Where has all the productivity growth gone?

Over on the Value Investing blog, James Carlisle asks “Where has all the productivity growth gone?”

Over the last 20 years or more we’ve been told that employee productivity – that is, us workers, has grown by something like 3% or more a year. Part of this has been due to productivity aids like automation, IT systems, removal of beaurocracy, less layers of management, and so on and on.

The commonly accepted wisdom is that IT systems (the PC on every desk, fuelled up with software from Mr Gates and others) have driven most of the gains.

I posted a long comment, which I reproduce below in slightly edited form, in which I argue that most of the gains are either illusory, delusional, or one-off in nature.


I expect a great deal of the claimed productivity growth to be illusory – the product of the fevered imagings of managers, economists, and the salesmen for IT companies.

We can split an analysis into 3 main activities in the broader economy:
- Administration, support and services
- Design and development
- Manufacturing

Administration, support & services

In general, jobs like waiting tables, doing tax, purchasing paper clips for the office and so on have not changed a great deal. A waiter can only walk so fast, doing tax gets more complex not less (though spreadsheets help to even the score a bit here), and purchasing is still purchasing.

The methods of doing these things might have changed a little: waiters have point of sale computer systems – but these primarily remove error, they don’t get orders in a kitchen faster, they don’t get meals cooked faster. A purchase order is still a purchase order. The typing pools for those disappeared in the 1980s, so IT systems have done the work for over 20 years. Some suppliers now allow on-line orders – but hasn’t this just replaced the phone call?

So in all these parts of the economy, which I collectively lump together as “services”, productivity gains have been pretty small. There may have been some one-off gains back in the 1980’s, but since then there is not really a lot more left to wring out.

Product Design and development

Once upon a time, before the days of the PC on every desk – in other words prior to about 1990 for you young ‘uns, things got done. Smart folk designed washing machines and bits of electronics, and bridges and buildings. Sure it was all done on drawing boards with ink, and the clever guys stood around a blackboard arguing the merits of their ideas. Paper based design documents were slow and painful to produce, and kept to the bare essentials.

Now we have computer systems that make it easier and faster to write software, design electronics, and design bridges, buildings, widgets and what-not. We have CAD packages and software source management systems, electronic circuits can be designed using software that’s free on a $1000 PC instead of needing a mainframe in a room full of white coated keepers.

At the same time, though, the complexity has increased (compare a car or washing machine of today and 20 years ago). In addition, the ligitious nature of our society means that more needs to be taken into account, more documentation produced and archived in case of legal action.

So there have been huge productivity gains in all the product design activities, across all the disciplines, but these are again mainly one-off in nature and are counterbalanced by demands for the thingies being designed to do more.

Then there are the costs – you need systems and processes in place to backup the data, run the IT systems, archive documents, repair breakdowns, install data lines, service the UPS, and so on. There are now a whole range of essential service provision jobs that did not even exist 20 years ago, and they need to be paid for from something.

Let’s also mention email – something that didn’t exist years ago. Sure, it replaced the inter-office memo. But you generally only ever made a couple of copies, you didn’t send then to 10 or 20 people including the big boss fellas just to make the recipient look bad. Modern managers can receive 20 to 200 emails a day, and it’s very easy to spend 2, 3 or 4 hours every single day just reading and responding to email. The modern miracle of email with instant delivery has been one of the biggest productivity-destroyers in the history of the universe!

And with the PC and internet on every desk, now it’s very easy for people spend their day at work playing solitaire and reading the news sites. I personally don’t think design and development productivity has increased, overall, AT ALL in the last 20 years. The combination of easy goofing off, tools to remove drudgery, greater support costs, and greater demands all balance out.

Finally, and fundamentally, you can’t make people think or innovate faster in spite of books by Bill Gates. Ideas come about at the rate and times that ideas come about.


There have been numerous gains over the years, with automation robots, and moving it all to low labour-cost countries. Sometimes the gains there are illusory too: for example, transport can eat up all the gains you made from a lower labour cost. And if that doesn’t stiff you, exchange rates can bite you instead.

But especially in manufacturing, any savings due to productivity drive competition. Competition allows one maker to offer a lower price and increase market share – and all manufacturers are forced to play a game descending the ever-decreasing spiral of  lower prices.

Here’s an example. I bought a dishwasher 20 years ago, and it cost me $1000 on discount. It died 2 years ago so I bought a replacement, which cost me $600 on discount. They both do exactly the same job, they are both base models without frills. In real terms the price paid for that dishwasher has more than halved in a period of about 15 to 17 years.

So any productivity gains in manufacture didn’t end up with the manufacturer, or their shareholders. Those gains materialised as lower prices and consumers took the benefit.


There have been big productivity gains in manufacture that go to consumers, small if any gains in design and development, and little or no real gains anywhere else.

I’m also deeply suspicious of those who claim general long term productivity gains, especially over the last 20 years. I get that Mandy Rice-Davies feeling: “well, he would say that, wouldn’t he”.


This is very interesting. It’s analogous to the question, “Why does the paperless office use more paper than the old type?”

I have a personal theory that people have a limit to how much useful creative work they can do in one day. It varies from person to person but it limits what can be done. No amount of technology is going to change that.

Comment by Jack | March 11th, 2009 7:34 am | Permalink

This is so well-written Wally – why not send it into the Traumatiser so they can eduhmucate themselves and their readers?

Comment by Kath Lockett | March 11th, 2009 1:22 pm | Permalink

I spent more than half of my working life [40 years and 3 months] designing and supervising the construction of major assets for ETSA. At the beginning, there were no environmental controls or guidelines and to build a major high voltage power line took about three years from decision to build to commissioning. 20 years later it took over 6 years and the bulk of that time was spent arguing one’s case with nervous politicians and in gaining environmental approvals and arguing with greeny groups. All of this was a waste of time as the asset had to be built anyway. I mean how do you get power away from a power station if there is no power line. You just have groups of educated people in totally unproductive disciplines [environmentalists] who were trying as hard as they could to slow down or stop those trying to do something useful. So productivity has not got better; it has got significantly worse. I believe that if one tried to build a major power asset like a nuclear power station or even an ordinary old coal power station today, it would be almost impossible because the unproductive arm of society would be trying as hard as they could to stop it. It is little wonder that most of our new power production is coming from peaking plant [gas turbines designed to run for a short period only] as they are small and can be slipped in here and there without drawing obvious attention to themselves. But what will happen when the major base load [coal burning] stations wear out and have to replaced? And please don’t give me any bullshit about replacing coal with wind or solar power. Those who advoacte that simply do not know anything about the economics of power production and what amounts of power are required in even quite small industrial processes.
And to look at another area, at Adelaide University, several years ago for the first time, the number of ancillary staff exceeded the number of academic staff. Now what do all those ancillary staff do? Check up on the academics and make sure that they are not spending too much of the budget, keep records to show how many students are taught per teacher per day or collect other useless data to report back to the politicians who allocate money to universities? I mean what is a university there for? To teach students surely so why the empasis on ‘controlling’ the teachers.
So, the prevailing emphasis in engineering today is on controlling [stopping or slowing down] the productive people in our society.
Is it any wonder that the Asian nations are going ahead so quickly with their development; there are minimal or no controls and when you see how major Australian companies such as Leighton perform in those environments, you can see that Australian know-how and expertise is second to none if only you give it a chance.
And all that is required is to specify the amount of particulates in stack emissions and precipitators and scrubbers will remove the airborne nasties. Engineers can be trusted. They are servants of society, not its enemies!
Productivity gains? Humbug I say!


Comment by Dad | March 16th, 2009 9:00 pm | Permalink

Leave a Comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Live Comment Preview

Comment by Somebody

Powered by WordPress 2.8    Rendered in 29 queries and 0.567 seconds.    CleanBreeze Theme