引っ越しタイム, I’m moving my blog

For anyone that’s watching, I am moving my wordpress.com hosted blog to a self-hosted one (using wordpress of course). A few reasons: One, more freedom to play around with the website design. Two, I can make my blog bilingual to increase my readership (I’ll probably shorten the content since I will be even more pressed for time) Three, ads! Four, it’ll be good practice translating back and forth. Five, to expand my blog neta repository. Anyways, the following are the links:

Achikule! or あちくれ！

I’ll probably nuke this site once I’m done with my official one.

Kuchikomi Prediction Analyses 2009 Japan Lower House Elections

As promised in a prior post, I’ve finished the data entry for the Kuchikomi (クチコミ総選挙) election predictions. The Kuchikomi election prediction algorithm was created by hottolink (ホットリンク) and it is based on internet chatter (口コミ). The algorithm parses through information on the internet from mass media to blogs on individual politicians and parties to produce a prediction percentage based on historical data. You can I guess think it as a souped up Google AdPlanner:

候補者個人に関するクチコミと所属政党に関するクチコミを集め、それぞれがどの程度得票率に影響を与えるかを過去の国政選挙をもとに分析し、予測モデルを構築して算出した値が予想得票率です。

In terms of very general results, Kuchikomi had about 80% of the single district seats correct (mind you, they didn’t calculate the proportional representation seat — a whole another issue). Both Asahi and Nikkei newspapers predicted more than 94% of the single district seats correct. But considering Kuchikomi is still new technology, 80% isn’t too bad. But as you’ll see, there are some major weaknesses.

Party Bias

First of all, Kuchikomi did really bad for minor parties (namely: YP = Your Party (みんなの党), SDP = Social Democratic Party (社民党), PNP = People’s New Party (国民新党)) and independents (= I). See the graph below as I plotted the predicted vote percentage vs the actual vote percentage of each candidate separated by party:

A perfect prediction will show a one-to-one lines. But for the aforementioned minor parties and independents, we see that Kuchikomi had very low predictions, yet resulted in considerably higher numbers (don’t worry about the mysterious acronyms, some of them are uber-minor parties). In the description of their algorithm, they mentioned that they gather information for both the candidate and the party. But it looks like they put more weight into on the party. For example, Your Party has former LDP members that were fairly well-known before the election. But because Your Party only formed as a political party three weeks before the election, Kuchikomi failed to gain sufficient information; thus, the low predictions. The leader (Watanabe Yoshimi) was predicted highly, but they had only two candidates in that district. Kamei Shizuka, the leader of the People’s New Party, was predicted to have 15% of the vote, but he actually got 60% — a semi-educated human being would have predicted that much better. In a similar vein, Kuchikomi just sucks for Independents. Either the media just doesn’t like to talk about them, or again, Kuchikomi doesn’t weigh the individual as much as the party.

Let’s look a little closer at the two main players: DPJ and LDP. The following graph is basically a mash up of the previous graph, where the red dots indicate the DPJ and the blue ones as LDP:

A one-to-one black line is passed through to indicate anything above it constitute underestimated predictions and anything below are overestimated predictions. The DPJ estimates are generally underestimated, and they are underestimated more strongly for lower prediction percentages. In general, the variance of the actual vote percentage is higher for lower prediction percentages. A paradox that happens here, is that even if Kuchikomi is underestimating the DPJ results in terms of percentages, it actually overestimated the number of seats that the DPJ won.

So what is going on? Well, it is hard to tell. One would think that underestimating the percentages would lead to underestimating the number of seats. This parallel can get broken if the prediction screws up (overestimates) big for a few seats, while underestimating moderately. I decided to plot the movements (difference between prediction percentage and actual percentage) on seats that were won by either the LDP or the DPJ and was predicted to win by either the LDP or the DPJ. Of course there are 4 cases: predict LDP -> observe LDP (upper right), predict LDP -> observe DPJ (upper left), predict DPJ -> observe DPJ (bottom left) and predict DPJ -> observe LDP (bottom right):

The red dots show the predicted percentage of both the DPJ and the LDP. The arrow protruding from the red dot “moves” to the actual vote percentage. So if the arrows going straight up, the prediction was underestimated for the LDP, and if the arrow is going to the right, the prediction was underestimated for the DPJ (and vice-versa). We see the red dots very close to black one-to-one line compared to the actual votes. This indicates that the predictions are a lot more flat than the actual results maybe due to lack of information. We see the arrows on the bottom right graph (predicted DPJ seat but actually went to the LDP) going vertically straight up. This of course means that the DPJ prediction was fine, but it fucked up on LDP results big time.

Now who were the candidates for the bottom right graph? Here is the list:

Aomori 2    “Eto Akinori”
Aomori 3    “Ooshima Tadamori”
Aomori 4    “Kimura Taro”
Chiba11     “Mori Eisuke”
Chiba12     “Hamada Yasukazu”
Ehime 1     “Shiozaki Yasuhisa”
Ehime 4     “Yamamoto Kouichi”
Fukui 1     “Inada Tomomi”
Fukui 2     “Yamamoto Taku”
Fukui 3     “Takagi Tsuyoshi”
Fukuoka 7   “Koga Makoto”
Gifu 2      “Tanahashi Yasufumi”
Gunma 4     “Fukuda Yasuo”
Hiroshima 1 “Kishida Fumio”
Hokkaido 7  “Itou Yoshitaka”
Ibaraki 4   “Kajiyama Hiroshi”
Ishikawa 2  “Mori Yoshirou”
Kagoshima 2 “Tokuda Takeshi”
Kagoshima 4 “Ozato Yasuhiro”
Kagoshima 5 “Moriyama Hiroshi”
Kanagawa 2  “Suga Yoshihide”
Kanagawa11  “Koizumi Shinjiro”
Kanagawa15  “Kouno Taro”
Kouchi 1    “Fukui Teru”
Kouchi 2    “Nakatani Gen”
Kouchi 3    “Yamamoto Yuuji”
Kumamoto 3  “Sakamoto Tetsushi”
Kyoto 5     “Tanigaki Sadakazu”
Mie 5       “Mitsuya Norio”
Miyazaki 2  “Etou Taku”
Nara 4      “Tanose Ryoutarou”
Okayama 1   “Aisawa Ichirou”
Okayama 5   “Katou Katsunobu”
Tokushima 3 “Gotouda Masazumi”
Tokyo17     “Hirasawa Katsuei”
Tottori 2   “Akazawa Ryousei”
Wakayama 3  “Nikai Toshihiro”
Yamaguchi 1 “Koumura Masahiko”

I’m not sure anything general can be derived from this list. I do however see former Prime Ministers (Mori, Fukuda, Koizumi (well, his daddy..)) and pols that appear on TV frequently (Kouno, Hirasawa, Gotouda) in this list. This might indicate a big-name bias that can’t be detected from more neutral mass-media articles (although I would think blogs would give some of this info). It could also mean that the LDP really pushed certain candidates at its last rush (many of the big-names also hold high positions).

How about the graph on the top right? The graph shows again the same movement: no bias for the DPJ but large underestimation for the LDP.

Fukuoka8   “Asou Tarou”
Gifu4      “Kaneko Kazuyoshi”
Shimane1   “Hosoda Hiroyuki”
Tochigi5   “Motegi Toshimitsu”
Tottori1   “Ishiba Shigeru”
Yamaguchi3 “Kawamura Takeo”
Yamaguchi4 “Abe Shinzou”

Former PMs (Abe, Asou), TV personas (Ishiba) and pols that made frequent media appearances (Kawamura, Hosoda).

Geo Effects

Another thing that I looked at, were prediction performances according to each prefecture and district. Each district has their own color that may influence media and blog-related chatter that goes onto the internet.

I first created an error measure as follows:

This basically gives me a proxy for how “bad” the prediction went. I put an exponential simply because I wanted to penalize mistakes at higher prediction percentages. The squared difference is to penalize large mistakes. What I did was calculate this measure for each candidate of each district. Then I can take aggregate statistics to explore how bad the predictions went given the location and the candidate.

In this case, I simply took the mean across districts and the candidates to arrive at a Prefecture-wide value. I can plot this (thanks Prof. Aoki (青木繁伸)) on a Japanese map to neatly see this information at a glance:

This basically shows poorly preforming predictions as a darker color. Is there some sort of pattern here? Again, it is hard to see. The error measure that I picked was fairly arbitrary, and I take the mean a couple times anyways — creating a major loss of detail. One thing though, Miyazaki had easily the largest error among the Prefectures. Can this be caused by media-crowding? Meaning, because the Governor caused a huge ruckus, maybe information let out about the election were superfluous.

A nice thing about these measures, is that we can play with them easily by making correlations with extraneous variables. Because I took Prefecture-wide statistics, I can regress it by, say Prefecture-wide population statistics:

We obviously see something interesting here. The Prefectures with high population has contain their error a lot more efficiently than the country lands. An error measure higher than 0.5 were observed at low population Prefectures (Akita, Gifu, Kagoshima, Kumamoto, Miyazaki, Okinawa, Ooita, Shimane, Tochigi, Toyama, Yamagata) except for Saitama. Can this be called the urban bias? I am guessing this can happen because there are more media outlets, a larger educated populace and a stronger dependence to government policies for populated areas.

Of course this is all very general, but I decided to stop here since my lower back is hurting. But it is easy to see that further analyses can be done.

Last but not least: Gender

A quickie on prediction performance according to gender. Here is a graph that plots actual vote percentage minus predicted vote percentage, separated into the DPJ and the LDP.

The female politicians going above the zero line means that the predictions were underestimated. This is definitely a good thing since Kuchikomi has stated that they use regression results from past elections to get their predictions — meaning, this is slight evidence that people are more likely to vote woman than in the past. Then again, the LDP women predictions were spot on, so it might just be a DPJ thing. The women in general didn’t perform as well as the men .

ah, So?

Aside from being a statistical exercise, what was the reason for doing this? Once the election is over, who cares about these retrospective-I-predicted-better-than-you-chest-pumping?

The obvious benefit is that we learn a few things about the real world. Any prediction is at best, abstract extrapolations of the real world, and based on how things are setup, it’s a good learning process to articulate what is happening. In this case, Kuchikomi data was based on information from the internet. If there are such biases that I mentioned (flattening, urban, big-name, gender, party, etc.), even though it might be bad in terms of prediction performance, these imperfections tell us something about the elections.

The most important imperfection, IMO, was that the LDP was underestimated by a fair amount (not that the DPJ was underestimated). This is somewhat surprising considering that Kuchikomi was using much of the data from historical elections where it was always the case that the LDP won. For whatever reason, certain LDP politicians just did way better in real life. I called it big-name bias, but it could also mean that the Regime Change (政権交代) meme was played so much in the media that Kuchikomi bought into the hype. Conversely, it could mean that Kuchikomi got it right, but the media overplaying the Regime Change meme influenced voters more than anything online. These tidbits you would not learn from opinion polls because opinion polls grab data from real people instead of information from the internet.

Another reason we should care about predictions is for data checking. Recently, Nate Silver of FiveThirtyEight accused Strategic Vision (a polling company) of having some shoddy, suspicious polling (Nate has huge balls doing this alone). Predictions can be applied to different datasets, and it usually outputs a general trend. When one results in a departure from the trend, it gives some evidence to look into information rigging. This came helpful in the recent Iranian election.

Anyways, for me, the biggest coup was my new ability to graph Japan in R. I also have maps for the districts (but I can’t show both the whole country and the districts at the same time); thus, now I can create pretty pictures on 県民性.

Feel free to contact me if interested in the data, code or the map.

If Yanba Dam (八ツ場ダム) is constructed, at least make it friendly to the Fish? part 1

The DPJ and the decision of whether to continue building Yanba Dam (八ツ場ダム) has been up to a lot of debate. Yanba Dam is a huge multi-purpose dam that cuts through the Agatsuma River (吾妻川). Proponents say the dam needs to be built for irrigation, water supply, electricity generation, flood control and the cost would be high if it were to pull out. Opponents say the dam will destroy the environment, overflow the infamous hot springs (don’t mess with the Japanese on hot springs!), force about 1100 residents to relocate, money will flow into non-local bureaucrats and the costs will be high. In any case, if there is there is going to be one, the purpose of making a dam should be very well-articulated — cuz there are some shitty ones out there. Specifically, what will they do for the salmon runs?

Dam It Japan

Japan is of course famous for being dam happy. The country contains 2675 dams and that is good for number 4 on the list. This isn’t too surprising given the love for public works projects, their large economy and the government’s disregard for the environment. What is surprising is the number of dams given the small area of the country:

Top 7 countries wrt # of dams
Country / Dams per km² x10⁵ / # of Dams / Area (km²)

1. Korea / 76.75 / 765 / 99,678
2. Japan / 70.78 / 2675 / 377,930
3. Spain / 23.64 / 1196 / 505,992
4. China / 22.82 / ~22000 / 9,639,688
5. India / 13.05 / 4291 / 3,287,263
6. U.S. / 6.83 / 6575 / 9,629,091
7. Canada / 0.79 / 793 / 9,984,670

Japan has a threefold higher dam per area rate than the next one on the list.  South Korea has even a higher rate, which presumably doesn’t include the huge one that restricts the flow of North Korean runs.

One of the reasons dams are so often needed in short intervals is because of the weird spiky geography. In addition, water storage is difficult; the combined water storage capacity of Japan’s dams is actually less than that of Hoover Dam.

Blocking Salmons’ Arterial

In terms of salmon, Japan accounts for about a third of the wild harvest in the Pacific Rim (British Columbia, Washington, Oregon, California, Hawaii, Russia, etc.). However, even though Japan has a lot of rivers, 98% of them that contain salmon are dammed (ibid.). We see courtesy of Dam Maps how crowded the rivers are:

0624 is Yanba Dam

Dam construction (along with other factors) can block salmon migration and spawning in a multitude of ways:

Problem	Dams	Logg.	Changing Climate	Agri.	Pop. Growth	Over Fishing	Other
Salmon Migration Barriers	Big Time	no	no	no	no	no	no
Barriers to Reaching Spawing Grounds	Yes	no	Yes	no	no	Yes!	no
Reduced Number of Young Salmon reaching sea	yes	yes	no	no	no	no	bird pred.
Pesticide Exposure	no	Yes	no	Yes	no	no	no
Industrial Pollutants Exposure	no	no	no	no	Yes	no	no
Loss of Streamside Vegetation and Functions	no	Yes	no	Yes	Yes	no	no
Increased Sediment Load	no	Yes	Yes	Yes	Yes	no	no
Natural Habitat Destruction	yes	yes	no	yes	yes	no	no
Reduced Fresh Water Flow in Rivers and Streams	yes	no	no	yes	yes	no	no
Abnormal water temperatures	yes	no	yes	yes	no	no	no
Reduced upwelling and nutrient/food supply	no	no	yes	no	no	no	comp. with hatchery fish
Estuary Degradation	no	Yes	no	no	Yes	no	no
Loss of cool deep pools in streams	no	Yes	no	no	no	no	no

Dam construction not only blocks the passage of the migration, but it impacts the flow regime of the river. Slight changes in the river temperature, turbidity, flow velocity and riverbed habitat make it difficult for the salmon to spawn. The actual dam effects on the magnitude of the declines in the population is however disputed. A nice animation of the population of the Columbia River salmon is presented here as dams get built. The general summary is that the dams initially did a very bad job maintaining survival rates of the salmon causing a huge drop in its population, but more recent renovations in the fish passaging, overfishing policies, stock enhancement programs (i.e. hatcheries) and other stuff has made the situation better.

Hatching its way

Japan of course experienced declines in their salmon as initial dams weren’t given any ecological consideration (only economic ones). The declines were responded with building massive amount of hatcheries (a natural response). Now Japan has 378 hatcheries as opposed to 191 from British Columbia, 3 from South Korea and 148 from Idaho, California, Oregon and Washington combined. Information on Japan’s wild salmon is scarce, and one of the reasons is that most of the salmon are now of hatchery origin (e.g. 95% of chum are hatchery). See map below for its distribution:

So basically, Japan has replaced most of its wild salmon with hatchery salmon over the course of a few decades. Compared to Canada and West Coast United States, Japan has a “hatchery-based management” on their salmon instead of an “ecosystem-based management“ (pdf). This is in part because of the economic success of certain stock-enhancement programs (Hokkaido chum).

Keep in mind there are of course exist many unsuccessful hatchery programs (probably more common) and those are the ones that hide information pertaining to its program. Since hatcheries are businesses, documents produced that explain the merits and demerits of hatchery programs are often partisan.

Any business has its fails and successes. If the hatchery business is successful, we have many jobs for the community. No one can the difference between the taste of wild salmon and hatchery salmon anyways — so who gives?

Some demerits for having a higher proportion of hatchery fish vs. wild fish:

1. lack of diversity: genetically homogeneous hatchery fish has a higher risk of getting wiped out by a chance event
2. decline of fitness: hatchery salmon are not as adept at adjusting to the changing environment (e.g. global warming)
3. water pollution: hatchery fish are reared in nasty waters that is comprised of their own poop
4. pollution within the fish: many fish feed have high amounts of PCBs and other toxins
5. diseases: hatchery fish can be reared resistant to sea lice, where wild fish are not
6. crowding out: if hatchery fish are replacing wild fish instead of supplanting it, why do it?
7. wasteful: salmon are carnivores; hatcheries often give fish meal and fish oil as feed (by certain calculations, one pound of farmed salmon = three pounds of wild fish)
8. bland taste: if you eat it enough, you’ll notice the difference

Basically, a good environment is good for the wild salmon and wild salmon is good for the environment. Moreover, instead of using hatcheries, prepping the environment to make it suitable for the salmon may be more beneficial.  Considering all the environmental, societal and economic consequences, if a dam + hatchery combo is to be built, how can it be made so salmon management is “ecosystem-based?”

part 2 will talk about dams comparisons of Japan and other countries — and features that make it “passable.”

part 3 will talk about dam effect on cultural and lifestyle issues (recreational fishing, etc.).

Global Climate Talk Stasis, thanks Obama

So we heard a series of speeches from world leaders at the UN Summit on Climate Change. Each of the leaders gave a speech indicating their intentions, leading up to the Copenhagen talks. President Mohamed Nasheed of Maldives gave an emotional speech indicating that their country will not exist if AGW is to be continued and all countries need to come together to solve this problem. President Obama of the United States gave a speech that was fluffy per usual.

…to reach the targets we set for 2020 and our long-term goal for 2050.

Eh? What target? The 17% reduction from Waxman-Markey relative to 2005 levels (= 0% reduction relative to 1990 levels)? Obama himself haven’t stated any targets and is pretty much letting Congress develop the bill (I guess this is his way of being very democratic). Meanwhile, China and India are playing the “we’re still a developing country” meme to force the hand on the U.S. Then, China and India can base their targets off the U.S. Meanwhile, having already articulated a 25% reduction target relative to 1990 levels, Japan’s PM Hatoyama Yukio mentioned a framework for helping the developing countries out (aka the Hatoyama Initiative).

Why is the Hatoyama Initiative (or something similar) needed? Because (of course) the negative effects of climate change is inversely related to how industralized a country is.

Top: Proportional map of AGW contribution. Bottom: Proportional map of AGW negative consequences

The two biggest economies (U.S. and Japan) basically shouldn’t even care about negative effects due to anthropogenic global warming (AGW) for the next 100 years if they’re only interested in their own country. Whee! Die Africans! Die Indians! Polar bears! Meanwhile, lemme play golf in mosquito-infested Alaska at 3AM!

So does the U.S. (especially) have the initiative to make bold moves towards alleviating AGW? The more people that are concerned about the issue, the easier to pass something bold. Let’s look at the concern level towards AGW relative to each country (2006).

76% of the people in the U.S. are concerned of climate change. Not bad as I thought.

Wait.

Who has less concern over AGW compared to Americans? — South Africa and Kenya. Hmm, these three countries may contain the most irrational people in the world?

How about the more recent data? Surely the citizens in the U.S. have realized the magnitude of the problem by now. We have a new president! He makes great speeches! He’s not anit-science! Crush those business interests!

Doh!

Is this some kind of “rebound effect”? When the problem gets much worse, we say F-this and give up — like in dieting?

Maybe the Americans don’t like the environment? Or maybe they say, it ain’t our business to fix African countries from the adverse effects of AGW! They could take this upon their own hands by civilizing, like us! Responsibility! But don’t come to our country even if it’s a safe haven; I’d still like to keep my back porch for my weekly BBQ.

Double doh-burger!

What’s so unbelievable about global warming? Because they (IPCC) use climate models? It’s not like climate science includes scientific measurements like in checking water pollution, or mathematical/statistical models to articulate relationships like in every physical science problem in the world?

…

Anyways, the conservative IPCC have adjusted their projections to include a 6 degrees Celsius increase in global temperature within this century as very possible. I understand President Obama has a lot on his hands and limited political capital from the healthcare fight. Given how Carter and Clinton have crashed and burned on each issue by presenting an ideologically pure position, it is only natural to be cautious. But it will be really historic if Obama can set a target similar to the Euros. The tide of AGW opinion is going the wrong way. Obama’s ryuuha of threading the needle between corporate interests and opinions of the general populace on his political stance is really hard to justify from a science perspective on a heavily science-y issue. How about taking a lead-and-influence approach to not only your own people, but other people in other countries.

pretty please?

Finally, here’s a gory description of what could happen @ +6°C by Mark Lynas.

If there is one episode in the Earth’s history that we should try above all not to repeat, it is surely the catastrophe that befell the planet at the end of the Permian period, 251 million years ago. By the end of this calamity, up to 95% of species were extinct. The end-Permian wipeout is the nearest this planet has ever come to becoming just another lifeless rock drifting through space. The precise cause remains unclear, but what is undeniable is that the end-Permian mass extinction was associated with a super-greenhouse event. Oxygen isotopes in rocks dating from the time suggest that temperatures rose by six degrees, perhaps because of an even bigger methane belch than happened 200 million years later in the Eocene.

Sedimentary layers show that most of the world’s plant cover was removed in a catastrophic bout of soil erosion. Rocks also show a “fungal spike” as plants and animals rotted in situ. Still more corpses were washed into the oceans, helping to turn them stagnant and anoxic. Deserts invaded central Europe, and may even have reached close to the Arctic Circle.

One scientific paper investigating “kill mechanisms” during the end-Permian suggests that methane hydrate explosions “could destroy terrestrial life almost entirely”. Acting much like today’s fuel-air explosives (or “vacuum bombs”), major oceanic methane eruptions could release energy equivalent to 10,000 times the world’s stockpile of nuclear weapons.

Whatever happened back then to wipe out 95% of life on Earth must have been pretty serious. And while it would be wrong to imagine that history will ever straightforwardly repeat itself, we should certainly try and learn the lessons of the distant past. If they tell us one thing above all, it is this: that we mess with the climatic thermostat of this planet at our extreme – and growing – peril.

Update: We’re already half-way there to our 2020 target thanks to the bad economy; therefore, don’t be afraid to set more liberal targets!

Update 2: Kerry-Boxer bill announced with steeper cuts:

At the end of the day, what happens early on is what’s most important, not what your goals are 50 years from now

Election Prediction Comparison 2009 Lower House Election (衆議院選挙)

Here is a comparison of election predictions by various outlets of this past election. The top part of the table only shows prediction results for only the single district seats and the bottom part of the table includes the proportional representation seats.

The prediction outlets are composed of internet portals kuchikomi (クチコミ＠総選挙) and shuugi.in (総選挙予測市場サイトshuugi.in), newspapers (Asahi), magazines (Asahi, Mainichi and Post) and TV (TBS, Fuji, Asahi, Tokyo, and Nihon).

So we see a general trend of overestimating the number of seats for the DPJ and underestimating for the LDP. It is so consistent across these media outlets (except for the magazines), it is worth looking into. As far as I know, only the kuchikomi people have posted data on their election results so I will be looking into this in future posts (once I get the data inputted — phew!).

BTW, most of these outlets do not get accurate predictions according to a chi-squared test. The Asahi outlets seem to do well, but shuugi.in is just amazing. They basically got the results exactly right except for a little overestimation of the DPJ results. This may be due to them giving away some candidates due to overflow and those things are hard to predict.

I believe the methods used for shuugi.in uses prediction markets that lets users buy “stocks” to bet candidates that they think are going to win the election. I haven’t studied the method in detail, but I will do that once I get the time. The other internet prediction site, Kuchikomi, uses kuchikomi (口コミ), which basically means internet chatter. Again, I’m not sure what their exact methods are, but it is different from analyzing internet polling.

As for the other media outlets, I am guessing they used traditional opinoin polls. They did perform better than Kuchikomi’s predictions, but not nearly good enough compared to shuugi.in.

More to come! Why did these predictions overestimate the DPJ results???

Keeping a Keen Eye on Media Openness of the DPJ

Japan’s Press Clubs have been in constant flurry of criticism from the citizens of Japan and the foreign press. Their insular nature produces an array of problems towards the ultimate goal of full transparency (see first hand account/analysis here). Although the DPJ have never really articulated their position on Press Clubs or the media openness, they’ve made hints that they would indeed open up (say) press conferences to outside media (foreign and internet).

On the campaign trail, unlike the LDP, the DPJ had their weekly press conferences open. Many accounts say that the open press conferences contributed to a healthy dialog from all kinds of outlets (internet, magazine, freelance, foreign). One foreign correspondent was described by saying that finally, the Japanese press openness is becoming more like the ones in foreign countries. Compared to the LDP, the DPJ had a lot more people at these conferences, and the reporters there freely raised their hands to ask questions.

On March 24th of this year, then DPJ leader Ozawa Ichiro was bombarded with tough questions on his funding scandal of his secretary. The usual stuck-up Ozawa was more upfront and honest to the questions, presumably trying to save his legacy from a freefall. One journalist who works for the internet media company Diamond Online, Uesugi Takashi (上杉隆), asked a simple question. He asked if the DPJ will open up the Press Clubs when they are in power. There, Ozawa affirmed Uesugi that he has no problem opening up to the foreign press, and he has thought that this was an important matter since his days in the LDP. Elated by this answer, Uesugi described this day as:

3月24日は、日本の政治ジャーナリズムにとっては記念すべき日となった。

Today is a day that should be celebrated for Japanese political journalism

On May 16th of this year, DPJ leader Hatoyama Yukio held a press conference, and Uesugi asked the same question to Hatoyama. Hatoyama answered with a pledge:

私が政権を取って官邸に入った場合、上杉さんにもオープンでございますのでどうぞお入りをいただきたい

If I win the election Uesugi-san, the press conferences will open so I invite you to enter the circle

However today, the Hatoyama administration announced that they would add only a few spots for foreign correspondents to their press conferences during the appointments, but no spots to the internet or independent press. They cited a lack of space as one of their reasons for limiting the spots. Internet video journalist, Jimbo Tetsuo (神保哲生), described this situations as:

しかも、就任会見に限って雑誌と外国メディアを招き入れることで、記者クラブ加盟のメディアに「民主党はこんなにオープンです」という記事を書かせる、そ の代わりに、就任会見以降の会見も白紙だし、ネットメディアやフリーランスの記者会見への参加は認めないことにするというバーターというか、取引のような ものが成立したようです。

They are opening up press conferences on only the appointments to Hatoyama’s administration to only the magazines and the foreign press. In doing so, the Press Club will write articles that say “oh the DPJ is finally opening up to the press.” Unfortunately, there are no plans of keeping these spots open once the administration is set, and the complete shut off of internet media and freelance writers seems to suggest that there is something going on behind the scenes.

Thanks to Uesugi’s questioning, the Hatoyama administration is under close scrutiny of how they are going to deal with the press. Historically, even the words “Press Club” was bleeped out of mass media, and there haven’t been any mass media coverage on media openness at all during the election campaign. Today’s news is an incremental step to media openness, but not nearly enough.

Update: Here’s a posted video on open media “pledge.”

Energy Output Breakdown in Japan

So I just watched a clip on the Japanese news, that claimed industry was producing less CO2 overall compared to previous years. Here is a quick look at the statistics from Ministry of the Environment of total energy output (国内最終エネルギー消費の推移). [*note, y-axis is labeled wrong]

In terms of energy output, the industry has done a very good job int he past 19 years. Lets look at an image capture of the TV program (youtube clip) that shows the CO2 output breakdown.

This says, industry has actually decreased CO2 emissions by 2.3%, while household increased by 41.2%, commercial business increased 43.8% and transporation increased 14.6% since 1990.

Industry has said that they’ve made the efforts to reduce CO2 emissions as much as possible, foreseeing the fact that emission standards will be stricter. They claim that any stronger regulation will force them to reduce their business, since efficiency cannot be improved further.

With the proposed 25% reductions in CO2, Japan will need to find the propr apportionment of CO2 reduction to make it as fair as possible. I’m actually a little surprised by the increase of household CO2. I thought migration of the population to big cities, reductions in population and would neutralize, say increase in housing appliances. Japan’s ecopoint system will be beneficial if it replaces a current inefficient appliance. But if those CO2 reduction benefits is outweighed by an increase of appliances in general, it may be all for nought.

Embarrassment of Hatoyama Miyuki?

So Japan is going to have a new, eccentric First Lady in Hatoyama Miyuki. Here are some of her quotes off the top of my head:

1. Rode a UFO to Venus
2. Have met Tom Cruise in her past life
3. I eat the sun

In addition, she’s supposedly writing up a movie that will change peoples’ perceptions of the world forever (with Tom Cruise of course playing the lead character). The Japanese press, who generally know her better than a few selected quotes, have reacted innocuously in either amusement or bewilderment. Many non-DPJ-supporting opinon makers however go a little stronger and feel embarrassment of having a “weird” First Lady represent the people of Japan. The netouyo have gone to accuse her of being a scientologist.

From the foreign press, views on Hatoyama Miyuki (who’s also an actress, playwright, interior designer, entrepreneur, cook, stylist, a lifestyle guru, a macrobiotics enthusiast, spiritualist, author, etc.) have been favorable. She’s already been selected to the top 10 most interesting First Ladies list (gah, can’t find link), and people are intrigued of what influence she can bring compared to the “quiet” First Ladies of previous incarnations. Weird, unique individuals are almost celebrated given that person has good morals.

Many of the “embarrassed” Japanese commentators are a little afraid that non-traditional acts by Hatoyama Miyuki distance Japan from foreign countries without realizing that those exact feelings of embarrassment are what makes the Western press feel “foreign” about Japan.

脳内メーカー ~ Peering into your Subconscious via Your Name

秘 ~ means "secret"

So I checked out my what lies in my subconscious by using 脳内メーカー (brain maker). It’s a simple program that semi-deterministically characterizes your brain by entering your name. I found out that I am full of secrets and a complete stranger to other people — which surprisingly, is not too far from the truth.

So I didn’t like my results; therefore, I used my pen name:

悩 ~ Worries 嘘 ~ Lies 劣 ~ Low Self-Esteem

Ugh, not a good result (again, it’s pretty accurate). There’s actually 52 different types of emotions that can be presented here and both brain characterizations got my current state of my mind spot on. Scary. Here are the 52 emotions.

01.愛 (Love)　02.欲 (Greed)　03.悪 (Evil)　04.遊 (Play)　05.Ｈ (Sex)　06.秘 (Secrecy)　07.食 (Food)　08.友 (Friend)　09.悩 (Worry)　10.休 (Rest) 11.嘘 (Lie)　12.家 (House)　13.妄 (Dreamy)　14.想 (Thinking)　15.気 (Chi)　16.無 (Nothing)　17.恐 (Afraid)　18.敬 (Respect)　19.好 (Like)　20.逃 (Running Away)　 21.怒 (Furious)　22.抱 (Caring)　23.寂 (Lonely)　24.楽 (Easy-going)　25.嫌 (Dislike)　26.苦 (Arduous)　27.虜 (Prey)　28.疑 (Suspicious)　29.告 (Warning)　30.疲 (Tired) 31.忘 (Forgotten)　32.敵 (Enemy)　33.餌 (Someone else’s lunch)　34.学 (Learning)　35.泣 (Crying)　36.羨 (Envy)　37.癒 (Relaxing)　38.幸 (Satisfaction)　39.妬 (Jealousy)　40.変 (Weird) 41.犬 (Dog)　42.猫 (Cat)　43.私 (Me)　44.国 (Country)　45.酒 (Alcohol)　46.謎 (Unknown)　47.淫 (Temptation)　48.負 (Loss)　49.金 (Money)　50.眠 (Sleepy) 51.迷 (Lost)　52.善 (Charitable)

Here’s a couple more:

The Yuuai (友愛) concept may not be disingenuous after all.

Hatoyama’s 25% CO2 Reduction Target, and the Old School Media

Couple days ago, soon-to-be Japanese Prime Minister Hatoyama Yukio announced an aggressive stance on combating climate change. He is vying to cut CO2 emissions by 25% by 2020 respect to 1990 levels. This is a huge departure from the 8% levels that the now defunct Liberal Democratic Party was proposing. The United States on the other hand, is only proposing a 4% cut (via the House Bill). The U.N. climate change advises even deeper cuts (25~40%) for rich countries. As I mentioned in my last post, Hatoyama looks to be a big proponent of doing something about the environment. The goal is for Japan to be the leader of energy efficiency and conservation.

It will be difficult for Hatoyama achieve this goal, as business lobby groups will be furious to keep these cuts to a minimum. There are major political obstacles as well; the biggest worries are “developing” countries like China and India, as they may not adhere to the cuts proposed by the U.N. and other rich countries. Moreover, there are a lot of worries if the cap-and-trade system would not only plunge the domestic economy, but it may lead to transfer funds to foreign countries which enables them to pollute even more.

Another concern is the media. Environmental issues are covered very weakly in the major newspapers. There’s a recent trend especially by the conservative media outlets to give opinion space to climate change deniers. The framing of the debate is turning into a believers vs deniers polarization with the intention of being fair. However, the actual debate among the experts on this issue is not whether anthropogenic global warming (AGW) is real, but it is on the causes of AGW and its corresponding effects. When the debate is framed in the mass media like it is now, uninformed readers will have the false interpretation that the science is not more or less settled.

Hatoyama will have to deal with the largely conservative press (at least on AGW) and will need to strongly reframe the debate portraying actual scientific discussion. I think a lot of the “fair” framing that is done in the media is largely due to the cultural objection to adding restrictions to business and their own lifestyles. The major opinion makers come from the baby-boomer era where resource and financial restrictions weren’t as livid, and culturally, the opinions made by the denier scientists are more palatable. For example, this article Nakagawa Nobuhiro is symbolic of conservative opinion on AGW and Democratic Party of Japan’s policy.

失礼な言い方になりますが、2020年には民主党があるかどうかも定かではありませし、また地球温暖化の原因がCO2ではないという研究結果が発表されることも考えられます。—そのような研究結果はすでに多くあります …

また米国政府、中国政府がCO2削減のテーブルに乗るか—乗させることができるか—どうかをみてからでも遅くはないと思います。あとだしジャンケンでも良いんです。

Excuse me if I may mention that it is unsure whether DPJ will exist in 2020. In addition, there are increasing research results that refute the human causes of global warming [then links to Itoh Kiminori's opinion article]. …

I believe it is advisable to wait and see what the governments of the U.S. and China does with their CO2 cuts. I don’t think it is too late to propose Japan’s plan then.

Mr. Nakagawa links to Itoh Kiminori and declares that there are research results showing AGW is false. Of course Prof. Kiminori has never published a refereed article on climate science, and the linked article mentions his opinion (IMO nitpicking) on climate change. The second paragraph he mentions that Japan should take a “wait-and-see” approach instead of a “lead-and-influence” approach. This latter point is just a healthy difference opinion, but the former point is plain disinformation. I am not accusing Mr. Nakagawa of fudging the facts (like the *cough*Republicans*cough*), but he does have a cultural bias against AGW that clouds his opinion.

In general, being against AGW is culturally more palatable for anyone. Once you disregard the effects of CO2, it opens up people to have more freedom to do stuff, like eat Kobe beef. However, science cannot be disregarded, and hopefully we have more articles that mention the scientific debate more accurately. Currently, people like Prof. Kiminori are seen as revolutionaries, and media will cover him disproportionately to achieve ratings/readership.

It is valuable to look at all sides of the problems and exaggerations; and I see opinions like the above valuable. But if I were to trust a scientists opinion, I will trust the expert. There are some scientists to be careful of: Tsuchida Atsushi, Takeda Nobuhiko, Akasofu Shun and Itoh Kiminori. The last two of which are very accomplished scientists in their own respective fields. But none of them have published a refereed article on climate science, yet has a large voice in the public sphere. Accomplished scientists tend to be more Libertarian, and it is even a bigger cultural hurdle to accept what the majority believes in. More opinions need to be based on the experts such as: Asuka Jyusen, Seita Emori and Kanya Kusano.

Hopefully the media won’t be polarizing as the healthcare debate in the U.S. Nobody would be opposed to the 25% cut if it were easy. The process is to reach the goal requires a complicated and nuanced national effort to maximize environmental benefits while minimizing costs. The media will play a large part of this.

Some good climate change blogs by experts: 環境問題補完計画, Climate Progress and Real Climate.

Update: Or take the approach of we’re-fucked-anyways-so-let’s-adapt.