Diversity in the metastudy

Wharton's Katherine Klein has a nice literature review up on the effects of corporate boardroom gender diversity.

The bottom line seems to be no effect.

Klein usefully contrasts consultancy reports on the topic with the findings of the academic literature:

Do companies with women on the board perform better than companies whose boards are all-male? Many popular press articles and fund managers make this claim, citing studies by consulting firms, information providers and financial institutions, such as McKinsey, Thomson Reuters and Credit Suisse.

Writing recently on Huffington Post, for example, one consultant observed the following:

“Companies with gender-diverse management teams have been proven to consistently perform better and be more profitable than those without them. There is overwhelming evidence to support the value of having more women in senior leadership positions. A growing body of research –including studies by McKinsey & Company — has proven that companies with more women in senior executive and board roles have advantages over those that don’t.”

But research conducted by consulting firms and financial institutions is not as rigorous as peer-reviewed academic research. Here, I dig into the findings of rigorous, peer-reviewed studies of the relationship between board gender diversity and company performance.

Spoiler alert: Rigorous, peer-reviewed studies suggest that companies do not perform better when they have women on the board. Nor do they perform worse. Depending on which meta-analysis you read, board gender diversity either has a very weak relationship with board performance or no relationship at all.

That's consistent with my read of things as well. But be careful here too: there being no relationship doesn't mean that quotas or mandates would be costless. You'd need to specifically sort through the studies that looked at effects of quotas, because changes in board composition that are board initiated might differ from ones that are compulsory.

Previously: Wishful Treasury Thinking

Good points on pay equity

Today's reader mailbag brings a few excellent points on the government's pay equity deal for homecare workers.

I think your discussion of pay equity (at least the bit you quote in the blog) does not clearly distinguish between two very different questions: (i) is there an equity issue (ie evidence of discrimination against women)? (ii) is government intervention likely to create more distortions than it solves. One of the reasons I think Hayek is very powerful (cf the "Constitution of Liberty" on anti-trust legislation) by being happy to allow there will be many distortions where real world markets do not deliver efficient outcomes. The force of his argument is that the justification for government intervention has to show how the intervention improves on than outcomes after intervention.

In this context, there is nothing inherent to the way markets operate that will deliver equity and if consumer (or government funding) preferences are racist and sexist, so will the outcomes be. You can demonstrate (as Becker did in the late 1950s) that people make themselves materially worse off by acting in this way, but this is just one of many examples of people trading material welfare to maximise utility (the other obvious ones are giving to charity, voting to pay higher taxes and paying to reduce risk).

But that is not enough to justify government intervention. To justify intervention, it needs to be shown that the outcomes are "better" after the intervention. For instance, if the care industry responds to additional funding by investing in human and physical capital that means many people currently employed in the industry lose their jobs, the outcome may be fairer in the sense of having removed gender inequity for those who are qualified, but have done so by punishing those unable to get qualifications and/or younger people. In fact, you will find in health and social care that occupational licensing has tended to do precisely this, but with a twist that many of the people on lower pay are simply reclassified. So nurses used to do a wide of range tasks. It is no longer worth employing a fully trained nurse to do the lower skilled tasks so "trainees" do some of them and others are done by people explicitly labelled for those lower skilled tasks (e.g. "cleaners").

Therefore my prediction of the impact of the legislation is this:

• After the initial redistribution there will be a joint effort by health unions and employers to increase the training barriers to be eligible for the higher pay;

• This will include an extended training period when younger people will be paid at a lower rate because they are "training". Much of the on-the-job component of training will be doing many of the lower skilled tasks previously done by everyone;

• There will be a plethora of new job titles created, where all will have lower certification requirements than "fully qualified" carers and therefore have lower pay.

• Most of the people in the lower qualified jobs will be women, earning slightly less than they would previously have earned. They will be disproportionately from poor backgrounds, more likely to be Maori and PI, and will find most practical means of promotion blocked because of the new licensing requirements...

I am not sure that counts as equitable...?

I agree with my correspondent that Becker models under-emphasise that the path to equilibrium can be slow. Taste-based discrimination only really holds up where the customers have a preference for discrimination (unlikely in the homecare case) or where there's a severe lack of competition - and that's the dominant funder problem that MBIE's RIS talked about.  

I fear that my correspondent is right about what comes next. I hope that what comes next is the DHBs shifting to allow more people to hire their carers through MyCare. It's better for the homecare clients and better for the workers and cheaper for the health system. But I'm a bit pessimistic there too - it would require back-end changes that DHBs seem to like throwing into a too-hard basket. 

Pay and equity

I went through the government's proposed pay equity legislation in last week's NBR and in The Initiative's Insights newsletter. Some snippets:

The bill would ultimately have the authority decide on allowed wages – the price of labour – if employees brought a claim that they were underpaid.

It requires that remuneration in female-dominated occupations be no less than remuneration in male-dominated occupations with substantially similar skills responsibilities and services, where work is undertaken under substantially similar conditions, and where substantially similar degrees of effort are involved.

But that gets the workings of prices in competitive markets completely backward.

Value and price do not derive simply from the skill involved in some type of labour, or from the working conditions, or from inherent responsibilities. What matters instead is how much consumers value another bit of the final product or service, and how many people are willing to provide the labour to supply it.

If two jobs have, as far as an employment tribunal can tell, equivalent skills, equivalent working conditions and equivalent responsibilities, and one pays far more than the other, that price difference is important.

It says there is a surplus of workers in lower-paid jobs, relative to demand for their services – and a relative shortage in the higher-paid jobs.

For example, translation between Maori and English, or between French and English, may require equivalent skill, training and working conditions but no tribunal could improve on existing market prices to tell you which job should pay more in New Zealand.

I went on to talk a bit about the pay equity mess in Ontario in the 1980s. I hit that part more directly in the Insights column:

The Ontario Tribunal that had to decide on pay equity claims had to choose which professions provided the most appropriate comparison for the group that claimed to be underpaid. The result, as one evaluation later put it, was a “litigation nightmare.”

New Zealand will face similar and substantial problems in assessing which occupations have similar skills and responsibilities, similar working conditions, and require similar levels of effort. While Treasury made the job somewhat easier by providing a hierarchy for selecting appropriate comparators, it would be surprising if New Zealand did not face litigation nightmares of its own.

In short, New Zealand learned little from Ontario’s experience. Neither Tribunals nor Authorities are well placed to set pay relativities between different occupations.

In this coming Friday's Insights column, I'll cover a better way for better pay. If you haven't subscribed already, the sign-up link is at the bottom of the column.

I noted in the column an OIA request of Treasury on their advice regarding pay equity and whether they provided any warnings about the mess that Ontario encountered. Doesn't look like Ontario came up. Here's the first reply, and the rest of the documentation.

Guaranteed Employment for Lawyers

Our excellent Research Fellow Dr Rachel Hodder warns about potential legal nightmares in the government's proposed pay equity framework.

Last week, the government announced that it would accept the Joint Working Group’s recommendations, which will allow employees to take pay equity disputes directly to their employers.

These changes were motivated by the landmark ruling in the TerraNova case where it was ruled that female-dominant aged care work has been historically undervalued due to gender discrimination. This has opened the door for a flood of similar claims to be made in other female-dominant industries.

The intentions of the changes are laudable but good intentions do not always make good policy. Overseas experience with similar pay equity laws is a good reason to be concerned.

There is little evidence that similar laws have reduced gender pay gaps. Adjusting pay in undervalued female-dominant industries is a blunt tool compared with adjusting pay for individual undervalued female employees in any industry.

Ironically, some studies have demonstrated that pay equity laws have widened the pay gap. Making employers pay more in whole industries reduces the number of jobs available, pushing many women into lower paid work.

A review of similar pay equity laws in Ontario is discouraging. The costs of administering the system ate up much of the available compensation change.

Establishing plans for compensation systems was much easier said than done. These costs were disproportionately felt by small employers where often the administrative costs were more than the recognised pay differences.

Difficult in practiceDisputes turned into a litigation nightmare. Rules to determine what counted as equivalent work seemed simple on paper but were difficult to evaluate in practice.

This led to endless disagreements about what male-dominant industries could be considered as relevant comparators. The main winners of the laws were consultants and lawyers, not women.

One would hope the government has been advised on the likely problems the new laws could encounter based on overseas experience.

It would be very interesting to know just how much warning Treasury or the ministries provided about the difficulties encountered in Ontario, and likely difficulties here. One should be especially careful about the quality of advice in areas where there is a lot of wishful thinking, and a lot of social sanction for even suggesting that outcomes might diverge from intentions.

Meanwhile, The Listener does not give us reason to hope for reasonableness:

The law change won’t be a magic wand. Comparing the relative value of different jobs in disparate industries was a contentious issue within the working party and threatens to be a sticking point in workplace negotiations. A suggestion by the New Zealand Educational Institute that pay rates for teacher aides be linked to those of Corrections officers, whose work entails risks not generally present in primary schools, isn’t a promising start.

The government hasn't drafted the legislation yet; I hope that the government gets some better advice around this stuff before it does.

New Zealand is totally different from Ontario

New Zealand may be heading toward 1988-Ontario-style Pay Equity rules.

Here's The Herald's summary:

The changes will mean employees who believe they are underpaid because they do work in fields dominated by women will be able to approach their employer to raise a pay equity claim.

The Government has today announced it has accepted the recommendations from a joint working group, led by Business NZ and the NZ Council of Trade Unions, that were delivered to Cabinet earlier this year.

When putting up a pay equity claim, employees need to find another non-female-dominated job they can compare their work to.

As well as establishing a process for employers and employees to follow to address pay equity, ministers have also decided to clarify how to choose an appropriate job for comparison when making a pay equity claim.

PSA national secretary Erin Polaczuk said Cabinet had proposed a "start close then move out" mechanism, where employees must try to find a comparison in their own business, then their industry, then their sector.

She said the PSA would now raise pay equity claims for thousands of low-paid women.

I am absolutely confident that our experience will be absolutely nothing like Ontario's. I only put the Ontario experience up here as illustration of what goes wrong in other places. Here it will be wonderful, as I am certain all right-thinking people would have to agree.

On administrative costs, here's a report by Morley Gunderson (University of Toronto) and Paul Lanoie (École des Hautes Études Commerciales) for CIRANO. You should read the whole thing, so you can satisfy yourself that the clear differences between Ontario and New Zealand mean we are entirely safe. Here is a snippet.

Based on these figures, in some back-of-the-envelope calculations, Gunderson (1994, p. 88) estimated the administrative costs for the 20 percent of the persons in an establishment who may receive an award, to be approximately $750 or about 20 percent of the average award of around $4,000. That is, about 20 cents of every dollar transferred is "eaten-up" in the employer’s cost of administering the system. This may be an underestimate to the extent that it does not include the actual cost of the job evaluation procedure, or the hidden indirect cost (e.g., managerial time) or the cost to other parties. It may be an overestimate to the extent it does not net out other benefits that may result from the process. Furthermore, many of the administrative costs may be one-time only, while the pay equity adjustments may continue.

Based on her consultations across the province, Read (1996, p. 4) also indicated: "For some smaller private sector employers, the cost of compliance sometimes exceeded the actual adjustment made." That is, more than 100 percent of the adjustment would get "eaten-up" in the real resource cost of administering the system. She cites (p. 38) survey evidence from the Canadian Federation of Independent Business indicating that for small businesses the total cost of consulting fees10 averaged around $5,400, which was considerably higher than the total amount of around $4,000 that went into wage adjustments.

As an illustration of the costs incurred, we present the process that was involved in the hospital that served as one of our case studies. First, a job description had to be done for all occupations by people of the HR services. The job description was then presented to the supervisors who submitted them to all workers for approval. Four pay equity committees (one for every bargaining unit and one for non-unionised workers), each of them involving 5 or 6 persons then proceeded to the job evaluation. This took many weeks, involving a countless number of meetings. The results of the job evaluation and the consequent wage adjustment were then presented to supervisors before being posted. Once the plan was posted, HR officers had to meet a large number of workers to discuss the results and the wage adjustment they received. The pay equity law also requires that firms maintain pay equity, which means that every time a job description is changed (because of a re-engineering, because a worker uses a new equipment, etc.), the pay equity process has to be done again.

When the pay equity system was first conceived in Ontario, it is not likely that the complexities and "litigation nightmare" were anticipated11. Nor were the long delays and legal wranglings of the Tribunal over seemingly simple, but inherently complex and important matters12. Examples of issues dealt with by the Tribunal include13: the definition of the establishment and the employer; gender composition of the job; gender bias in job evaluation; appropriateness of certain male comparators; exemptions for bargaining strength; retroactivity in the wage adjustment; language difficulties in understanding the pay equity plan; disclosure of documents; the possibility of an "officially induced error" if a Review Officer gives an opinion; and the rights of employees to paid time off to participate in the Tribunal hearing.

The job-to-job comparison procedures were likely utilised because of the apparent simplicity of comparing one job to another once the appropriate comparison group was established -- not anticipating the complexity of selecting the comparison group. This could be minimised by allowing the estimation of paylines, but that too can give rise to endless technical wrangling. The problem of a lack of male comparator jobs was not likely anticipated, nor was the complexity of proportionate value methods (with their paylines) or proxy comparisons to close those loopholes. Regulations beget further regulations to close the loopholes, with the process often bordering on the impenetrable to all but a few. Of course, there is money to be made by such complexities, but that money does not go to those for whom the system was designed to assist.

The CIRANO piece concludes:

Pay equity can be a complex procedure even if it is intended to be simple. That very complexity can give rise to negative features when judged according to program evaluation procedures. Not all persons in the target groups are assisted and in a manner sufficient to fully redress their problem. Significant spillover benefits can leak to non-target groups. Otherwise similar groups can receive very different treatment. The process may foster rather than break down the sex segregation of jobs. Significant real resources can be used up in the administration and implementation of the complex procedures. Such procedures are not transparent and well understood by the parties.

Our assessment is that these negative features especially associated with the complexity of the policy are a significant barrier to the effective implementation and wider adoption of the policy. Pay equity runs the very real risk of "imploding" under its own weight. The problem is exacerbated by the fact that simple rules for its application are elusive. Rather than closing loopholes, further regulations seem to add more complexity, begetting even further regulations. The pie gets eaten up in the process of administering the pie.

These problems appear to be inherent in the nature of the policy, at least when it is applied proactively and on unit-by-unit bases, as is inevitable when applied to the private sector. This raises the danger that the legislation will be ignored because it is unwieldy, a possibility enhanced by the fact that settlements have been much smaller in the private sector than the public sector.

Pay equity was instituted in large part because of the limitations and ineffectiveness of conventional equal pay policies in reducing the male-female wage gap17. Unfortunately, we know remarkably little about the overall effectiveness of initiatives like Ontario’s pay equity initiatives. We know that Ontario is a leader in the world in the area of pay equity, and we know that its system has turned out to be complex, but we do not have an answer to the bottom-line question: has pay equity been successful in closing the wage gap or in achieving other social objectives?

They didn't have an answer to their bottom-line question in 1999 when that piece was written. What have we learned since? Here's McDonald and Thornton (sci-hub link)

Evaluating the effect of pay-equity laws is important and yet difficult as one needs to deduce what would have occurred without the policy intervention. We use a new tool, synthetic-control method, to examine the effects of Ontario's Pay Equity Act on the gender pay gap. This tool enables us to create a “Synthetic” Ontario, which resembles Ontario more closely than does any other single province. Using Synthetic Ontario to compare what actually happened in Ontario to what would have happened, we find that the act has had little or no effect on the female-male wage gap in Ontario.

New Zealand will be totally different. Here we never create systems that layer up tons of bureaucratic processes imposing massive cost for little discernible benefit, and surely that will continue to be as true as it ever was.

Part time work and the wage gap

The Herald's reporting on a new tool they've produced to investigate the gender wage gap. Enter your broad age and profession (ANZSCO 2006), and it will tell you how much more (less) you earned than a woman (man) of the same age and profession. How?

This annual calculation makes the assumption of working 40 hours a week and 52 weeks a year. These rates are before tax.
There are a range of reasons which can influence the number of hours men and women work.
If an occupation has been selected we calculate the difference in male and female median hourly wages for that ANZSCO 2006 category and age group. If no occupation has been selected the overall difference in median hourly rates is displayed.

Last year, I'd noted that there are some pretty substantial differences in full-time and part-time work choices that affect the overall pay gap results:

One simple change could almost halve measured pay inequality. What is it? Have women replicate men’s split between full-time and part-time work. Ok, maybe that’s not so simple. But while there are 7.05 full time male workers for every part-time male worker, there are 1.96 full time female workers for every part-time female worker. And part-timers always earn less than full-timers. Interestingly enough, part-time female workers earn more than part-time male workers – for the obvious reason given those numbers.

If you re-weighted the pay gap using women’s median earnings for full-time and part-time work, but men’s split between full-time and part-time work, the pay gap would drop from 11.9% to 6.6%. And that’s without controlling for a pile of other important stuff, like time spent outside of the workforce or differences in education background and the like.

The mix between part-time and full-time work matters. Part time work generally pays less per hour than does equivalent full-time work - except in some higher end contracting that won't much be evident in median earnings anyway.

If we look only at full-time earnings, men make $25 per hour and women make $23.02. Or 92 cents for every $1 of male earnings. If we look only at part-time earnings, men earn $16 per hour and women earn $17.65 per hour, or $1.10 for every $1 of male earnings. If we bundle those both together, because a greater proportion of women are in part-time employment which pays less overall, you get a larger pay gap: women earn $0.88 for every $1 of male earnings.

So, what to do about it? I'd suggested (as a first cut) equivalising things so that you were comparing men and women on the same basis, such that they'd have equivalent proportions of men and women.

When I asked the Herald why they hadn't accounted for full versus part-time employment, they noted they were following Stats' recommended procedure:

We can measure a gender pay gap for either full-time or part-time workers separately. When we do this, we find that for full-time workers only, the gap is smaller but has a similar up-and-down pattern over time as the ‘total’ gender pay gap. For part-time workers, the gap reverses – women who work part time typically earn more (per hour) than men who work part time.

When we separate workers into full-time and part-time groups, we hope to remove the differences caused by the types of jobs that offer (or don’t offer) part-time hours. However, splitting workers into full-time and part-time work can change the balance of other factors that affect pay, such as age. For example, females working part time are more likely to be older than males working part-time.

Overall, we recommend using the median hourly pay across all workers, rather than using full-time or part-time workers separately.

That's all well-and-good, but you can't really go from there to talking about the resulting gap reflecting failures in equal-pay-for-equal-work, as some of the quoted folks in the story then do. Part of the salary bundle in part-time work is the flexibility it gives you. That is worth something to those in part-time work or they wouldn't be choosing it. And because it is valuable to employees and costly to employers, it draws a pay penalty.

I really can't recommend strongly enough that folks running the pay equity beat read Claudia Goldin's summary of the state of the literature. It was her 2014 American Economics Association Presidential Address. Or, just listen to her interview with Kathyrn Ryan from 2014.

Wishful Treasury thinking?

Here's Treasury Secretary Gabriel Makhlouf in The NBR:

The business case for diversity in the workplace is clear. In the case of gender diversity, international research shows companies that have a balanced representation of women and men on their boards perform better.

I was curious what Makhlouf was citing when he referred to international research, so I asked. Treasury says Gabs was citing a Deloitte piece; the Deloitte piece cites work by Catalyst (2011 and 2007). The 2007 piece is a one-page infographic. The 2011 version splits firms up by quartiles of gender diversity and compares returns across quartiles as comparisons of means in what's basically a cross-tab and t-test. And it also cites a McKinsey piece, which is also just comparisons of returns across quartiles of firms sorted by gender diversity without any correction for other firm-level differences.

Now if there's any endogeneity in selection of board members, there could be problems. And there could also be problems if there are other things that correlate with both board choices and with performance. That's why the more standard method will run a pile of other stuff we know predicts firm performance then add the new variable, like gender diversity, in a multivariate regression. And it might try to find an instrument to get around the endogeneity issue.

Here's Adams and Ferreira in the Journal of Financial Economics, who do things properly and correct for all the other firm-specific bits:

We show that female directors have a significant impact on board inputs and firm outcomes. In a sample of US firms, we find that female directors have better attendance records than male directors, male directors have fewer attendance problems the more gender-diverse the board is, and women are more likely to join monitoring committees. These results suggest that gender-diverse boards allocate more effort to monitoring. Accordingly, we find that chief executive officer turnover is more sensitive to stock performance and directors receive more equity-based compensation in firms with more gender-diverse boards. However, the average effect of gender diversity on firm performance is negative. This negative effect is driven by companies with fewer takeover defenses. Our results suggest that mandating gender quotas for directors can reduce firm value for well-governed firms.

I'd quoted from their conclusion last year, emphasis added here:

Our results highlight the importance of trying to address the endogeneity of gender diversity in performance regressions. Although a positive relation between gender diversity in the boardroom and firm performance is often cited in the popular press, it is not robust to any of our methods of addressing the endogeneity of gender diversity. The true relation between gender diversity and firm performance appears to be more complex. We find that diversity has a positive impact on performance in firms that otherwise have weak governance, as measured by their abilities to resist takeovers. In firms with strong governance, however, enforcing gender quotas in the boardroom could ultimately decrease shareholder value. One possible explanation is that greater gender diversity could lead to overmonitoring in those firms.

More generally, our results show that female directors have a substantial and value-relevant impact on board structure. But this evidence does not provide support for quota-based policy initiatives. No evidence suggests that such policies would improve firm performance on average. Proposals for regulations enforcing quotas for women on boards must then be motivated by reasons other than improvements in governance and firm performance.

It is interesting that the Secretary of the Treasury missed this piece in the Journal of Financial Economics on a question of financial economics. Surely he has a research staff that can point him to these things.

There are other pieces out there in reputable journals, but this seems the one that has to be answered. Infographics or comparisons of means where firms are sorted by gender representation* aren't a great place to start.

I agree with Makhlouf's ultimate conclusion that providing flexible arrangements** is an important part of enabling greater female representation in senior leadership. But we shouldn't necessarily expect it to yield great dividends on the dividends front.


* Treasury's Living Standards Framework likely has a weighting in which it's ok to rank infographics ahead of the Journal of Financial Economics on matters of Financial Economics.

** On a related Wellington-specific whinge, think about how many offices put on after-work work events in the 5:30 - 7 pm time slot. Things held during office hours that are work-related, well, you balance it against other work stuff and make a call. Things held after kid bedtime that are work-related - if you're a two-parent family, it isn't hard to work something out. But the 5:30-7 pm slot is the absolute worst. If you have two working parents, somebody is then stuck on after-school duty if the other has to be at a post-work-but-still-work thing. And sole-parent households would have to sort sitters.

It's hard to find a day when there isn't an invite for some 5:30pm event that would be worthwhile, is great networking, and that would matter for work one way or another. They're great for the not-yet-kidded, and for those with adult children. But it's a reasonable hindrance on everybody else. We have those events too. Would that there were a better equilibrium.

Wage gaps and part-time work

I had a look at the latest numbers over at The Initiative's blog:

Last week’s New Zealand Income Survey sure led to a lot of headlines. Here are some of them:

• NZ Gender Pay Gap Increases
• Gender pay gap balloons to almost 12% [and the Minister for Women wants to know why!]
• Hourly wages rise, but gender gap back to six-year high
• Human Rights Commission urges action over growing gender pay gap

One simple change could almost halve measured pay inequality. What is it? Have women replicate men’s split between full-time and part-time work. Ok, maybe that’s not so simple. But while there are 7.05 full time male workers for every part-time male worker, there are 1.96 full time female workers for every part-time female workers. And part-timers always earn less than full-timers. Interestingly enough, part-time female workers earn more than part-time male workers – for the obvious reason given those numbers.

If you re-weighted the pay gap using women’s median earnings for full-time and part-time work, but men’s split between full-time and part-time work, the pay gap would drop from 11.9% to 6.6%. And that’s without controlling for a pile of other important stuff, like time spent outside of the workforce or differences in education background and the like.

How is this different from last year? Last year’s wage gap was 9.8%. In the 2014 NZIS, there were 6.65 full time male workers for every part-timer and 2.02 full-time female workers for every part-timer. So the proportion of male workers in full time positions increased while the proportion of women in full-time positions decreased. The absolute number of part-time male workers dropped; the number of part-time female workers increased.

How big is the effect? It looks like the change in the relative proportions of part time and full time workers accounts for about a fifth of the increase in the wage gap from 2014 to 2015. The rest looks due to a much stronger increase in male full-time earnings than in female full-time earnings – though part-time male earnings have stagnated while part-time female earnings grew.

Meanwhile, the Herald is blaming "bro culture" for that women take longer to pay off student loans than men.

The latest Education Ministry report on the student loan scheme says male students who graduated in 2011 will pay off their loans in an average of 6.7 years, but women will take a further six months.

The gap is far smaller than 2005, before student loans were made interest-free.

In 2005, men would take an average of 14 years and women 28 years to pay off their debts.

Hmm. Any guesses about something that might happen for at least some women in the first 5-10 years after college graduation that would mean an income-contingent loan repayment scheme might take longer for women than for men? Any?

And recall that student loans here don't accrue interest if the debtor stays in New Zealand. If the numbers above are right, then inflation eats far more of female graduates' debt burden than it does for males. Think about it this way. If you owe a big pile of money on which zero interest is accruing, would you prefer to be able to pay it off really quickly or have it deducted from your estate on your deathbed? The longer the term of a zero-interest loan, the better for the borrower!

Update: Brennan McDonald points out that student loan balances are written off when the government sees your death certificate.

Gender-blind economists

In a new audit study, male economists are the only ones who came out as gender-blind in hiring preferences.

The underrepresentation of women in academic science is typically attributed, both in scientific literature and in the media, to sexist hiring. Here we report five hiring experiments in which faculty evaluated hypothetical female and male applicants, using systematically varied profiles disguising identical scholarship, for assistant professorships in biology, engineering, economics, and psychology. Contrary to prevailing assumptions, men and women faculty members from all four fields preferred female applicants 2:1 over identically qualified males with matching lifestyles (single, married, divorced), with the exception of male economists, who showed no gender preference.Comparing different lifestyles revealed that women preferred divorced mothers to married fathers and that men preferred mothers who took parental leaves to mothers who did not. Our findings, supported by real-world academic hiring data, suggest advantages for women launching academic science careers.

Via @clairlemon

Treasury learnings

Things we didn't know before the latest set of Treasury working papers came out:

• Income mobility in NZ is high: From one year to the next, over 60% of the population shifts income decile group; only 22% stay in the same income decile group eight years later. Remember that when looking at numbers for the bottom 10% or bottom 20%: this largely isn't persistent poverty.
• Most people talking about inequality don't have a clue what they're talking about. Every one of the different statistical measures of inequality has an underlying set of background value judgements. Creedy and Eedrah draw the links. Creedy explains further here.
• Workers in foreign-owned firms earn 2 - 4% more than Kiwis employed in NZ-owned firms, after correcting for that foreign firms tend to be larger. But I doubt it'll matter for those who want to ban foreign investment in New Zealand. 
• Working For Families, by increasing the effective marginal tax rate for second-earners, reduced married women's labour force participation rates and hours worked. But it also reduced inequality and poverty. 
• If you increase the age of eligibility for public pensions, people save more. If Key announced today that the retirement age were to go up in 2025, we'd see saving rates increase now in anticipation of it. 

The full set of papers is here.

The gender wage gap and hours worked

From my piece with Bryce Wilkinson in last week's National Business Review:

Suppose a management consultant pulled you aside and said, “Hey, have I got a deal for you. With this one simple trick, I can lower your labour costs by 14% and it won’t cost you a dime in productivity. Trust me, I know what I’m doing.” You’d probably be a little sceptical about the claim, or at least I would.

Statistics released in the last week by the State Services Commission reportedly found that the pay gap between men and women in public sector management roles averaged 14%.

...
Sadly, there are few one-simple-trick paths to success. More rigorous analysis of gender pay gaps that accounts for differences in work experience, training, time outside of the workforce, choice of industry, and hours worked generally wipes out the bulk of the headline pay differences.

The recently released New Zealand Income Survey data also makes things very clear. Hourly earnings for full time male employees, at $23.97/hour, is only about 6% higher than women’s $22.54; women in part-time work, averaging $17.26/hour, beat men’s $16/hour. Much of the headline pay gap comes from that more men are in full-time work rather than part-time work.

Treasury released analysis a fortnight ago showing that Working for Families (WFF) has reduced the number of hours worked by married women. Prior to WFF, among married men in employment, about 10% worked part-time, 59% worked a 40-hour week, and 32% worked a 50-hour week. For women, 55% were in part-time work, 30% worked a 40-hour week, and 15% worked a 50-hour week. None of those figures changed greatly with WFF, but women in part-time work shifted to working fewer hours, and about 9000 married women dropped out of the workforce entirely.

In a lot of professions, wages start climbing substantially for employees willing to put in the longer work-weeks, as economist Claudia Goldin pointed out in her 2014 Presidential Address to the American Economic Association. Much of the measured pay gap comes down not to discrimination, but to choices.

Update: NBR subscribers can catch it here.

A few relevant links:

• Treasury's analysis on Working for Families;
• NZ Income Survey;
• Claudia Goldin's Presidential address, and related discussion;
• A few other tidbits from Offsetting on the gender wage gap.

Pay gaps, computers, and the public sector

The NZ EEO Commissioner, Jackie Blue, is unhappy about the gender wage gap in the public service.

The woman in charge of equal employment opportunities is set to "crack the whip" on government departments for failing to close the gender pay gap.

Women in the public sector are paid an average 14 per cent less than their male counterparts, despite making up 60 per cent of the sector's workforce, according to analysis by Equal Employment Opportunities Commissioner Jackie Blue.

"I'm very unhappy with the public service. They should be doing much better than they are. They're not taking it seriously - some are, but many aren't."
Preliminary analysis by Dr Blue's office of structural discrimination across 29 government departments shows little has changed in the two years since the last survey.

The gender pay gap was greater in 21 departments than the labour market average of 11 per cent. Seven had a gap of more than 20 per cent, including the Treasury and the Department of the Prime Minister and Cabinet.

The most recent American work suggests that the pay gap that remains after correcting for differences in education and time out of the workforce is largely due to gender differences in valuation of non-pecuniary benefits like flexible work environments and differences in the rewards for putting in long hours.

Paul Walker notes Claudia Goldin's excellent AER Presidential Address on the topic. I had a chat with Kathryn Ryan on Nine To Noon a couple of months back about the empirics around natality and baby bonuses; I'd then suggested she might have a chat with Claudia Goldin. Goldin was on with Kathryn a couple of weeks ago, discussing the evidence on the gender wage gap. While there's the general impression that the public sector is pretty 9-to-5, a lot of the senior bureau staff and senior analysts will be putting in rather more than that. It isn't implausible that the same differences found in the US will be driving things here. The Herald piece notes that much of the public service pay gap comes down to different occupational choices:

"The reality is the gender pay gap in the public service is heavily influenced by the relative pay rates in occupation groups where women are over-represented, such as clerical and administrative roles, contact centres and social, health and education workers," [Deputy State Services Commissioner] Ms Beatie said.

It could be useful for the EEO Commissioner to listen to Kathryn Ryan's interview with Claudia Goldin.

In related news, a new paper in the AEJ: Applied Economics [ungated] suggests that much of the decline in the US gender wage gap comes from the rise in PC adoption. Manual labour, more frequently undertaken by men, drops in value; returns to education rise.

Moar kids

Overshadowed (predictably) by Lorde's Grammy wins was Labour's policy announcement. Should Labour form government, they'd like to pay parents $60 per week for the first year of the child's life. 95% of children are meant to be covered by the plan, which is not income contingent for the first $150,000 of family income but abates to zero immediately at $150,000. Parents of 2 and 3 year olds would also receive up to $60 per week. Those with family income less than $50k/year would receive the full $60/week for the extra two years, abating to zero by $80k annual income for those with one child and somewhat more slowly for those with two or three children.

The policy will likely increase fertility rates. We can look to two decent studies for evidence. First, Josh Gans and Andrew Leigh found that an Australian baby bonus programme induced a very strong timing effect: women delayed giving birth until 1 July, when the programme came into effect. This suggests that people do respond to these kinds of incentives, but it doesn't tell us whether the programme increased total fertility or just affected the timing.

Of more relevance for present purposes, Kevin Milligan found that Quebec's baby bonus increased fertility in Quebec relative to that in other provinces for the duration of the programme. For those without NBER access, here's the CD Howe summary. Canada's federal system provided him a nice way of running difference-in-difference to be able to infer causality. While the programme mostly paid families for children they would have had anyway, he did nevertheless find a substantial effect on fertility. He writes:

...the estimates suggest a strong, positive, and robust impact of the policy on fertility. In the model containing the full set of control variables, the fertility of those eligible for the new program is estimated to have increased by 12 per cent on average, and by 25 per cent for those eligible for the maximum benefit

As the Quebec programme's intention seemed to have been to produce more Quebecois,* Milligan figured out the cost per child-who-would-not-otherwise-have existed. It cost about $15,000 per additional child. Is that value-for-money? Depends what you think a child is worth. I think life is worth at least that much to each of the children who wouldn't otherwise be born, and that New Zealand is so far below optimal population size that it would be a bargain, if results here were similar.

Milligan also found some really interesting demographic effects.

Suppose your model of the world is as follows. Poorer people have lower IQ on average and are more likely to be income-constrained against having another child. We might then expect that a lump sum baby bonus, like Labour's, and like Quebec's, that doesn't scale with income, would differentially encourage poorer people to have more children. Then, this.

Milligan found instead that the greatest policy effect on fertility was in the middle income ranges - those earning $50,000 or so in Canadian dollars at the time. The poorest groups didn't see much fertility increase. Milligan speculates that the group either didn't know about the policy or were farther away from being at the margin where another child would be desired. There also was less fertility response among rich cohorts, for whom income would not have been the binding constraint anyway.

Some of the demographic separation could be due to the structure of the Quebec programme. After the programme was made a bit more generous in 1992, it paid $500 on the birth of a first child. The birth of a woman's second child also generated a $500 payment as well as another $500 on that child's first birthday. Things got far more generous for larger families: third and subsequent children drew twenty quarterly payments of $400 ($31/week) for a total of $8000. Milligan then, as expected, finds the largest effect on the likelihood of a woman's bearing a third or further child. Where the very poorest women cannot afford a first child, a $500 payment is unlikely to change things much. Only those who were more affluent to start with could afford to reach the hurdle necessary for accessing the payments for very large families.

We then shouldn't conclude that the differential fertility effect will necessarily hold under Labour's proposed scheme, which provides fairly generous payments for any newborn regardless of parity. As Labour's proposed payments for children after their first birthday abates quickly for families earning more than $50,000, we might expect a humped effect where there would be little effect on fertility decisions among the very poorest cohorts, rising effects up to $50,000, then attenuating effects thereafter. I wouldn't be surprised to see negative fertility effects in the upper-middle income cohorts due to income effects, but that would depend on how Labour finances the scheme.

The tldr; summary then:

• Labour's programme is near certain to increase total fertility rates, regardless of how much snark leftie folks on Twitter want to throw around about "Oh, would you want to take care of my kid for $60 a week then?" They're nuts to suggest it won't happen. You don't need the programme to cover all the costs of child-rearing - you just need that some people had just barely decided against having a(nother) kid due to the cost and that these people change their mind under the new programme. I hate seeing tweets from people who know better pretending the opposite. You're not idiots - why pretend to be idiots on Twitter? Go read Kevin Milligan. 
• We should perhaps worry less about "Oh, well, poor people are just going to have a pile of kids in order to draw the payments then, and anybody who'd decide to have a kid on that basis is exactly the kind of person who shouldn't be a parent." Milligan's results suggest that they don't, but we need to be very careful on that one as we just don't know what would have happened if the Quebec payments had been really generous for first and second children. It's possible that Matthew Hooton's right here and that there will be a strong fertility response among the poorest cohorts. 
• One way of perhaps checking which is more likely here to hold: does anybody yet know whether the changes to the Sole Parent Benefit had any effect on fertility? The rules recently changed such that women can defer the Sole Parent Benefit's work requirements by having an additional child every few years. 
• It would be worth MSD's keeping an eye on whether there's been this kind of effect. It would be pretty bad if the programme really did start creating kiddie-mills.

Other observations:

• Bill Kaye-Blake is right that the $150k family income cutoff is a bit odd. 
• I would love to see somebody put together a new Effective Marginal Tax Rate schedule incorporating Labour's proposed changes. The abatement rates from $50k-$80k household income are pretty sharp. We'd also see a very large EMTR spike at $150k: the $1 that pushes you from $149,999 to $150,000 in family income will cost you $3120 if you have a newborn. 
• Because of the large EMTR effect based on household income for those with newborns, I expect this will induce many second-earners, predominantly women, to spend longer outside of the labour force on the birth of a child. We'd see this in higher-earning couples, where the second-earner's income pushes family income above $150,000, and in middle-earning couples in the high abatement ranges. 
• The plan's proposed extension of paid parental leave to 26 weeks will intensify the effect above. Consequently, small businesses may be more reticent to hire women of prime childbearing ages. The costs of bearing a half-year's likely leave are not trivial for employers. The wage gap will consequently increase. Sadly, National's jumping onto this bandwagon as well. 

* Milligan infers this from the increasing payment schedule. As there are economies of scale in child-rearing, a model designed simply to compensate for child-related costs would have a large payment for the first child, with lower payments thereafter. Political discourse at the time, if I recall correctly, was entirely pro-natalist. Citizens and Permanent Residents had access to the payments, immigrants who hadn't made PR weren't. Quebec nationalists later blamed "money, and the ethnic vote" for the failed sovereignty referendum; breeding more pure-laine was a potential solution.

Maternity penalties: average is over edition

Matt Yglesias makes the case that, in a world in which workers are not interchangeable widgets, compulsory jury duty places an unreasonable burden on firms.

The implicit model of jury duty is one in which workers are basically interchangeable widgets. John is going to be missing for a while, so Jane will pull some extra shifts to cover for him. Since different people have different tastes and circumstances, the odds are pretty good that at any given time there's someone on staff who actually prefers to work longer hours in exchange for higher pay. Consequently, John is put out a bit but basically the company trundles on.

There obviously are some work roles that function this way. But there are a lot of work roles that don't. Even when we have people doing similar jobs, that doesn't mean anyone can actually double-up. If I'm on a jury, Weigel can't write my blog while also writing his blog. He could try to write his blog during regular business hours and then write my blog at night, but all the news will have already happened. By the same token, even if other editors step up to fill in for a missing editor and work longer hours the reduction in parallel processing still means articles will be published more slowly. The impact of the slowdown then trickles through the entire enterprise and lowers everyone's productivity. And I think we're hardly unique in this regard. People are collaborating in complicated ways in workplaces all across America, and while nobody is irreplaceable, it's also fairly unusual to just be able to swap a given person in and out without consequence.

Jury duty is random-draw. At least here in New Zealand, you'll have a few weeks' notice and trials might last a fortnight. If your employer can make a case that you're irreplaceable during that period, then you can get an exemption or deferment.

Imagine that jury duty were longer: three months to a year if called. Imagine further that one identifiable group of employees - half of all potential employees - would never ever be called for jury duty. And imagine further that, among those in remaining group, there were a pretty high chance of being called for jury duty a couple of times while in their 20s and 30s. And there's no opportunity for employers to get deferment for critical staff. Would you, as an employer, lean towards hiring the first group for task-critical roles, even though you think jury duty is really valuable and important?

Maternity leave isn't jury duty: it's chosen by the employee. We then have two effects going on. Women who wish to have children disproportionately select into jobs allowing flexible time arrangements and less overtime. And, employers fearing incipient maternity leave may be reluctant (all else equal) to assign women of higher childbearing likelihood into mission-critical and higher paying tasks.

Claudia Goldin's superb Presidential Address at the AEAs makes the case that the structure of mission-critical jobs in some industries is what's driving the gender wage gap. Women disproportionately prefer workplace flexibility over pecuniary benefits. And some high-paying jobs require massive time investment.

Residual differences by occupation in earnings by gender, I will demonstrate, are largely due to the value placed on the hours and job continuity of workers, including the self-employed.²³ Individuals in some occupations work 70 hours a week and receive far more than twice the earnings of those who work 35 hours a week. But in other occupations they do not. Some occupations exhibit linearity with respect to time worked whereas others exhibit nonlinearity.²⁴ When earnings are linear with respect to time worked the gender gap is low; when there is nonlinearity the gender gap is higher. 

... In many workplaces employees meet with clients and accumulate knowledge about them. If an employee is unavailable and communicating the information to another employee is costly, the value of the individual to the firm will decline. Equivalently, employees often gain from interacting with each other in meetings or through random exchanges. If an employee is not around that individual will be excluded from the information conveyed during these interactions and has lower value unless the information can be fully transferred in a low cost manner.

The point is quite simple. Whenever an employee does not have a perfect substitute nonlinearities can arise.²⁵ When there are perfect substitutes for particular workers and zero transactions costs, there is never a premium in earnings with respect to the number or the timing of hours. If there were perfect substitutes earnings would be linear with respect to hours. But if there are transactions costs that render workers imperfect substitutes for each other, there will be penalties from low hours depending on the value to the firm.

I've not seen a better single compilation of the existing evidence on the gender wage gap.

Career Lesbians

First, a short summary on the gender wage gap. Men on average earn more than women. When we correct for differences in education, work experience, industry, and time outside of the workforce, the differences are much smaller but they do not disappear entirely. There's been a lot of work done on that women often choose professions or jobs that provide less in salary but more in non-pecuniary benefits like more generous leave, easier flex-time arrangements, or fewer expectations to work more than a 40 hour week. These differences in choices can result in differences in observed pay packets even in the absence of any wage discrimination. See the posts linked at the end for some of the evidence.

One hypothesis that's intrigued me: while wage regressions will adjust for number of children, and find that having kids hurts women's wages, they don't really account for employers' forward looking expectations of job separation or time out of work due to fertility. If an employer is weighing up two job applicants, one of whom, statistically, has a much higher chance of taking 3-12 months off on maternity leave sometime in the next three or four years, then the employer might shy away from offering the kinds of jobs that require strong job attachment - the same kinds of jobs that lead to progression through the ranks and to the upper echelons of industry. Worse, this can easily lead to the kind of bad equilibrium that Glen Loury posits for human capital acquisition across races: if you expect that employers will reckon you'll drop out of the labour force to have kids and that you'll have a hard time getting to the top because of it, why make the human capital investments that would be necessary to do it?

Occupational segregation arguments then start having more force: women are disporportionately employed in industries with lower skill depreciation rates and lower penalties for temporary workplace separation, or in the public sector, not only because those sectors are friendlier to families but also because making investments to go into other industries are riskier where you can't credibly signal that you won't separate from the workforce.

My Masters student, Hayden Skilling, presented his thesis to the Department on Tuesday. Lesbians earn more than heterosexual women. Some of this is due to occupational choice, some of it is due to having had fewer kids. But some of it also seems due to that lesbians are less likely to have children than are heterosexual women. Hayden shows this in a few ways. First, the age patterning of realised fertility outcomes among lesbians and heterosexual women explains some of the wage gap between lesbians and heterosexual women after correcting for a host of other variables in Census and ACS data. Second, state-level policies that differentially affect fertility by sexual orientation also affect the wage gap between lesbians and heterosexual women in the predicted direction. For example, mandatory insurance coverage for IVF makes it easier for heterosexual couples to have kids; where the insurance coverage typically requires conception failure subsequent to an extended period of natural attempted conception, married heterosexual couples will be at an advantage. Similarly, legislation making it easier for homosexual couples to form stable legal recognized couples make it easier for them to have or jointly adopt children. Both of these affect the wage gap in the predicted ways.

I'll be blogging more of Hayden's work after he's given me the final draft. In the meantime, here's another great piece of evidence supportive of the basic hypothesis we've been exploring. Tobias Schmidt points me to this piece over at IZA. Stijn Baert at Ghent ran an audit study where Belgian employers were sent vitae for homosexual and heterosexual women. Apparently it's not uncommon in Belgium not only to list your marital status but also your partner's name on your vita. This is great because it lets them provide a signal of sexual orientation that doesn't simultaneously signal, for lack of a better word, stroppiness: membership in a gay rights organisation as an "other interest".

They found that there was no discrimination against homosexual women in general but that younger lesbians are more likely to get positive callbacks from potential employers than are young heterosexual women. They also expect this is due to employers' expectations about fertility.

If an important part of the pay gap is employers' expectations about the costs of employees' fertility decisions, then optimal policies to address the pay gap and optimal policies to support family formation will vary. For starters, we should be reticent to impose costs on employers who hire women who then go on to form families.

Previously:

• Wage gaps and Maternity
• Mommy track vs partner track
• That pesky pay gap
• Odd Japanese labour markets
• Fertile Employment
• More pay gaps
• Parental leave and benefits
• Gender pay gaps and nonpecuniary benefits
• The lesbian pay gap
• Wage discrimination: the evidence
• Critiquing the Ministry of Women's Affairs paper on pay gaps (and here)

Wage gaps and maternity

Another for the "the gender wage gap has less to do with sexist employers and more to do with productivity characteristics" file: having children reduces the intensity of market work. In a nice study on Norwegian data, where effects are identified by comparing against women who had miscarriages, women having children see earnings reductions of 16%, driven mostly by reductions in the number of hours worked rather than by labour force exit. HT: Ole Rogeberg

We find that each of the first children has a large impact on female earnings, which on average are reduced by 16%. The largest decrease in earnings are driven by women with average earnings reducing their labor supply and thereby earn less. There are small effects on earnings for those who are marginally employed or women in the higher end of the earnings distribution. Labor force participation is not affected much by having children, indicating that in the Norwegian context, the combination of market work and family obligations is clearly feasible. The intensity of market work is however reduced. Women work on average 2 hours less per week per child, and the effect does not decrease (much) when the child grows older. Motherhood therefore plays a significant role in explaining female part-time work. We find no evidence of an adverse health effect of having children as neither sickness absence not disability seem to increase due to motherhood.

Since they're identifying on the difference between women having miscarriages and those with successful births, the estimates will be the partial effect of having had a child rather than the total effect of both preparing to have had a child (perhaps reducing investment in on-the-job training) and of having the child. The total effect will be larger.

Previously:

• Mommy track vs partner track
• That pesky pay gap
• Odd Japanese labour markets
• Fertile Employment
• More pay gaps
• Parental leave and benefits
• Gender pay gaps and nonpecuniary benefits
• The lesbian pay gap
• Wage discrimination: the evidence
• Critiquing the Ministry of Women's Affairs paper on pay gaps (and here)

Fertile employment

One potential explanation for lower female wages: employers fear that female employees will take maternity leave and condition wage and employment offers on that risk. I'm not saying it's a good thing. It's almost certainly illegal. But it's potentially a thing employers do. My Masters student Hayden Skilling is currently investigating whether this can explain any substantial part of the unexplained gender wage gap; I'll report on results later this year.

In the meantime, here's an interesting bit of evidence. Spending a lot of time in casual employment rather than full-time employment increases the risk that a woman is childless by age 35. 

MAIN RESULTS AND THE ROLE OF CHANCE The likelihood of childbirth by around age 35 was reduced for every year spent in casual employment, irrespective of socioeconomic status, partner's education and parents' birthplace. The likelihood was reduced by 8, 23 and 35% for 1, 3 and 5 years spent in casual employment, respectively.

Now run the thing in reverse and assume rational expectations. You're an employer. Two equally capable potential employees are up for a job. One comes with a non-trivial risk of taking maternity leave. The other is male. In case of maternity leave, you need to sort out cover and you can't guarantee that the employee returns. So you're taking on risk and you need to be compensated for taking on that risk.

There exists no real way for young women credibly to signal that they have no plans on having children, or that they promise to work for n years prior to any childbearing. It's generally illegal for the employer to ask, and both the candidate and the employer knows that any promise made is unenforceable - in fact, the employer would likely get in trouble for trying to enforce it.

Again, I'm not saying any of the resulting equilibrium is in any way good or desirable. If you wish to promote gender equity, though, you have to set incentives such that good results obtain. Maternity leave mandates that are costly to employers will likely result in fewer women being given positions that draw maternity coverage. Employment subsidies proportionate to the cost of the risk imposed might socialise the costs more efficiently.

Previously: Parental Leave and Benefits.

More pay gaps

My colleague Andrea Menclova responds to some of the recent foofarah about the latest data on the male-female wage gap in New Zealand. From her email:

The New Zealand Herald has recently reported that the New Zealand gender pay gap has increased in the year to September by 1.3 percent (from 12.85 to 14.18 percent) and that it is the largest it has been in a decade. These figures are based on the latest Statistics New Zealand Quarterly Employment Survey and have lead to considerable dissatisfaction for some. Angela McLeod, spokesman of the Pay Equity Challenge Coalition, said for the Herald that "The best [the gap] has ever been using the measurements we use is 12 per cent. It's still a gap and it's still unacceptable." And Lucy Johnston from the University of Canterbury told Scoop that “active intervention is needed to address these inequalities”

Why?! Gender pay gaps arise for a number of reasons and only some may be worrying. Unless we know the reasons for a pay gap, we can’t conclude anything about its desirability. See the famous Oaxaca decomposition and the ton of research that has been done on the topic. For example, women often choose part-time employment as it suits their personal lives (they may prefer part-time employment to no employment but would prefer no employment to a full-time job). It honestly frightens me when people draw angry conclusions from raw numbers. Angela MacLeod admitted the Coalition was not sure of reasons for the pay gap change but said “a higher proportion of women in part-time work had not helped”. Helped whom? Helped what? Help us someone! Help us write better commentaries on such issues.

I agree entirely. In particular, it is nonsense to compare average wages without correcting for education, time outside the workforce, industry, or full-time and part-time status. The QES's tables only give a gender breakdown for average hourly wages, though I'm sure Stats NZ could provide more detailed gender cross-tabs on request. Part-time workers typically earn less per hour than full-time workers; if more women than men take on part-time arrangements voluntarily, and we don't correct for that, we overstate the gap. Same for correction for industry and the like.

The New Zealand Income Survey gives us a somewhat different picture - again, only comparing raw figures. For the June 2012 quarter, average female hourly earnings from wages and salaries were $23.35 to the $26.75 earned by men: men earned 14% more on average. But if we look at median hourly earnings, the gap is only 10%: $22 for men compared to $19.95 for women. And, better, the NZ Income Survey lets us compare full time and part time earnings. Again, the raw gaps above were 14% in averages and 10% in medians. If we compare full-time workers, men earn (median) $23 per hour to the $21.58 earned by women: a 6% gap. Comparing part-time workers, men earn $15.50 per hour and women earn $16.25 per hour. So, in the Income Survey, you can abolish almost half of the wage gap by considering only full-time workers, and reverse the darned thing if you look at part-time workers.

Table 11 of the last NZ Income Survey is awfully telling. Where twelve percent of men work part-time, thirty-five percent of women do. Cohort differences in the choice to work full or part time drive the bulk of the observed wage gap. And it would be surprising if correction for industry, work experience, time outside of the workforce, and education didn't wipe out most of the rest.

Professor Johnson, in the Scoop press release, cites other research:

Research, including some conducted by Professor Johnston, has shown the continued existence of such stereotypes and the impact they have on recruitment and promotion processes - often without the awareness, or explicit endorsement, of the evaluators.

Women were evaluated less positively than men with identical qualifications and experience when applying for jobs typically filled by men, such as management positions. Women are more likely to be successful in such applications if they dress and talk in a more masculine way – wearing a suit rather than a dress and having a deeper voice, she said.

``Role models and encouragement are essential for young women and we must make every effort to promote and endorse successful women and we must ensure that to become successful women do not need to become `like men’.

It's true that in audit studies, we can find employers who discriminate. But, in equilibrium, it's not likely that they drive aggregate wage differences. There's also evidence that women are less aggressive in salary negotiations than are men, which also contributes to wage gaps. But is that really discrimination?

Finally, suppose that you ran all the regressions and controlled for all the human capital characteristics and still found on wage decomposition that young women earned less than comparable men. There's still something for which you haven't corrected: an employer hiring a young woman faces greater risk that his employee will take 3-12 months of leave. Sure, maternity leave is paid for by the government rather than by the employer, but the employer still bears costs in accommodating the leave.

There will be more here sometime in the next few months on maternity risk and wage gaps. Stay tuned.

Parental leave and benefits

Put yourself in the place of an employer faced with two excellent and similarly qualified candidates for a position. You'd be very happy with either. But one, a young women, comes with maternity leave risk. If she decides to have a child, you will bear costs of worsening productivity over the course of the pregnancy despite her best efforts, costs of finding a temp worker to cover her position while she is on maternity leave, and the uncertainty of whether she will indeed return when leave concludes. She may also wish to move to flexible time arrangements on return. The other, male, doesn't. You're running a small business where losing a skilled worker for a short period is a very real burden, even if somebody else is paying her salary while she's on the government's paid parental leave scheme. Whom do you choose?

Cactus Kate makes the case:

I will never apologise for being honest enough to say that I don't like employing women of child bearing age especially if they have just got married or are loved up with a boyfriend because you know the next step. Babies. It is bad enough for a small business losing a staffer for 12 (as it is in HK at 4/5th pay) or 14 weeks, try employment laws where you can't sack a woman while she is pregnant (that's nine months of secure employment) even if she is hopeless at her job or not turning up, try the woman who at 11 weeks and a few days of investment and patience waiting for her to return to work, then hands you their bloody resignation. Try co-workers having to pick up the slack while she is away as you can't afford a temp.

In many cases they cope fine which means on return to the workforce it's pretty clear the new mothers position can be made redundant anyhow. This is the reality of parental leave. It indicates pretty quickly to an employer just how crucial a woman is or isn't to an operation. In many ways it's a rehearsal for redundancy.

We can wish that employers would willingly take on these costs. And many who do find that they wind up with a very loyal and committed employee if they do. But it is a risk. And it's a risk that, at least in data from a very nicely designed field experiment in France, has employers shy away from employing women with high maternity risk. Lower employment isn't the only way that the policy's costs can be shifted; Jon Gruber finds that costs of mandated maternity benefits through US employer-provided health insurance tends to be borne through lower wages for women [HT: @KevinMilligan]. And it's a pretty plausible candidate explanation for the lesbian pay gap; my excellent honours student, Hayden Skilling, is investigating this as his honours project this year.

New Zealand currently requires employers to hold a woman's position open for a year if she takes maternity; the government provides paid leave scaled to the woman's salary (and subject to a relatively low cap) for 14 weeks. The Labour Party proposes extending this to 26 weeks; the bill has been drawn from the ballot. It is likely to pass first reading, but likely to be killed afterwards because of the budgetary implications.

Were it implemented, I'd expect that the policy will increase the amount of time that women spend on maternity leave. In Canada, Baker and Milligan found that a doubling of the compensated maternity leave entitlement significantly increased the amount of time women spent on maternity leave.* Employers will bear costs despite that the paid leave entitlement is covered by IRD: it will be harder for employers to cover leave internally and so more of them will have to find replacements willing to work on temporary contracts. A longer time outside of the workplace means skills have longer to erode. Women are also more likely to want to return on part-time or flex-time arrangements after longer periods outside of the workforce; Schott finds that the American Family and Medical Leave Act increased women's likelihood of returning to work part-time rather than full-time.** Finally, we may expect increased labour market participation among women anticipating maternity leave, but also increased employer reluctance to take on women of higher maternity risk except at lower wages. But, I don't have a great sense of the incremental cost above existing leave entitlements; what's true at the margin might not cash out as much in the aggregate.

If Labour's economics were just a bit stronger, they'd be trying to couple their policy with some kind of compensation mechanism for employers whose workers take maternity leave rather than embedding the lump of labour fallacy into the bill's explanatory note:

Extending paid parental leave from the current entitlement of 14 weeks to 26 weeks would support families and also create jobs across the economy as employers engage staff to replace those on paid parental leave. As the majority of paid parental leave is uplifted by women, it has the added benefit of creating jobs in areas of the economy where women work, while supporting families and the well-being of children.

Why not advocate for a maximum 35-hour work-week to encourage employers to hire more temp workers to cover the work not done?

* While the Canadian change increased breastfeeding rates, one of the NZ bill's other stated purposes, it had no effect on child health outcomes.

** While Schott finds increased workplace flexibility encourages post-natal female employment, we might reasonably worry that increased likelihood of moving to part-time or flex-time arrangements reduces an employer's willingness to invest in an employee's human capital or to take on the worker in the first place except at lower wages.

Note: updated a couple of times for clarity and to add links to a couple of helpful tweets from Kevin Milligan and Frances Woolley.

Update 2: @askessler recommends this IZA piece showing no long term benefits to kids from paid maternity leave extensions in Germany.