Saturday, 16 April 2016

Trade Unions, productivity and wages: cause or effect?

PaulC156 left some interesting comments on an earlier post, to wit:

1. Stagnant median hourly pay in the US over the last few decades coincides nicely with the steady erosion of unionised labour in the States:

2. Strong unions improve productivity. Plenty of evidence in a six nation study here:

3. Though not good for profitability… some good anecdotal on the UK motor industry from this article by Professor David Bailey of the Aston Business school here:

4. CEO compensation is inversely correlated with stronger union presence:

On the facts, I don't dispute any of this, but I question which is cause and which is effect.

From general knowledge, there are certain industry-specific factors which say which people are more likely to belong to a trade union (in the traditional sense) or a guild (for example the British Medical Association, the Institute of Chartered Accountants, the National Farmers Union etc).

a) The employer is a large employer and especially a monopoly employer/provider.
b) The business can't go bankrupt (because it is state-run or has a monopoly).
c) There are fixed pay scales.
d) Workers are highly skilled and/or less easily replaceable.
e) There are barriers to entry to that job, profession or business.

Some of these categories overlap and some are mutually exclusive. They also operate on different levels (employees vs employers or industry vs customers). But to give some examples:

Historically, the first trade unions were local guilds, higher skilled people who could club together to try and exercise a bit of monopoly power and raise barriers to entry. If you look at the progression to trade unions, it was the higher skilled trades first and domestic workers were the last to be unionised, which never really took off. Higher skilled people are by definition more productive than unskilled people.

In London, buses are run by competing private companies (although regulated by the TfL) and the Tube network (a monopoly) is run by TfL (a government body) directly. Each bus company sets its own pay levels but train drivers are paid pretty much a flat rate. So bus drivers are less unionised than train drivers.

Doctors are very expensive to train up or poach from abroad, the NHS is a state-backed near-monopoly provider, so all doctors are in the BMA (as far a I am aware) and they negotiate from a position of strength.

The importance of pay scales is, if you know that you get paid the same as thousands of co-workers (more likely in large organisations and the public sector), there is no harm in putting up a united front and asking for a collective pay rise. If pay is set individually according to each worker's skills, experience and working hours (more likely in small businesses), it is every man for himself and less likely to be unionised.

The importance of the employer having a monopoly position is, if he doesn't, pushing for higher wages will bankrupt him, as happened to some extent with the UK motor industry in the 1970s or 1980s; all that happened was cheaper cars were imported from abroad. If the government had imposed an import ban on motor vehicles, then the British motor industry - and trade unions - would have survived (producing crapper cars for higher prices and paying much higher wages).

Stock Exchange rules and the Companies Act mean that larger companies have to pay for an audit whether they like it or not; although auditing is not particularly difficult, it can only be carried out by Chartered Accountants (although Certified Accountants rank equally in law, they are brushed off as second class). This is bureaucracy for the sake of bureaucracy, so those at the top can cream of rental income and impose a strict hierarchy and lengthy training period on new entrants who want to get their snouts in the trough.

There is a limited amount of farmland so there is a limited number of farm owners, they can club together to form the NFU and put pressure on the government to keep the CAP subsidies flowing. Farm workers on the other hand are relatively easy to train up, it is part time and seasonal work and they are unlikely to be unionised.

The importance of collecting rental income/monopoly profits is that it frees up plenty of your time for campaigning, organising and putting pressure on the government to rig things in your favour. Which is why the most vocal groups are people like pensions companies, who live off the tax breaks nominally given to savers.

As a slightly separate topic, the pay of senior people tends to be much higher in businesses which collect 'rent'. So Chartered Accountants (as explained above), home builders/land bankers; oil and gas companies, banks and insurance, entertainment companies, top footballers etc. Top executives skim off some of this rent from their shareholders; if they can depress wages, this boosts profits so gives them more to skim off.

To sum up, it is not that trade unions or guilds lead to higher productivity and wages, it is that higher skills/wages/barriers to entry/profits (whether in the 'free market' or in a monopoly or state-backed organisation) are the fertile ground on which trade unions and guilds (BMA, ICAEW, NFU etc) can flourish.


Lola said...

Agreed. Real world experiment. In my own trade - retail financial advice services - the FSA/FCA introduced a thing called the Retail Distribution Review which, among other things, mandates minimum training standards. The effect was to force a decline in the adviser population of about 16,000 to 20,000 in number, about 20% to 30% of the existing adviser population. At the same time the private representative bodies won the Royal Charter special privilege.

And what has happened to the average adviser remuneration? Well, what happens when you introduce a scarcity and increase arbitrary barriers to entry? The price goes up. I would estimate the remuneration and profitability have risen by about 50%. It's pure rent seeking. Indeed the average adviser is more 'productive; in the sense that he can generate more fee income in a given time, but the quality of that advice has hardly changed at all. And, FWIW, the vast majority of my peer group remain woefully ignorant of the economics of this.

Mark Wadsworth said...

L, ta, that is a more or less perfect example.

Bayard said...

"Farm workers on the other hand are relatively easy to train up, it is part time and seasonal work and they are unlikely to be unionised."

Which is why the NFU is the National Farmers' Union and not the National Farmworkers' Union. I had, without thinking about it too hard, assumed the NFU was a trades union, but of course it's not, it's a landowners' union, the same as, but just less posh than the CLA.

Shiney said...


I can give you a very specific example of this. I bought my business out of receivership 13 years ago. Previously it was a division of a large industrial company that, historically, had dominated the town where it was located.

At one point in the 70s the parent of the parent employed 4-5,000 people in the locality (the direct parent about 500). By the time it went bust the parent was down to around 200-250 and the big factory up the road which had owned it was virtually closed - it is now... a pile of rubble and contaminated land is all that is left.

We took our division on with 42 people.

When we took over the workforce was pretty much 100% T&G (Unison now?). The former parent's workforce were all T&G/AEEU (?) and even the the staff were in the MSA.

Within a year the whole union recognition issue had gone away... we were now a small business fighting to survive (if you've ever done an MBO/Start-up/recovery hybrid you'll know what I mean) and that was the least of the workforce's worries. Several of the long standing employees kicked against this and lost.... but hey, that's life.

As for productivity? Well, year one we did 8m units with 42 people, last year 18.5m units with 51 people. And average wages/skill levels are definitely higher because the amount of 'kit' we use is higher (we have more engineers/technicians than before). That is what drives average wages... NOT unionisation or otherwise.

So in summary - my experience is that unionisation of the workforce or lack thereof has no effect whatsoever on increased productivity or average wages, its about skill levels and mechanisation.

DBC Reed said...

@Shiney: was the original parent company British Cellophane?

Mark Wadsworth said...

B, glad to have cleared that up

Sh, thanks for anecdotal. So in your case factors a and b no longer applied which made all the difference.

The Stigler said...

Unions existed because of two barriers: on the one hand, skilled people had to work at one place. On the other hand, employers had few choices but those people. When you had few choices of where to work, you had to use collective power to get a better deal. You could threaten a strike and the employer could pay up or make no product.

Along comes the motor car and better trucks and better roads, and well, the employer can now go all over to get bits of his things made. On the flip side, the employee can now go elsewhere. People just don't need unions like they once did. You can get a better deal? Leave, work elsewhere. The problem is that everyone's read the cause as being politicians, which is just wrong as union membership has fallen everywhere in the West.

And yes, the reason why Airbus and Rolls-Royce still have strong unions is that both employer and employees have this monopoly thing. Rolls-Royce needs people with very specific skills and those people with those skills have few other options. You don't get unionisation with F1 teams because there's a competitive market for people in the Oxfordshire/Northants borders with a couple of dozen teams (plus those skills are portable in other sorts of companies outside).

The main reason unionised skills pay more is that yeah, people have skills. There isn't this monopoly/monopoly with cleaners and staff in Starbucks. Don't like it in Starbucks? Go work in Costa or Nero. And even historically, those people weren't unionised. It was generally factory workers or professions, and in those days, factory work was "higher" skilled.

The heart of the problem is simply people without much in the way of valuable skills. We can only deal with that in one of two ways: paying more benefits or getting people more valuable skills. And it really irritates me that almost no-one seems to care about cuts to FE colleges, when the cost of these is small, and the results are very good, while everyone's crying about the cost of degrees, most of which are just a very expensive signalling mechanism. A lot of people who do photography or psychology degrees would do better to go and learn to be a chef or a welder. Plenty of demand for both. Photographers? There's sodding tons of them. I know really good ones and it's a struggle to make a living.

Derek said...

One worrying trend in the US recently is for unions to back a push in some place for a high minimum wage, $15 an hour or whatever, and then negotiate an exemption for their members. The latest example of this is in Los Angeles. As a result employers can get union labour cheaper than non-union labour and the unions can get more members. Too bad for the low-paid, non-union workers who are being forced out of the labour market by this ploy. Unless they join a union.

Mark Wadsworth said...

TS, more good examples, ta.

D, now that is spiteful!

Shiney said...

@DBC... Nope. What made you think that?

@MW... Exactly

@TS - generally (I know there are exceptions) manufacturing jobs tend to be more skilled than in days of yore as there is more kit - that is why we make roughly the same amount of 'stuff' in this country as we always did.... its just we need fewer people to babysit the machines and fix them when they go wrong. And the proportion of 'blue collar' to 'white collar' roles in most manufacturing companies is now less heavily skewed than it was..... which means less unionisation (in general).

Mark Wadsworth said...

TS and Sh, I think that's point f, people are less likely to be unionised if they can switch easily between employers/jobs.

Anonymous said...

"if he doesn't, (have a monopoly position) pushing for higher wages will bankrupt him, as happened to some extent with the UK motor industry in the 1970s or 1980s; all that happened was cheaper cars were imported from abroad." MW

Mark, I posted a link to a study (Warwick Uni) of Indian cotton mills spread across India and compared those with and without unions and the respective response to wage increases in those mills that were unionised:

"This paper uses a new data set of cotton mills from all regions in India. Unionisation and worker militancy differed greatly across the regions. I use the regional variation in unionisation to test if the militant workers in Bombay cotton mills were less productive. I find that regions with higher wages had higher labour productivity.These were also regions where the workers were unionised. The presence of unions did not lead to lower productivity. On the contrary, by raising wages the unions contributed to raising labour productivity in the region. This reinforces the argument
that the causality may go from wages to productivity."

Much of what you have written is fair and I agree that there is movement both ways in terms of causality. I will just including this for the sake of balance.
One also finds support for the hypothesis that stronger unions DO beget stronger investment/efficiencies in looking at Toyota. Heavily unionised in Japan it has become one of the most innovative and productive motor manufacturers in the world since it's early abysmalforay into the motor industry.
Higher wages, leading to more motivated staff, encouraging more staff involvement in implementing innovations all buttressed by typically longer tenure times as opposed to that which exists in non unionised factories. Greater emphasis on training, and a strong tendency for 'less managers' in such workplaces. All stemming from the more lateral management techniques where worker participation is encouraged as opposed to the fractured unionisation and 'us and them' culture that existed(exists?) in the UK. The class riven society that characterise post war Britain was arguably a hangover from the days of the British empire. It wasn't that unions were bad per se for British motor industry rather in a class ridden culture dominated by management ideas about top down hierarchical management techniques (as Thatcher called it- manager's 'right to manage') unions became a limiting factor. The problem isn't unions or the nature of the industry, as Toyota showed but rather the UK's rotten to the core class culture. To talk about Britain's poor performance in this sector (as a few have done here) in the 70's and 80's and the linkage with unionisation without acknowledging this reality, is highly misleading in my view.

DBC Reed said...

@Shiney re above. I thought you were based in Bridgwater and your mention of a contaminated site brought British Cellophane to mind: once the town's biggest employer.
I am loth to feed Lola's obsessions, but the decision to concentrate cellophane production in the loss-making plant in Tecumseh (USA) rather than in the profitable Bridgwater was entirely a political decision with the Statist Americans offering a bigger bribe than we did: we only offered tax breaks in six figures while the yanks offered millions of dollars up front.
From the purely practical point of view cellophane has commercial potential by being biodegradable.