Japan’s Unified Local Elections: Key Figures

Commentary using these figures appeared at Tokyo Review.

Figure 1: Prime Minister Kishida’s Cabinet Support Rate

The moving average includes data from surveys conducted by JNN, ANN, NHK, Jiji, Kyodo, Mainichi, Asahi, Yomiuri, FNN (Sankei), and Nikkei news organizations.
Figure 2: Prime Minister Kishida’s Net Support Rate

Note: The moving average includes data from surveys conducted by JNN, ANN, NHK, Jiji, Kyodo, Mainichi, Asahi, Yomiuri, FNN (Sankei), and Nikkei news organizations. Calculated by subtracting the percentage of respondents who support the cabinet from those who explicitly said they did not support the cabinet.
Figure 3: Unified Local Elections Unification Rate (1947-2023)
Figure 4: Unified Local Election Voter Turnout and No-Contest Percentages (1947-2019)
Figure 5: Prefectural Assembly Seat Acquisition Percentage (Unified Local Elections 1959-2019).

Note: The LDP only contested its first ULE as a merged party in 1959. Socialist Party trend line includes both the performance of the Japan Socialist Party and its successor, the Socialist Democratic Party. The Democratic Party trend line includes both the DPJ and CDPJ incarnations.

Figure 6: Comparison of Official Prefectural Assembly Candidates of Five Smaller Political Parties between the 2019 and 2023 Unified Local Elections
Figure 7: Ishin no Kai Prefectural Assembly Candidates 2019 and 2023 Unified Local Elections.

Note: Pink= 1-9 candidates; Yellow=10-19 candidates; Orange=20+ candidates. Grey is no candidates or no prefectural assembly held.
Figure 8: Constitutional Democratic Party Prefectural Assembly Candidates 2019 and 2023 Unified Local Elections.

Note: Pink= 1-9 candidates; Yellow=10-19 candidates; Orange=20+ candidates. Grey is no candidates or no prefectural assembly held.
Figure 9: Opposition Contestation Ratio 2019 and 2023 Unified Local Elections.

Note: ratio calculated by pooling the official candidates from four opposition political parties: Ishin no Kai, the CDP, the JCP, and the DPFP, and comparing it to the total number of assembly seats available. Light blue is a ratio of between 0 and 0.29; moderate blue is between 0.3 and .69; dark blue is a ratio of over 0.7.
Figure 10: Prefectural Assemblies with Increased Opposition Contestation Ratios between 2019 and 2023 Unified Local Elections.

Note: Light blue is an increase of between 0 and 25% in the ratio; moderate blue is between 25% and 100%; dark blue is a doubling or larger of the ratio between 2019 and 2023.

Political Survey Data Points for 2022

This is an assortment of tables and figures I will be using for a year in review article.

AgencyMonthSupportOppose
FNNApril5738.9
NikkeiApril5533
KyodoMay54.843.1
FNNMay6236.1
NikkeiMay5631
MainichiMay7623
NHKMay5236
KyodoJune53.139.1
NikkeiJune5437
FNNJune72.725.3
YomiuriJune5341
ANNJune5037
JijiJune49.735
NHKJune5237
KyodoJune47.845.8
AsahiJuly3942
KyodoJuly55.537.8
KyodoSeptember56.340.8
NHKOctober5529
FNNOctober62.529.8
KyodoNovember7124.9
ANNNovember6327
KyodoDecember3953.6
MainichiDecember4841
FNNDecember45.848.3
AsahiDecember4648
NikkeiDecember4745
JijiDecember35.533.6
YomiuriDecember5142
JNNDecember5336
ANNDecember4441
*Opposition in most cases is a composite of preferences for either keeping spending the same or decreasing it.

Constitutional Democratic Party’s Deterioration in Support between Elections
(Average of 10 Media Survey Organizations)
Average Party Support Rate 2017 (November)13.63%
Average Party Support Rate 2019 (July)8.01%
Average Party Support Rate 2021 (October)7.56%
Average Party Support Rate 2022 (July)5.96%
Difference between November 2017 and July 2022-7.67
Difference between July 2019 and July 2022-2.05
Difference between October 2021 and July 2022-1.6

Movement in PM Kishida’s Net Support Rate (Average of 10 Media Surveys)
October 2021 to May 2022+12.38%
May 2022 to December 2022-54.21%
Overall (October 2021 to December 2022)-41.83%

Japanese Responses to Surveys on Foreign Territory Strike Issues in 2022

AgencyMonthAgreeDisagreeContention
FNNJanuary43.550.6In response to North Korea’s repeated missile launches, PM Kishida expressed his belief that Japan should consider acquiring the capability to attack enemy bases.
KyodoFebruary49.145.1The government & ruling parties are considering the acquisition of “enemy base strike capability” to target missile bases and other facilities in a foreign country before being attacked. 
YomiuriFebruary4346Japan should have an “enemy base attack capability” to destroy missile bases and other facilities in other countries before Japan is attacked .
JijiFebruary34.326.1Japan to possess the ability to attack enemy bases.
YomiuriApril4646Japan should have an “enemy base attack capability” to destroy missile bases and other facilities in other countries before Japan is attacked.
MainichiMay6622The LDP has proposed that the SDF possess a “counterattack capability” to attack enemy missile bases and other targets.
AsahiMay4449The SDF does not possess the ability to attack enemy missile bases as Japan relies on the U.S. military for foreign territory strike. Japan should have this capability.
AsahiMay3360The SDF does not possess the ability to attack enemy missile bases as Japan relies on the U.S. military for foreign territory strike. Japan should have this capability.
NikkeiMay6030PM Kishida is considering allowing the possession of a “counterattack capability” to strike adversaries missile bases or command centres to enhance deterrence.
FNNMay64.728.6Japan to have an “enemy base attack capability” in order to respond to North Korean ballistic and other missile attacks.
NHKMay5529To cope with ballistic and other missile attacks, LDP has argued that Japan should possess a “counterattack capability” to destroy enemy missile bases, command and control facilities.
JijiJune60.919.2Japan to possess a “counterattack capability” (enemy base attack capability) as recently proposed by the LDP.
KyodoJuly51.624.8Japan possessing an “enemy base attack capability” (“counterattack capability”) for the purpose of self-defence to prevent missiles from being launched from an opponent’s territory.
AsahiJuly5040SDF possessing a “counterattack capability” to strike foreign missile bases that are trying to attack Japan.
KyodoSeptember53.538.4The government is considering possessing an “enemy base attack capability,” allowing Japan to attack missile bases and other facilities in other countries before Japan is attacked.
ANNOctober6027Given repeated missile launches by North Korea, the government is examining the option of the SDF possessing “counterattack capability” (The ability to attack enemy missile bases and other facilities).
FNNNovember62.130.1PM Kishida has stated that he will consider allowing the SDF to possess “counterattack capability” for the purpose of self-defence so Japan can attack enemy missile bases.
ANNNovember6129Given repeated missile launches by North Korea, government is examining the option of the SDF possessing “counterattack capability”. (The ability to attack enemy missile bases and other facilities).
YomiuriNovember5241Japan should have a “counterattack capability” to destroy enemy missile bases for the purpose of self-defence.
KyodoNovember60.835An expert panel proposed Japan possessing a “counterattack capability” so that the SDF can attack missile bases and other facilities in other countries to stop missile attacks.
NikkeiNovember6524PM Kishida is considering allowing the possession of a “counterattack capability” to strike adversaries missile bases or command centres in order to enhance deterrence.
FNNDecember60.632.9New security documents clearly state that the government should have a “counterattack capability” to attack enemy missile bases and other targets for the purpose of self-defence.
NHKDecember5531The Japanese government to change its policy to allow the SDF to possess a counterattack capability.
AsahiDecember5638The SDF should have the capability to strike a foreign country’s missile base or other facilities in the event that a foreign country attempts to attack Japan.
MainichiDecember5927The government has decided that the SDF can possess a “counterattack capability” to attack enemy missile bases and other targets.
KyodoDecember50.342.6To strengthen its defence capability Japan purchase Tomahawk cruise missiles for a “counterattack capability” for self-defence (SDF can destroy missile bases in other countries). 
NikkeiDecember6031In order to enhance deterrence, the government has decided to allow the SDF to possess a “counterattack capability” to strike missile bases and other targets.
JNNDecember5730The SDF to possess the ability to destroy an enemy’s missile launch sites and other facilities (“counterattack capability”) for the purpose of self-defense.

Pre-Election Key Indicators

Figure 1: Evaluation of Government COVID-19 Response (2020-2021)

Figure 2: Comparison of Suga, Kishida and Abe’s Initial Cabinet Approval (Disapproval) Ratings

Figure 3: 2017 House of Representatives (HoR) Party of Hope Voters 2016 House of Councillors (HoC) Proportional Representation (PR) Vote (2017 UTokyo-Asahi Voter Survey).

Figure 4: Constitutional Democratic Party of Japan Party Support: 2017 versus 2021

Figure 5: House of Representatives Proportional Representation (PR) Voting Intention (Monthly Average Various Media Agencies)

Figure 6: Who Would You Prefer to Gain More Seats at the Next House of Representatives Election?

(Average of Various Media Agency Surveys Leading up to Election)

Figure 7: Prime Minister ‘Premium’: Selected Dates 2001-2021

PM Premium calculated by adding difference between cabinet approval rate and party support rate and difference between cabinet disapproval rate and party support rate.

Figure 8: Majority Risk Index: Selected Dates 2001-2021

Majority Risk Index is found by adding net cabinet approval to the ruling party ‘offset’: the difference between the support rate for the ruling party and the main opposition party.

Figure 9: Continue Abe-Suga Policies? (October 2021 Surveys)

More Updated Election Indicators

Where are Ishiba and Koizumi? And a Third Coming of Abe?

I have a few words over at Tokyo Review regarding the LDP leadership race. In addition to supplementary data and figures (see below), I have a few extra words about who is not running.

Ishiba was apparently lining up the support of Nikai (47 members) at one point to add to his own 16-member faction. While broadly popular, Ishiba has lost all four times he has run (2008, 2012, 2018, 2020). He may be evaluating whether his influence on the last three races where his opponent has benefitted from strong anti-Ishiba sentiment within LDP parliamentarians. One of the major reasons for Suga’s victory last year was the ‘anyone-but-Ishiba’ movement engineered by Nikai Toshihiro where the four largest factions uniformly backed Suga, making his triumph a fait accompli. He may have gotten the hint that his party simply does not want him. He has added his voice to those supporting Kono’s appointment as the candidate most likely to protect the LDP current House of Representatives domination. Whether that will galvanize the faction heads to oppose Kono even more, or bring Kono support from Ishiba’s supporters in the party—especially among the faction-less—remains to be seen, especially now that Noda Seiko has entered the race.

Koizumi continues to bide his time and recently publicly backed Suga and looks likely to back Kono. Koizumi has enjoyed the brand enhancement that has come with the notable progress made by the Suga Cabinet on Japan’s more aggressive climate change commitments. If Kono were to win, Koizumi likely gets a cabinet promotion. This will give him further opportunities to enhance this brand as well as get more experience managing a more prominent and difficult portfolio to fend off criticism that he is more substance than style.

An Abe third coming has certainly been rumoured in Nagatachō, but the man himself has said little. Abe is instead engaged in a covert intra-party struggle with Secretary General Nikai Toshihiro over control over party funding, top party executive posts, and China policy. Ultimately, Abe may well be satisfied with exercising influence behind the scenes within the party for years to come, much as his grandfather Kishi Nobusuke did after he was deposed during the 1960 Mutual Security Treaty crisis. Abe is backing ideological fellow traveller Takaichi in this race, although if she does not make it to a second round run-off, Abe is likely to back Kishida as the candidate most likely to protect his legacy and preserve his influence going forward⁠—especially as Kishida will be sure to reduce the influence of intra-party rival Nikai.

Data and Figures

Updated Figures

Updated figures with polls only from September. In the first two graphs the vertical dotted line signifies the point when polls started to ask respondents about their support for official candidates only.

July 2021 Political Update

As we move towards both the LDP Presidential Elections (definitely September) and the House of Representatives election (before November), I will use this blog to post a few updates on polling trends. At this point, which contest comes first still seems to be a point of vigorous contention. Various personalities in both ruling parties have mentioned the possibility of delaying the lower house election until after the LDP leadership race. Deputy Prime Minister Asō Tarō yesterday even went as far as saying (jp) the possibility of an October national election was, in fact, ‘very high’.

The first two figures below are an update of how the public views the government’s COVID-19 response. The small bounce from the rapid improvement in vaccination numbers from mid-May into June appeared to be rather short-lived.

The next figure shows proportional representation vote intention at the next lower house election under Prime Minister Suga. Incremental deterioration of support for the LDP after a very positive start continues. While the recent Tokyo Metropolitan Assembly results might point to such a trend imperiling an LDP-Komeito majority in the national election, the continuing inability of the opposition to capitalize on deteriorating sentiment (the story of the last 9 years) remains a major mitigating factor. The point of contention at this point in time remains the size of LDP and Komeito’s majority.

The seeming lack of enthusiasm for Suga to stay on as prime minister past September’s LDP election (including amongst LDP voters themselves) remains a problem for the incumbent, however. With ambitious leaders waiting in the wings, and an internal party struggle over control of the LDP Secretary General role and party funds also playing out (jp) in the background, the situation remains volatile and unpredictable.

Finally, who could throw their hat into the ring should Suga be forced to step down or simply internal ambitions prevail? Both Kōno and Ishiba Shigeru remain ahead of other potential replacements in overall popularity. This dynamic will affect the first round of voting in the LDP leadership race assuming it is contested. The favourite throughout this year has been Kōno Tarō, but there does appear to be a hint of softening of support for the minister responsible for vaccination roll-out. Certainly remains in the discussion but he is not pulling away and creating some kind of fait accompli. Currently, it points to no candidate having distinguished themselves as an obvious candidate to rally around in order to displace the prime minister—inertia and caution in the party remain the current prime minister’s best friend for now.

Political Elements of Japan’s COVID-19 Response

First of all, here is a pre-print of an article written by Giulio Pugliese and myself on “Japan in 2020”. It is not short, essentially being a combined COVID-19 response review article together with a year in review article.

I was also asked to write something for East Asia Forum about the relationship between COVID-19, the Olympics, and Japan’s upcoming House of Representatives election that must be held by the end of October this year. Current reports are pointing to the beginning of September for the election if the government’s calculated risk of holding the Olympics (with spectators) turns out to be prescient. There is also an LDP election for president mandated for September, the winner of which is almost certainly going to become Prime Minister of Japan. The figures and commentary below are useful data points supporting assertions in that article, which you can read here.

This first figure shows the PCR Positivity Rate for Japan’s various PCR testing facilities throughout 2020 as reported by the MHLW.

This is essentially the number of individuals receiving a positive COVID-19 PCR test divided by the number of individuals tested (not the positives to test ratio as people are tested multiple times). A rolling seven-day mean is used to smooth out the day-to-day fluctuations in reporting and testing.

Knowing how much testing to do is difficult when testing everyone regularly is not an option. Especially at the outset of a pandemic when various systems have not been streamlined or stress-tested, there are labour and resource trade-offs that have to be considered. Also important is to avoid throngs of infected people crowding medical facilities where other vulnerable patients also congregate. It also seems the more one tests, the more one finds COVID-19 positives. The so-called R2 for testing to cases is often over 0.9 (the degree to which the number of positive individual cases is determined by variation in the number of PCR tested individuals).

Obviously, the cause of this is that testing is not a random sample, but self-selected or demand driven; nevertheless, it is still difficult to know what level of testing gives you at least a reasonable indication of the trends, even if not the exact prevalence of COVID-19 (which would require mass scale testing, which some countries have gone for). The WHO, however, suggests that a PCR Positivity Rate around 5 percent or under is an appropriate level and an indicator of sufficient testing, while anything greatly above that suggests insufficient testing likely to result in poor information and the underestimation of prevalence by orders of magnitude.

In Japan’s case, the above figure shows that for most of 2020 the PCR Positivity Rate was under or around the WHO indicator. It actually worsened following the announcement of the postponement of the Olympic Games. Testing began to rapidly scale up just prior to the announcement, but the curious thing is that, even though testing capacity and administration increased after Olympics postponement, so too did the PCR Positivity Rate. If Olympics postponement was the ultimate cause of increased testing, and therefore the increase in confirmed cases, we should have seen the exact opposite trend. In essence, the demand for testing increased as symptoms resembling COVID-19 also increased rapidly around that time. Researchers subsequently identified the ‘European Strain’ as the culprit of the big jump in infections starting in March through to July, suggesting it was the government’s failure to lockdown the borders to Europe and the US, rather than undertesting that let the original ‘Wuhan Strain’ to proliferate undercover, that was the driver of the first state of emergency in May 2020. This failure of course is still a political driver—connected as it was to the Olympics and diplomatic considerations. But such considerations are not a good explanation for (purposeful) undertesting (especially as it would be self-defeating in the long-run). In any respect, antibody testing in June and December 2020 lent credence to the claim that COVID-19 just simply was not that prevalent in Japan during 2020.

Political Evaluation of Government Responsiveness During the COVID-19 Pandemic

Subjectively, the story is quite different. The following figures relate to the Japanese public’s evaluation of the central government’s COVID-19 response as well as general political sentiment during the pandemic crisis up until June 21, 2021. This is of course something different from the evaluation of Japan’s overall response, where a number of different actors, including local governments, medical authorities, and the public itself, were involved. In the first figure, we see low public evaluation of the personal leadership shown by both Prime Ministers Abe and Suga during 2020 according to surveys conducted by both the right-leaning Yomiuri Shimbun and left-leaning Asahi Shimbun.

This next two data points were constructed using survey data from 7 major media companies’ regular surveys. All organizations asked virtually identical questions about how respondents rated the central government’s overall COVID-19 response. The sources were the Asahi Shimbun, Kyodo News, Jiji News, Yomiuri Shimbun, Nikkei Shimbun, Mainichi Shimbun, and NHK. I considered Sankei/FNN surveys for other items in 2021, but due to a fake survey scandal, the data from 2020 is unusable. The company took a break from surveying anyhow.

Beyond evaluation of government response, the media companies asked a variety of other questions.

The first figure below shows that, perhaps unlike a lot of overseas jurisdictions battling COVID-19, there was little pressure on the Japanese government to open up economically in 2020—if anything, quite the opposite. Of course, had there been more severe lockdowns like elsewhere, the outlook may have changed, but we will never know.

This next figure shows a rare drop in LDP party vote intention in the next House of Representatives election of about 10 percent (Yomiuri and the Asahi). This is not of a scale that would imperil an LDP majority at a House of Representatives election, especially as the LDP still retains overwhelming advantages in the small member districts due to its strategic voting relationship with Komeito and the vote-value disparity advantage for rural electorates. Nevertheless, if this trend continued through to the next election, it would cost some seats and is the sort of thing intra-party rivals could point to in any subsequent leadership contest, say one due to take place in September 2021.

The next two figures show that the public is currently not too enamored with Prime Minister Suga and that at this point they would not mind seeing the LDP go in another direction in September’s leadership contest. Of course, there is the small matter of the Olympics and a national election to traverse. A ‘safe and secure’ Olympics and policy proposals that interest the public in the general election, such as a minimum hourly wage boost to ¥1000 that would also apply to rural areas, could still give Suga a big boost that might allow him to unite the party behind him.

If the calculated risk of holding the Olympics turns bad, however, the House of Representatives election can still be delayed until October to give the government time to deal with whatever problems arise. However, given Suga’s already precarious position, the LDP would also need this time to select a new leader ahead of the House of Representatives election as Suga stands down to ‘take responsibility’ on the LDP’s behalf.

Japan’s Foreign Territory Strike Missions: Technical and Tactical Debate

I recently published something at the Tokyo Review that looks at broader strategic considerations pertinent to Japan permitting the Self-Defense Forces to conduct long-range strike missions against ballistic missile positions in foreign territory (North Korea). However, I felt the technical and tactical details were also important to consider when evaluating the possibility that Japan might explicitly allow its military to exercise the use of force inside the territory of another country for the first time since World War II.

Starting with weapon systems, for many years the Tomahawk Land Attack Cruise Missile has been the main focus of Japanese hawks who believe Japan should be able to project power into the territories of other countries. On paper, the Tomahawk addresses Japan’s military need for a precise, limited but effective at-launch strike option against North Korean missile positions. The Tomahawk is battle-proven and accurate, has a small to medium radar cross-section as it skims along the surface, and its turbofan propulsion means that little heat is emitted for infrared detection. Requiring only minor modification of the Mk41 VLS already onboard Japan’s so-called ‘Aegis’ destroyers, ship-launched Tomahawks could enable long-range strikes against North Korean missile positions without putting the JSDF in harm’s way like fighter-launched strikes would. Furthermore, a Tomahawk ($1.8 million for Block IV) is considerably cheaper than a SM-3 Block IIA BMD interceptor ($18.4 million).

Not everyone—including former defense ministers—is convinced of the immediate need for Japan to field Tomahawks for ‘missile interdiction’ missions, however. Despite his past advocacy, even renown military otaku (and former defense minister) Ishiba Shigeru is questioning the contemporary wisdom of Japan pursuing long-range strike capabilities for missile defense missions. As Ishiba, current defense minister Kono Taro, and most commentators recognize, missile acquisition would be the easy and relatively inexpensive part of any autonomous Japanese strike capability. However, an increasingly contested East Asian aerial and maritime environment where the People’s Republic of China (PRC) and North Korea are fielding solid-fueled intermediate range ballistic missiles launchable from tracked TELS (maneuverable, multi-terrain and reusable mobile launch platforms), greatly complicates the effectiveness of pre-launch long-range strike missions. Japan’s East Asian adversaries now require fewer support vehicles and less preparation time to launch ballistic missiles (as little as five minutes), have greater choice of launch location, and can take advantage of concealment. Even if launch locations are detected, the ‘mere’ subsonic speed of the Tomahawk also gives well organized adversaries time to react and intercept, especially if launched from the Tomahawk’s supposedly ‘safer’ outer ranges.

The Japanese government will therefore need to purchase and acquire familiarity with large numbers of high-end, high maintenance and expensive support systems to have even partial success. These support systems include strategic early warning satellites, ISR-enabled airborne systems to identify, locate, and track targets once warned, the ability to suppress enemy air defences and the destruction of radar installations, robust cyber and electromagnetic capabilities (both offensive and defensive), the ability to intercept and interfere with enemy communications, and the ability to sustain firepower beyond a single decisive attack.

Japan has already started (E-767 AWACS, E-2D Advanced Hawkeye, KC-46A in-air refuelling tanker aircraft) or thought about acquiring systems (EA-18G Growler, Global Hawk) that can perform useful roles in such missions. Much greater unit numbers will be needed than the SDF currently possesses, however, as well as more UAVs and loitering munitions. This is due to distribution, disaggregation and redundancy requirements in a contested air environment. Assuming Japan pays the initial outlays or guarantees access to all of the necessary equipment, simply maintaining this all could still come to about US$1 billion dollars a year.  

Furthermore, these support systems will still need to operate in these contested air and maritime domains while trying to identify and track North Korean missile positions. Trade-offs in terms of operational proximity to the adversary’s territory, effectiveness, and safety will remain an issue. Base hardening and the building of underground tunnels, hangars and military facilities by North Korea (and the PRC) also further complicates the likelihood of success. Support systems or not, cruise missile-enabled strike is not going to be very useful against submarine-launched missiles.

Experts are ultimately sceptical that pre-empting a missile attack is a viable mission, and even a ‘Nodong hunt’ to remove threats after an initial attack will likely be only partially successful—at best. This strike capability will certainly not be the silver bullet some proponents advertise to the public. Jimbo Ken outlines the most plausible scenario for Japan attacking North Korean missile positions: a combination of layered BMD plus long-range strike on overseas military assets playing some role in limiting damage to Japanese and American bases after an initial saturation attack. Takahashi Sugio also notes potential value for Japan where the United States decides to prioritize missiles targeted at US-ROK facilities in South Korea.

However, even for missile defence-focused mitigation there is a priority hierarchy. First, traditional interceptor-based BMD is essential to minimize as much damage as possible from an initial saturation attack on Japanese and/or American positions in Japan. Second, dispersing bases, airfields, and ports, wider distribution (deceptively, if need be) of units, equipment, fuel and ammunition, and in general building in resilience or survivability to US-Japan military posture on the Japanese archipelago, are all essential on the assumption that some missiles—or perhaps many if Aegis Ashore remains cancelled—will make it through the midcourse and terminal layers of alliance BMD. With the former two priorities addressed, and the alliance still able to function, a Japanese long-range strike capability could then contribute to suppressant effects by disrupting the rate and simultaneity of subsequent attacks, as well as decrease the raw number of attack nodes capable of striking military positions in Japan. This would in turn increase the possibility of interception.

For pure geographical reasons, Japan has the most important, alliance-enhancing role to play for the first two priorities, while the third would be an incredibly taxing mission to undertake from Japan. The United States is by far better positioned to address the third priority (cooperating with the Republic of Korea) due to its ROK operational bases and better chances to establish air dominance over the Korean Peninsula, as well as the capability to inflict greater damage on North Korea. The full costs of implementing a robust and credible long-range strike capability for ‘missile interdiction’ (ミサイル阻止) need to be considered alongside the fact that such missions are likely to be exercises in mitigation and may not be the best use of resources to support the alliance.

Foreign territory strike also raises some politically tricky questions regarding command-and-control structures for Japan, and not only with the United States. The US-Japan alliance does not currently have a combined command system to be activated during wartime like the US-ROK alliance does. The relationship with South Korea will also have to be enhanced in advance, as the United States will not want the burden of connecting disparate command and control structures during wartime. Realistically, for Japan to operate effectively in this mission, it will also need to conduct these activities from ROK bases on the Korean peninsula.

That’s a big problem. The US-Japan alliance provides essential strategic depth for the US-ROK alliance, and the US-ROK alliance could provide essential intelligence and real time information on North Korean activities as well as a forward operating launchpad for Japan during Peninsula contingencies. However, neither side appears to be willing to turn down the heat on pointed historical issues, let alone conceptualize the strategic relationship in such a constructive way—recent history suggests it may even be counterproductive to hector them into doing so.

These are all issues that the Japanese government should thoroughly consider when deciding on whether to make foreign territory strike an explicit JSDF mission, in addition to opportunity cost and budgetary limitations. The above is not an argument against Japan acquiring any long-range or ‘stand-off’ missiles. Japan is, after all, already acquiring air-launched cruise missiles with considerable ranges and developing hypersonic missiles that take advantage of boost-glide trajectories. Most notably, while lacking the Tomahawk’s range, Japan’s incoming JASSM-ER is faster, stealthier, and cheaper ($1 million unit cost). Future iterations will be capable of delivering more effective bunker busting warheads as well as Electronic Warfare packages like CHAMP.

While such missile systems could be repurposed for hitting fixed military sites in foreign territory, their main value is, however, in allowing the JSDF to operate outside or at the outer limits of the enemy’s threat envelope in anti-air or anti-ship operations or in aid of remote island defense in the East China Sea and Japan’s southwest. In any respect, when considering the purchase of new missiles like the Tomahawk or repurposing current missiles, a fuller accounting of the barriers to mission effectiveness should be undertaken before making a policy determination and investing precious resources in fleshing out new systems and missions.

Keio’s Jimbo Ken is particularly blunt on this point:

If Japan doesn’t make clear the strategic rationale for allocating scarce resources to specific security or defense priorities, there is a major risk of Japan implementing an inefficient or half-baked defense posture. For example, by introducing a strategically meaningless level of long-range strike capability while leaving the defense budget essentially as it is, the MSDF’s Aegis destroyers will be (over)burdened with missile defense duties [both missile interception and foreign country strike]. This will ultimately result in the neglect of the defense of the southwest maritime regions around Japan. It is thus easy to see how this might lead to the collapse of Japan’s overall defense portfolio.

What if this is not really about missile defense?

What if the current discussion about foreign territory strike is not really about missile defense/interdiction focused on North Korea, but something much more ambitious? After all, the alliance provides opportunities for the JSDF to access and familiarize itself with various high-end capabilities that could be a path to autonomy or a hedge against abandonment. A precautionary hedge is smart, but it still requires balancing the acquisition of high-end capabilities with continuing to do well what you do currently as an alliance contribution. Japan needs to consider what being a good ally entails more than what would be the case if military autonomy was inevitable and/or the motivating desire for Tokyo’s military enhancements. Otherwise, hedging against abandonment might become self-fulfilling. Furthermore—and at least for the time being—not all Japanese defense experts or even conservative politicians are so sensitive to idea that abandonment is imminent and perceive foregoing strike capability as a lost opportunity to hedge.

Another possible foreign territory strike mission would be targeting PRC coastal bases, ports, airfields, ammunition depots, chain-of-command systems and other fixed facilities to stunt a Chinese response at the start of a PRC-Japan maritime conflagration around the Senkaku Islands. As a pre-emptive measure, this would raise a whole lot of questions of escalation that seem absurd to consider outside of the US-Japan alliance context. In the alliance context, Japan could plausibly offer help striking fixed coastal assets where the United States might be stretched given China’s own military build-up. This, however, does little to relieve the more intense burdens the United States would have to carry in such a contingency, such as hunting mobile missile platforms or establishing maritime control/air dominance versus a Chinese posture designed to prevent just that. Enhancing Japan’s military posture and the alliance through other means, including spending more on maritime and area denial capabilities, and finding an alternative plan for Aegis Ashore and the already purchased, US$300 million SPY-7(V)1 radar, will free up the United States Navy to an even greater degree than a limited Japanese strike capability would. 

A final possibility is that the Japanese government is accelerating the move from a posture based on deterrence by denial to one based on deterrence by punishment. The former is the effective communication of the willingness, ability, and preparation to prevent an opponent from achieving their immediate military goals, thus deterring the initiation of hostilities. Deterrence by punishment is the effective communication of the willingness, ability, and preparation to exact an unacceptable cost on the opponent that deters them from initiating hostilities. This might be through attacks on the political regime that weakens or removes it, destruction of both military and socio-economic infrastructure, or even national annihilation. A military posture based on deterrence by punishment would truly be a complete departure from Japan’s post-war concept of defensive defense. It is also a departure that public discussion has not touched upon. Transparency issues aside, one would have to question both the credibility and advisability of committing to conventional long-range strikes to deter nuclear-armed enemies by threatening them with unacceptable punishment. Particularly when those enemies have a much higher threshold for civilian casualties, and are probably not all that sensitive to Japanese civilian well-being given their own historic suffering at the hands of Japanese military forces and contemporary state propaganda.

What Happened to Japan’s Influenza Season?

Being in Japan right now and seeing the crisis footing of many countries around the world can feel a bit surreal. How can it be that Japan has the number of cases it does, the population density it does, and the government has not until now felt the need to declare a state of emergency or completely shut down its borders? This is especially true if you live in the Kantō area,* where even a reduction of 20 to 30 percent of people on the street or passengers in trains can still seem…crowded.

Of course, (most) members of the public have engaged in ‘behavioural modifications’ and the government has pursued ‘non-pharmaceutical interventions’ (NPIs) such as requesting school closures and the cancellation of big, crowded events. It is easy to wonder, nevertheless, whether self-restraint in terms of going out (ie. the ubiquitous calls for jishuku), physical distancing, avoiding closed, poorly ventilated spaces, engaging in proper ‘cough etiquette’, or other hygeinic measures to avoid contacting or transmitting SARS-CoV-2, are actually having any effect? The recent dramatic uptick in COVID-19 cases in Tokyo would seem to suggest that discipline and vigilance dropped mid-March, and more drastic measures will be needed to avoid overwhelming medical facilities.

Pertinent to this issue of overburdening medical facilities—and possibly a direct outcome of aforementioned ‘behaviour modification’ and NPIs—is the fate of Japan’s influenza season. To look at this, we can use Japan’s Ministry of Health, Labour and Welfare (MHLW) regular in-season influenza tracking reports. Using 5000 fixed ‘sentinel’ medical institutions to track weekly reported cases of influenza, and 500 fixed in-patient core medical facilities to track hospitalizations, the Japanese government can get a good sense of whether the influenza season is tracking better or worse than previous seasons without having to do a comprehensive, nation-wide survey.

And, at the beginning of 2020, it looked like the 2019-2020 influenza season was going to be brutal. It started earlier than usual in September, with 7 times the number of hospitalizations at the core medical facilities than previous two seasons for that month. Hospitalization cases in December were still 3-4 times higher than the previous two seasons. This is something, because influenza-induced deaths have been rapidly trending upwards since 2010 according to MHLW annual statistics. The previous two seasons each resulted in national totals of 3308 deaths (2017-2018) and 3,267 deaths (2018-2019 season) from influenza directly or influenza-induced pneumonia. The 2018 calendar year was the deadliest year since 1970 for influenza fatalities. This year was on track to be even worse.

Then, suddenly, it was not to be.

Weekly observations at the 5000 fixed sentinels never reached the heights of previous seasons. The comparatively very bad December ended up being the peak of this year’s influenza season based on sentinel reports. The figure below taken from MHLW regular influenza reports tracks the peaks and troughs of the respective 2017/18, 2018/19 and 2019/20 influenza seasons. The y-axis is the number of reported cases per sentinel per week, while the x-axis is the week of the year. 

Observations Influenza

A very similar trend can be seen when I collated the monthly totals for hospitalizations at core medical institutions, an arguably more concrete measure of influenza impact.

Japan Influenza Hospitalisations by Month (MHLW Weekly Reports)

Like sentinel observations, influenza hospitalizations dramatically dropped after December. While December’s hospitalizations were four times up on the previous season, January had only just over 1/3rd the hospitalizations of the previous January. In February 2020, hospitalizations were 31 percent of the 2018/2019 season, and 19 percent of the 2017/2018 season. March was even kinder, registering 20 percent and 10 percent of hospitalizations, respectively, for influenza compared to previous years. Despite the earlier start to the season, 2019/20 hospitalizations for influenza at the 500 core institutions will register only 2/3rds of the previous two seasons.

To be sure, by the start of 2020 there was already significant concern about how deadly this influenza season was going to be. The MHLW had raised its alert level. Many classes and class years in Japanese schools were being closed to contain influenza. Students were told to dispense with any gaman (perseverance) and to stay home even if they had the slightest influenza symptoms. Then Japan got its first COVID-19 case in mid-January.

Perhaps, I wondered, hospitalizations were driven down by the moderately influenza-afflicted avoiding hospitals or being told to avoid hospitals given the risks of contracting the novel coronavirus? However, when one looks at the ‘serious’ hospitalizations from influenza requiring special care—admission to the ICU, the use of artificial respirators, or the conduct of a CT scan, MRI or EEG—a similar pattern emerges.

ICU Influenza Admissions
Artificial Respirator Use Necessary
CT, MRI, EEG Scans

For each of these categories, the December numbers were at least three times higher than those for the 2017/2018 and 2018/2019 seasons. Then, in the new year, the numbers dropped off rapidly. This does not suggest any kind of rationing in favour of serious cases versus moderate cases. From this data, it seems Japan as a whole had one of the lightest influenza seasons for many years with total hospitalizations in the 2019/2020 season coming to 66 percent of the previous season’s total, and hospitalizations requiring special attention coming to approximately 78 percent of the previous season (despite the very high early numbers).

Another Japanese rapid notification system using indicative measures is administered by the National Institute of Infectious Diseases. It uses notifications from 21 major Japanese urban jurisdictions of deaths from influenza and pneumonia. Rather than simply reporting raw numbers, the collated reports are measured against an expected mortality rate baseline modeled on factors including the presence of underlying diseases and the aging of the population.  ‘Excess mortality’ from influenza and/or pneumonia takes place when actual deaths exceeds a threshold—defined as the upper limit of the 95% confidence interval for the baseline. The figure below shows that in December 2019, actual influenza-related and pneumonia deaths (blue line) were at the excess mortality threshold (pink). However, by the end of January we start to see a consistent downward trend towards the baseline (green line), and then a big plunge around the start of March as many jurisdictions did not report any deaths.

influenza pneumonia deaths

Regrettably, as can be seen below, Tokyo is letting the team down (again), with reports of excess mortality at the end of the year, and the end of February.

tokyojinsoku

Like the MHLW weekly influenza reports, this data is more indicative rather than complete, and the MHLW statistics yearbook will have the final word on influenza and pneumonia deaths nationwide. From what I can tell, even the 2019 data will not be published until the middle of this year.

Nevertheless, there is good reason to believe that Japan as a whole avoided a high death rate from influenza as the COVID-19 crisis continues to unfold.**

Perhaps the most striking result, however, came from when I broke down the number of hospitalizations into age groups. Unlike COVID-19, influenza and influenza-induced pneumonia are equal opportunity menaces in that both the elderly as well as infants and younger children are likely to be particularly susceptible. Interestingly, if we look at the statistics for hospitalizations, under 60 age cohorts were hospitalized in similar numbers to the previous two years, but hospitalizations of those over 60 were down considerably.

Japan Influenza Serious Hospitalisations by Age Group

Furthermore, influenza hospitalizations requiring greater attention and care (ICU admission, use of artificial respirators, and the need for various scans) revealed similar year-to-year patterns.

Influenza ICU Admissions by Age Group
Artificial Respirator Use Necessary by Age Group
CT, MRI, EEG Scans by Age Group

In fact, this year’s influenza looks like the reverse image of COVID-19 in terms of the group it most afflicts. Why is this? I am not really sure, although the answer could be as simple as people being more careful around older people and giving them greater leeway in public (what I’d like to believe), although it is more likely that over 60s were the most ardent practitioners of self-restraint and seldom left home out of fear of COVID-19.

I collated these statistics mainly for my own interest as a proxy indicator of whether behavioural modification was actually taking place in the Kantō region of Japan. Deciding which non-pharmaceutical intervention was the most effective at mercifully cutting short the influenza season is better left to the experts; but whatever it was, given the possibility of much stricter limitations under a state of emergency declaration for the Kantō region, we can take some comfort from the thought that the two months of on-and-off jishuku and other behavioural modifications was not for naught. It is likely many lives have been spared from both COVID-19 and influenza.


*The Kantō region is the Tokyo Metropolitan Government plus the surrounding prefectures. Population of around 40 million people.

**There was also no major increase in the units of influenza vaccines procured in advance for the 2019/2020 season (see this MHLW document, p.3). Since the early part of the season was particularly bad, and younger age groups suffered as much as usual, enhanced vaccine effectiveness for this year is unlikely to be an explanation.

***For those interested, early season reports found Influenza-B to be most common at 59%, the ‘new influenza’ from the 2009 season (AH1pdm09) registered 39%、and Influenza-A (subtype H3N2) was 2%.