Since the death of the Bretton Woods system, the yen has actually seen several episodes of solid appreciation, consisting of in the so late 1970s, after ~ the 1985 Plaza Agreement, the early and also late 1990s and also after 2008. This appreciations have actually not just been associated with “expensive yen recessions” result from an unfavorable effects ~ above exports; due to the fact that the late 1980s, the strong yen has also raised concerns around a de-industrialisation that the Japanese economy. Versus this backdrop, the write-up investigates the effects of transforms to the yen exchange rate on the hollowing the end of the Japanese industrial sector. Come this end, the post uses both aggregate and industry-specific data to gauge the impacts of yen fluctuations ~ above the output and also exports of different Japanese industries, exploiting new data for industry-specific real efficient exchange rates. Our findings support the check out that the periods of yen appreciation had more than simply transitory results on Japanese manufacturing. The outcomes also provide indication of hysteresis effects on manufacturing. While there are certainly also other factors that have added to a hollowing the end of Japanese industry, a solid yen play a role, too.

You are watching: An appreciating yen makes japanese products


A solid yen – endaka (円高) in Japanese – has actually repeatedly brought about distress among Japanese policymakers and manufacturers. Because the demise of the Bretton Woods device in 1971, the yen has actually seen number of episodes of solid appreciation, consisting of in the so late 1970s, after the 1985 Plaza Agreement, the early and also late 1990s and after 2008 (Fig. 1). These appreciations have actually not just been associated with “expensive yen recessions” – endaka fukyo (円高不況) in Japanese – result from negative effects on exports; because the late 1980s, the strong yen has likewise raised concern about a de-industrialisation the the Japanese economy. Indeed, the share of manufacturing in full output has decreased from 26.7% in 1980 come 20.5% in 2015, if the share of production in total number of employed persons has declined from 27.0% come 16.2% between 1970 and also 2016 (Fig. 2). The decline in the share of manufacturing in full employment or GDP is not unique to Japan. That is without doubt a phenomenon that deserve to be seen throughout industrialised economies, with varying degrees. What provides Japan a specifically interesting situation to analyse is the the yen’s real effective exchange price has displayed much larger volatility than the currencies of other huge advanced economies. Between 1980 and also 2018, the standard deviation of Japan’s annual real reliable exchange price was 17.0, contrasted to 12.3 because that the U.S., 10.4 because that the unified Kingdom, and 5.3 for Germany. Exchange rate impacts on production may because of this have been much more pronounced than elsewhere.

Fig. 1



While the solid yen and also its potentially de-industrialising impact have received much attention in the political and economic plan discourse in Japan, there has actually been surprisingly tiny research on this in the scholastic literature. As mentioned by Hamada and also Okada (2009: 200), many research top top Japan’s “lost decade” and also hollowing out of the Japanese industry “have been extensively focused top top its real and domestic aspects, such as complete factor productivity (TFP), expansion decline, non-performing loans, and governance.” Hamada and Okada (2009: 200) argue that monetary and also exchange price policy have actually played an essential role in this, and also that “Japanese industries endured a heavy burden” because of a greatly overvalued genuine exchange rate.

Against this backdrop, the article intends to inspection the effects of yen fluctuations on commercial production and also the hollowing out of the Japanese industry. The significant contribution that this study is an econometric analysis using new data for industry-specific real efficient exchange rates (RIETI 2018) to gauge the impacts of yen fluctuations on the output and exports of various Japanese sectors at monthly frequency. Together recently shown by fort et al. (2018) because that the us economy, alters to manufacturing employment have actually differed considerably across sectors. The use of monthly, industry-specific data permits us to better capture potential impacts of real effective exchange rate movements. We additionally use yearly data since 1970 to inspection potential long-term results of yen exchange rate transforms on production employment. Complying with Rowthorn and also Coutts (2004: 767), we define de-industrialisation – come which we likewise refer as “hollowing out” – together “a secular decline in the re-superstructure of production in nationwide employment”.

The object is not only of relevance to the Japanese economy, it also bears relevance because that other advanced economies. Indeed, recent years have seen an intense discussion in several member nations of the Eurozone about purported an unfavorable effects of a also high euro exchange rate on manufacturing and employment (e.g., Belke and Volz 2015). Because that instance, Bénassy-Quéré et al. (2014: 7) calculation that for France “a 10% depreciation in the euro in relation to a partner country outside the eurozone boosts the worth of the average exporter’s sales come this nation by around 5-6%”, with many of this result realised in the exact same year as the depreciation. Likewise, the U.S. Has actually recently watched a renewed conversation on de-industrialisation, through a prominent function of the dissension exchange rate in this discourse (e.g., Campbell 2017; Levinson 2017). A much better understanding of any type of exchange rate impacts on Japanese production should also provide insights because that economic and also monetary plan elsewhere.

The reminder of the post is structured as follows. Ar 2 provides a evaluation of the literary works on the effects of exchange rate changes on de-industrialisation. Section 3 comprises our empirical analysis. The empirical part is split into two sections: Firstly, we conduct an empirical analysis with annual, accumulation data indigenous 1970 come 2016 that allows us to investigate potential long-term effects of yen exchange rate transforms on manufacturing employment. Secondly, we use monthly, industry-specific data indigenous January 2001 come June 2017 come gauge the dynamics the yen exchange rate transforms on Japanese sectoral manufacturing employment. Ar 4 summarises and also concludes.

It is well established that depreciated genuine exchange rates can assist to wake up industrial breakthrough and economic growth (Rodrik 2008).Footnote 1 return there seems to exist a “fear the appreciation” (Levy-Yeyati et al. 2013) fairly little research has been carried out on the result of overvalued exchange rates on economies. The first systematic empirical cross-country evaluation to investigate the results of large exchange rate appreciation on present account balances and also on actual output was conducted by Kappler et al. (2013), who discover that illustration of strong exchange price appreciations are 2175forals.comed with deteriorating current account balances and a slow-down of genuine export growth, but no significant effects ~ above output. Likewise, Bussière et al. (2015) examine to what extent huge and rapid real exchange price appreciations impact on financial growth. Utilizing a sample that 53 emerging and progressed economies, they find that while big appreciations dampen export growth and boost import growth, output expansion is higher on average.

The literature on de-industrialisation has concentrated less top top exchange price valuations than on transforms in specialisation, consumption, technical progress and also productivity, worldwide trade and also investment trends (Rowthorn and Coutts 2004). The major exemption is the U.S. Economy, for which the result of the dissension exchange rate on the U.S. Industry has actually been studied widely. The result in this literature are mixed, through some studies finding a big effect (Branson and also Love 1986, 1987, 1988a, b; Revenga 1992; Gourinchas 1999; Kandil and also Mirzaie 2002, 2003; Klein et al. 2003; Campbell 2017) and others a tiny or no impact (Glick and also Hutchison 1990; Goldberg 1993; Campa and Goldberg 1995, 2001; Bahmani-Oskooee and also Mirzaie 2000; Bahmani-Oskooee et al. Bahmani-Oskooee et al. 2007).Footnote 2

According come Rowthorn and Ramaswamy (1997a, b) and Rowthorn and also Coutts (2004, 2013), the most vital internal/domestic components contributing to de-industrialisation are productivity expansion in the manufacturing sector and also shifting trends of domestic expenditure, i.e., with climbing income usage tends to move towards an ext services and also away native manufactured products (known together Engel’s law or Bell’s Law). The most vital external determinants are established as “North-South” trade and the outsourcing of labour-intensive manufacturing to low-wage countries. However, domestic factors, particularly a greater productivity development in manufacturing than in services, space widely viewed as the dominant factor behind de-industrialisation (Rowthorn and also Ramaswamy 1997a, b; Rowthorn and also Coutts 2004, 2013; Lawrence 2017).Footnote 3

Surprisingly couple of studies have looked right into the perhaps de-industrialising effect of endaka, also though Obstfeld (2010) pointed out that Japan’s real economic growth rate has actually been strong negatively associated with the level the the yen’s real reliable exchange rate. Only couple of studies deserve to be discovered that systematically shot to verify the theory that the strong yen has added to de-industrialisation and also outsourcing of market in Japan, including Dekle (1996), Dekle et al. (2010) and also Yamashita (2013, 2015). Dekle (1996) and also Dekle et al. (2010) find that yen appreciations that 1985 and 1995 substantially hurt the ability of Japan to compete with the U.S. By raising the loved one production costs of Japanese industries. Yamashita (2013, 2015) finds the yen appreciation and outward FDI contributed to de-industrialisation and also suggests that a weak yen may contribute to reshoring.Footnote 4

A cable of studies examines the role of Japanese outward direct investment in the hollowing out process. Japanese outward straight investment took turn off in the mid-1980s (Fukao et al. 2003). The number of Japanese multinationals raised by 290% between 1985 and 1992, while abroad production that Japanese firms raised from 3% in 1982 come 17% in 2002 (Ryan and also Toubal 2017). Analysing data top top Japanese repurchase in the United says from 1975 come 1992, Bloningen (1997) finds that genuine dollar depreciations enhanced the likelihood that Japanese acquisitions in U.S. Industries. Likewise, the BOJ (2011) deadline the renewed moving of production abroad because 2008 come the relative appreciation that the yen compared to the levels before the Lehman shock.

Empirically, the impact of outsourcing on residential manufacturing employment is mixed. According to Baldwin (2006), offshoring labour-intensive production processes to nearby lower-cost places in Southeast and Northeast Asia due to the fact that the mid-1980s go apparently enable Japanese service providers to retain components of production production domestically if maintaining international competitiveness. The BOJ (1989) find that raised outward straight investment added to long-term structural mediate in the fiddle sector. This added to a short-term surge in the export of funding goods and also parts to overseas production bases that Japanese companies. Yamashita and also Fukao (2010) find no indication that development of abroad operations through Japanese production multinational enterprises diminished home employment end the period 1991–2002; indeed, their evaluation suggests that it may have added to maintaining the level of house employment. A current study by Ryan and Toubal (2017) supplies Japanese firm-level data for the duration 1982–2001 to compare employment in between firms with and also without international affiliates and also the level to i beg your pardon the advance of international operations influence firm-level domestic employment. It finds “some restricted evidence” that the hollowing out of the Japanese economic climate by Japanese multinational firms shifting manufacturing abroad native 1991 onwards.

Using the same industry-specific exchange prices that will certainly be to work in our empirical evaluation later, Sato et al. (2013) analyse export price competitiveness and also export performance of Japan, China and also Korea and also find that the solid yen after ~ 2007 destroyed the Japanese electronic devices sector. Likewise, Thorbecke (2012) finds in a study focused on the Japanese electronic devices industry the the yen appreciation in between 2007 and also 2011 had actually a very damaging result on electronics exports. However, neither research controls for employment effects. Kato (2018) examines the results of exchange price changes and also productivity on manufacturing exports because that the period 2002–2012 and also finds exchange rates to be important factors to affect firm-level exports.

An important issue in this paper definition is the exchange rate appreciations and also depreciations might not have symmetrical effects. That is, a period of solid appreciation may lead to outsourcing and also the lose of production capacities – and also employment – in the house country, yet a depreciation later on may not bring around a return (or “re-shoring”) the manufacturing tasks and employment, given that long-term investments have been made abroad, which will not be reversed. This hysteresis results have been explained in sunk cost models of trade hysteresis (Krugman 1988; Baldwin and Krugman 1989; Baldwin 1990; Baldwin and Lyons 1994) and also empirically investigated greatly for the U.S. And a variety of European economies.Footnote 5 Giovannetti and also Samiei (1995) conduct an empirical analysis of the importance of hysteresis in global trade, using production exports for the united States, Germany, and also Japan. They discover “strong evidence” in favour the the presence of hysteresis just in the instance of Japanese exports.

For our econometric analysis, we use both accumulation and industry-specific data to gauge the impacts of changes of the real effective yen exchange rate. Taking account that the availability of different data series, us firstly command an empirical analysis with annual data native 1970 come 2016 that allows us to inspection potential long-term impacts of yen exchange rate changes on production employment. Secondly, we usage monthly data indigenous 2001:M01 to 2017:M06 to examine the dynamics that yen exchange rate alters on Japanese sectoral manufacturing employment, exploiting a brand-new monthly dataset top top Japanese sectoral exchange prices which are only available from 2001 onwards. Because that this purpose, we use a combined time series-panel framework.

Analysis with annual Data, 1970–2016

Data and also Empirical Model

Building on Rowthorn and Ramaswamy (1997a) and also Rowthorn and Coutts (2004, 2013), we estimate the complying with model because that Japan and annual data varying from 1970 to 2016Footnote 6:

$$ EMPMAN=eta_0+eta_1 EMPMAN_t- 1+eta_2 CAP+eta_3 EXPMAN+eta_4 OFDI+eta_5 TFP+eta_6 EXPCHIN+eta_7 EXR+varepsilon, $$

with EMPMAN = Share of production in the total variety of employed people (%), CAP = Gross domestic fixed funding formation (% that GDP), EXPMAN = Manufacturing exports, OFDI = Outward straight investment, TFP = Total element productivity at constant national prices, EXPCHIN = Chinese exports, and also EXR = Real reliable exchange rate. Attachment Table 14 provides more details top top the variables and sources.

The choice of explanatory variables in (1) is encouraged by the typical variables offered in previous research studies on de-industrialisation. To this we include EXR together the change of our specific interest to inspection the hypothesis that activities of the yen’s real effective exchange price have added to a de-industrialisation of the Japanese economy.

Increases in fixed capital formation can theoretically have actually both optimistic and an adverse effects on production employment. On the one hand, an enhanced investment in the residential manufacturing sector might have hopeful employment effects. On the other hand, however, funding investment could advancement the automation of manufacturing production, with neutral or even an adverse effects on manufacturing employment. In contrast, rise in manufacturing exports should clearly have a positive impact on residential manufacturing employment. As debated above, the impact of outward direct investment deserve to be ambiguous: on the one hand a shifting of production abroad might have an unfavorable employment effects, but at the exact same time there might be optimistic outcomes since it may increase the competitiveness of residential operations which would certainly remain essential parts of regional or worldwide value chains (Baldwin 2006; Yamashita and Fukao 2010).

As stated above, productivity increases have actually been identified in the de-industrialisation literary works as a significant factor behind the loss in production employment, so we would suppose a an unfavorable coefficient. We usage TFP as our measure up of productivity. We taken into consideration using alternate measures the productivity, such as multifactor productivity, but were not able to acquire data ranging ago to the 1970s.

The impacts of Chinese export development are, again, ambiguous.Footnote 7 if Chinese export development is a have fun of China’s rapid growth process, i m sorry has also opened up good opportunities for Japanese companies, Chinese exporters have increasingly become competitors of Japanese production firms. Last however not least, we would expect the coefficient calculation for the real reliable exchange price – our main variable of interest – come be negative if the hollowing out hypothesis were to hold.

Figure 3 graphically screens the time series of the variables provided in ours estimations based on annual data. It conveys a an initial impression that the stationarity nature of the time collection used. At first glance, every time series appear to contain, most likely stochastic, trends. We will certainly explicitly inspect for stochastic trends prior to we start our estimation practice in order to make sure that the stochastic properties of the contained variables meet the standard presumptions of our regression analysis. Moreover, some an initial indications the correlations can be acquired as well. Even if it is this way “causation” in one econometric feeling will it is in checked after that in this section.


The peak left panel of Fig. 3 reflects the two main variables the interest, the re-publishing of manufacturing in total variety of employed persons and also the real efficient exchange rate. In between 1970 and also the mid-1990s, the real effective exchange rate showed a clear long-term appreciation trend, but with two durations – 1978-1982 and 1988–1990 – the depreciation. After a optimal in 1995, the real reliable exchange rate then reflects a declining trend, albeit with durations of strong appreciation in 1998–2000, 2007–2011 and also 2015–2016. Over the entire sample period, the re-publishing of manufacturing in total variety of employed persons shows a long-term decreasing trend, although the share of production remains stable over the durations 1987–1992, 2005–2007 and also 2012–2016.

Unit source Tests

We proceed with unit source tests the the variables to be employed in ours empirical model. The outcomes of our single time collection unit source tests are displayed in Table 1. Every variables rotate out to be combined of stimulate one (I(1)), except Chinese exports EXPCHIN which appears borderline I(2).

For the real effective yen exchange rate (EXR), with exception of the ADF-test, all complementary unit source tests consisting of the Dickey-Fuller GLS (ERS) test, the Phillips-Perron test, the Kwiatowski-Phillips-Schmidt-Shin test and also the Ng-Perron unit source test suggest into the direction of one I(1) process. Hence, we continue with the yen exchange rate considering it as an I(1) variable. The results are obtainable on request.

As a robustness check, we additionally conducted Dickey-Fuller GLS tests, Phillips-Perron tests, Kwiatkowski-Phillips-Schmidt-Shin (KPSS) and Ng-Perron unit source tests not just for the yen exchange price but also for all various other variables. The results as whole stayed the same. Among others, they support treating EXPCHIN finally as an I(1) change in ours regressions.

Given the unit source tests conducted in this section and their quite unambiguous results, us feel legitimised to employ the levels of the variables in the structure of a cointegration strategy (option 1) and/or usage the first differences of all variables as stationary variables in our Autoregressive distributed Lag (ARDL) estimations (option 2).

Regarding ours regression evaluation two issues concerned mind. Firstly, over there is a fairly broad agreement in the literature that one might well find impacts of exchange rate changes on employment also if there is less evidence of an influence on trade (Belke and Gros 2001). Secondly, a thorough regression analysis is crucial to take into consideration leads and lags that the variables when it involves establish any kind of one-way “causation”. This is dramatically presented by the key results that the fiddle hysteresis literary works which reflects that a positive coefficient the the exchange rate in export, employment, invest or commercial output equations may simply hide a more facility dynamic the the partnership which may include stepwise an adverse effects that an evaluation of the home currency (Krugman and Baldwin 1987; Belke et al. 2013).Footnote 8

We now turn to the first option, the estimation of a cointegrating equation in stimulate to assess the influence of the yen exchange price (EXR) ~ above the share of manufacturing in total number of employed people in Japan (EMPMAN), taking the various other variables consisted of in eq. (1) right into account all at once in the long-run relationship.

A Cointegration Approach: searching for Long-Run “Causation” that “Hollowing out”

As our cointegration frame we select Dynamic Ordinary least Squares (DOLS). This is because DOLS allows us to check for causality running indigenous the independent to the dependency variables. The DOLS estimation technique controls because that endogeneity the explanatory variables (Stock and also Watson 1993; Wooldridge 2009). Endogeneity in the form of feedback effects or reverse causality between the dependent and independent variables would bring about a misspecification that our approximated models. More specifically, the DOLS procedure controls for endogeneity of all explanatory variables by inserting leads and also lags the the alters of all exogenous variables. Hence, DOLS is a powerful estimation technique, where typical errors space corrected because that heteroscedasticity and cross-section correlation (Saikkonen 1992; Stock and Watson 1993). By inserting the leads and also lags the the exogenous variables in an initial differences, the explanatory variables in levels end up being (super-) exogenous and also the regression outcomes unbiased (Wooldridge 2009).

Table 2 display screens the outcomes of a DOLS estimation of our totally specified empirical model (Eq. (1)). The share of production in total number of employed people (EMPMAN) to represent the dependence variable. The approximated coefficient of the yen exchange rate (marked in bold) is an adverse as meant from theory and also highly significant. Checked out on the whole, likewise the other approximated coefficients are far-reaching and their signs corresponding with our priors. TFP shows a an unfavorable coefficient, arguing that higher productivity has actually resulted in a reduced share of production employment, however this effect is not statistically significant. Gross domestic fixed funding formation mirrors a confident sign and is highly significant. Together expected, production exports have a optimistic and significant effect on manufacturing employment, whereas outward straight investment has a an adverse and far-ranging effect. Chinese exports present a confident sign, arguing that the positive results of Chinese economic advancement on Japanese manufacturing overcame the an adverse competition effect. Finally, the empirical realisations the the goodness-of-fit criteria, among them the really high R-Squared, show the appropriateness of our selected empirical model.

Our robustness checks reveal that the full goodness-of-fit does not come to be significantly reduced if EXPCHIN is removed from our empirical model. This does not come as a surprise since both variables represent indications of the world organization cycle, i.e., the so-called “global factor”. For exactly this reason, we leave the end EXPCHIN in our next DOLS-specification. The matching regression results are displayed in Table 3. Again, the estimation results clearly are in accordance v our priors.

As robustness checks we additionally conducted DOLS regressions with White mediate of the coefficient covariance matrix (White 1980). There is no substantial readjust of the estimate results, v respect come the magnitude and the sign of the estimated coefficients. If at all, the definition of, because that instance, the yen exchange rate, also slightly boosts once again in the specification not included Chinese exports. Moreover, we also enacted DOLS regression through Newey-West correction of the coefficient covariance matrix (Newey and West 1987). Again, the estimate results perform not adjust much v respect to both the magnitude and the authorize of the estimated coefficients. Every these results are obtainable on request.

See more: The Dream Is Still Alive By Wilson Phillips The Dream Is Still Alive

Application the the DOLS procedure requires the time series to be non-stationary and to be cointegrated end time. We are able to display that both conditions are solve in our case. Proof of non-stationarity has been delivered already by the outcomes of our unit root tests in ar 3.1.2. And empirical support of cointegration among the variables contained in our empirical models is conveyed by cointegration test (Table 4 for the completely specified model and Table 5 for the version without Chinese exports). Figures 4 and also 5 display the residuals (which type the basis because that the Engle-Granger cointegration tests explained in Tables 4 and 5) and also the actual and fitted worths according to the DOLS regression models described in Tables 2 and 3, respectively.