> You seem to imply that this is voluntary without
offering any polls as evidence.

Indeed. I followed the general pattern in this area - assertion without evidence. If others can do it, I think I should be allowed to do so as well. I am a consumer of research in this area, not a producer.
But, I don't think one needs polls or a PhD in rocket surgery to figure out that not everyone who wants to work wants a full time job. The only real issue is what fraction want full time tenured jobs but cannot get them.

As just one example of women (and men) who deliberately do not take full time jobs when they could have, I would offer recent university graduates who are studying for civil service or other exams. I have had numerous female and some male students who took part time jobs upon graduation so they would have some income but also time for study.

Other cases I know of, beginning with my wife, involve women who were offered full time tenured jobs, but chose to work for the same employer on a part time basis because by their own calculation the increase in pay and benefits was not sufficient reward for a sharply increased time at the work place coupled to an almost complete loss of scheduling flexibility.

Another very general reason many women want part time jobs rather than full time jobs is because their husband loses them as a tax exemption if they earn more than 1030000 yen in a year and if they earn more than
1300000 yen, they have to make their own health insurance and public pension payments rather than being covered as a dependent of their husband.

Further, there is survey data that specifically indicates that some women want part time jobs.

パートを選んだ理由別のパートの割合（複数回答）をみる
と、「自分の都合の良い時間（日）に働きたいから」が
55.8％と最も高い割合となっており、次い で「勤務時間・日
数が短いから」35.2％、「就業調整（年収の調整や労働時間
の調整）ができるから」19.3％の順となっている。

　男女別にみると、女では「家庭の事情（育児・介護等）で
正社員として働けない」が19.3％と男に比べて高い割合と
なっている。

　 年齢階級別にみると、いずれの年齢階級においてもおおむ
ね「自分の都合の良い時間（日）に働きたいから」が最も高
い割合となっているが、「勤務時間・日数 が短いから」は40
歳以上が、「就業調整（年収の調整や労働時間の調整）がで
きるから」は45～59歳が、「簡単な仕事で責任も少ないか
ら」は55歳以上 が、「正社員として採用されなかったから」
は25～29歳と55～59歳が、「正社員としての募集が見つから
なかったから」は25～29歳が、「家庭の事 情（育児・介護
等）で正社員として働けないから」は30～44歳がそれぞれ他
の年齢階級に比べて高い割合となっている。

http://www.mhlw.go.jp/toukei/list/132-23e.html
That many women want some work to earn some extra money but do not want a full time job may also be imputed from questions about they are working.

　働いている理由別のパートの割合（複数回答）をみると、
「主たる稼ぎ手ではないが何らかの家計の足しにするため」
が56.0％と最も高い割合となってお り、次いで「生きがい・
社会参加のため」31.3％、「自分の学費や娯楽費を稼ぐた
め」27.7％、「家計の主たる稼ぎ手として、生活を維持する
ため」 26.2％の順となっている。

　「主たる稼ぎ手ではないが何らかの家計の足しにするた
め」と回答したパートの働いている理由（複数回答）をみる
と、「生活を維持するには不可欠のため」が49.2％と最も高
い割合となっており、次いで「子どもの教育費や仕送りの足
しにするため」26.8％、「住宅 ローン等の返済の足しにする
ため」15.7％の順となっている。

　男女別にみると、男では「家計の主たる稼ぎ手として、生
活を維持するため」52.6％が、女では「主たる稼ぎ手ではな
いが何らかの家計の足しにするため」70.9％が最も高い割合
となっている。

　年齢階級別にみると、15歳～24歳では「自分の学費や娯楽
費を稼ぐため」が、25歳～59歳では「主たる稼ぎ手ではない
が、何らかの家計の足しにするため」が、60歳以上では「家
計の主たる稼ぎ手として、生活を維持するため」が、それぞ
れ最も高い割合となっている

The same survey reports that of males working part time, 30% want full time tenured jobs, whereas less than 20% of the females do.
I make no claim that this survey is the be all and end all, but I would be surprised if other surveys produce wildly different results. A large fraction of women are not looking for full time work. Indeed, there is no small number of men not looking for full time work.

> 1) The share of female non-managerial employees
having
irregular jobs virtually doubled from 29% in 1984 to 57% by the end of 2013 because this was what women preferred, as you hypothesized.
2) It's because that's the kind of jobs employers were willing to provide.

> If option 1 were the case, that would mean that
employer demand for full-time women workers was greater than what women were willing to supply. In that case, the price would go up as it always does when demand exceeds supply. In short, real wages would rise.

> If it's option 2, that would mean employer demand for
full-time women workers was less than what women were willing to provide. In that case, the wage would go down.

I profess I cannot follow your logic. In the nearly 30 year span you are citing, there was a sharp decline in decent, predominantly male blue collar factory jobs.
In their place, came a proliferation of relatively low pay service sector jobs. Many of these jobs are inherently part time because they involve tidal flows.
Restaurants are not equally busy every hour of the day.
Department stores are busier on weekends than weekdays and at certain times of the year. Businesses that cater to people with full time jobs need people willing to work short hours outside of regular business hours.
　(This structural change has been dealt with by Noguchi Yukio and many others).

http://www.asyura2.com/11/hasan73/msg/688.html
http://diamond.jp/articles/-/14498
When the service sector was smaller, there were fewer jobs for women. I think it would be more correct to say that structural change has led to a proliferation of short hour service sectors jobs and that these jobs tend to be relatively attractive to women. Even if these jobs were being taken primarily by men, it is likely that the result would be a decline in wages.

> The evidence is clear. Real wages have been going
down
for years.

Indeed. And Noguchi Yukio has explained this, at least to my satisfaction, in terms of structural shift, not gender -- the rapidly decline in well paying industrial jobs and their replacement by low pay service sector jobs.

> The same, by the way, is true of men, where the share
of irregular employees rose from 8% in 1984 to 22% by the end of 2013.

Indeed. And, as your numbers show men have been hit proportionally harder than women. I have even heard a labor economist (a woman) describe the change as "the feminization of male jobs" by which she meant that low pay, short term jobs that had historically been taken by women were now being taken by men because that's all they could get. But, to the extent that this represents a convergence between the employment of men and women, it is a positive step in the direction of gender equality, right? In terms of both logic and arithmetic there are two ways to equalize indicators where one is high and the other is low. Lowering the high indicator is just as good as raising the low indicator in terms of equality, right?
Punch line: there's more going on than gender discrimination. It's a factor but it's far from the whole story.

EHK