

In Figure 5, we report our findings for the effect of accommodative monetary policy changes, that is interest rate decreases. The findings indicate that making the ECB's monetary policy less restrictive has a stronger impact on GDP in countries with a higher manufacturing share. The opposite is true for countries with a larger service share. Importantly, the effects are overall symmetric between restrictive and accommodative changes as Figure 5 is inversely mirroring Figure 4. This provides support for the view that GDP in euro-area economies with larger manufacturing sectors might increase more as the ECB cuts rates.¹⁰

Conclusion

Our analysis provides evidence that ECB restrictive monetary policy has a stronger impact on real GDP in euro-area economies with a larger manufacturing sector. This evidence suggests that the interest rate hikes between 2022 and 2024 have had a stronger contractionary impact on GDP in euro-area economies such as Germany. We also find that the effects of restrictive and accommodative shocks are broadly symmetrical which indicates that an easing of the ECB's monetary policy stance is likely to benefit countries with larger manufacturing sectors more than those with smaller ones.

References

Bellifemine, Couturier and Jamilov (2023), "The Regional Keynesian Cross", working paper

Bernanke and Mark (1995), "Inside the Black Box: The Credit Channel of Monetary Policy Transmission", Journal of Economic Perspectives, 9 (4): 27-48.

Datsenko and Fleck (2024), "Structural Transformation and the Efficacy of Monetary Policy", working paper

Debortoli, Forni, Gambetti and Sala (2020), "Asymmetric Effects of Monetary Policy Easing and Tightening", working paper

Galesi and Rachedi (2018): "Services Deepening and the Transmission of Monetary Policy", Journal of the European Economic Association, 17(4), 1261-1293.

Jarocinski and Karadi (2020), "Deconstructing Monetary Policy Surprises - The Role of Information Shocks", American Economic Journal: Macroeconomics 2020, 12(2): 1-43

Jorda (2005), "Estimation and Inference of Impulse Responses by Local Projections", American Economic Review, 95(1), 161-182.

Kreamer (2022), "Sectoral Heterogeneity and Monetary Policy", American Economic Journal: Macroeconomics, 14 (2): 123-59.

Newey and West (1987), "A Simple, Positive Semi-Definite, Heteroskedasticity and Auto-correlation Consistent Covariance Matrix", Econometrica, 55(3), 703-708.

Public link

Please use the above public link if you want to share this noodl on another website.