Wednesday, June 12, 2013

An Impossible Result

It's impossssssible!
To put a Cadillac in your nose.
It's just imposssssible!

- Steve Martin, Let's Get Small (1977)

Public opinion polls are supposed to be a reality check for politicians, keeping them from getting either too big for their britches or too far out ahead of the curve.

However, sometimes the results of public opinion polls demonstrate that the pollsters, the public or both have a slippery grip on reality.

NHK on Monday released the results of its June 7-9 public opinion poll. The results were largely in accordance with expectations:

- the Cabinet is still popular (62% supporting, 20% not supporting)

- the turnout for the July 21 House of Councillors election will be poor, but not historically poor (58% of the respondents said they "definitely will vote" - indicating that actual turnout will be around 50%)*

- a solid majority of the voters (64%) want the Liberal Democratic Party-New Komeito alliance to win enough seats in the House of Councillors in July to be in control of that House. Only 29% want the twisted Diet (nejire kokkai), the current splitting of control of the Houses of the Diet in between the ruling and opposition parties, to continue

- a more than two-thirds majority (69%) of the voters appreciate the government's economic programs, a.k.a., Abenomics

- the LDP is the only major party around today

Q: Which party do you support?
LDP 41.7%
DPJ 5.8%
New Komeito 5.1%
Communist 2.2%
JRP 1.5%
Your Party 1.5%
SDP 0.4%
Life 0.1%
None 34.5%
There is one unexpected result -- and it is a doozy.

To the question of whether the respondents personally felt they had accrued benefits from Abenomics, the replies this month were:
I am feeling them 11%

I am not feeling them 46%

I cannot say either way 37%
What is so charming about this result? It is impossible.

Last month (10-12 May 2013) NHK asked exactly the same question. The responses NHK pollsters recorded then were:
I am feeling them 21%

I am not feeling them 36%

I cannot say either way 40%
You read that right.

In May, 21% are feelin' it, baby.

In June, only 11%.

So, the Abe administration is in place one more month, initiating and implementing more and more of its policy program each day. During that month the percentage of those declaring that they have felt the effect of Abe's economics policies shrinks by 10 percentage points.

For the NHK statistician, the options are not terrific. Either

1) the poll has a colossal margin of error, capable of accommodating a 10 point drop and a 10 point rise -- which means the polling is unreliable, or

2) at least one poll has a perverse and crippling sampling error, or errors, wiping out 10% of the population at a go -- which means your polling is unreliable, or

3) when asked about their own lives respondents tell the poll data collector not what is happening in their lives but what they think is happening in their lives -- or what they think the pollster wants to hear is happening in their lives -- which means your polling is unreliable, or

4) combinations of the above, which means...well, you know already.

The most likely reason for the evidently possible (it has been recorded and reported) but obviously unusable set of results is #3, with the respondents getting caught up in the earlier zeitgeist and answering "Yes!" without looking at their own situations with objectivity. However the party support numbers for the JRP and the Your Party are also suspiciously low in this poll. In most other polls support for these two parties hovers in the 4% to 5% range. So options #1, #2 and #4 could also be in play.

Whatever the cause(s), the loss of 10% of the "Abenomics - Are you feeling it?" believers over a month should be attracting a heck of a lot more attention.

Later - In comments, a reader offers an Option #5.


* The lowest turnout ever for a House of Councillors election was in 1995, when only 44.5% of the voters showed up.


The Failed Trader in a Strange Land said...

Surely the effect of falling stock prices has a lot to do with the reduction in the numbers who have personally benefit?

Philippe said...

I'm not a statistician nor do I pay much attention to those polls, but a 10% flip doesn't sound so strange. It could very well be that since the previous poll, the reality of abenomics has started to bite a little more: rising prices for most imported stuff, and in particular daily necessities (cooking oil, toilet paper,…) on top of the rise of electricity, gas and so on. People are negatively affected in their daily lives with (possible) improvements still hanging in the mist of the future (I'm charitable!). The bubbly look of the stock market will certainly affect the views of the upper middle class. And finally, despite every effort of the mass media to hide it, the absence of details on further steps in the abenomics program (see mist above).

MTC said...

The Failed Trader in a Strange Land -

A clever retort...but no.

Too few Japanese hold equities or equities linked-accounts for recent market volatility to have affected the mindset.

MTC said...

Philippe -

You and I agree on the causes for the increase in those saying Abenomics has not benefited them. But let us accept as a basic proposition that persons, having once said they have felt the benefits, cannot turn around say they have not unless they confess, "I'm sorry. I got caught up in the euphoria and ended up mininterpreting events. Forget about what I said last time."

"The Failed Trader in a Strange Land" provides an explanation that would produce the fluctuation without euphoria, if only enough Japanese owned equities.

Peregrine Lancelot-Smythe said...

I don't know, for the past six months Japan has been a reasonable place to live again.

People seemed somehow more civil or optimistic - totally unscientific of course - just on my impressions of commuting around Tokyo.

This week, however, I have started to see the return of the grumpy salaryman, devoid of manners or decency.

It got me thinking that either it's the rain or the stock market.

Martin J Frid said...

There ought to be a percentage point, say 2/3 voter turnout, or even better in the range of 80-90%, and if that didn't happen, the election would be invalid. And the politicians would have to make a better case for their future job application.

Remind me, just how many people actually voted for Abe, in real numbers...?