Productivity and Employment (and Technology): In the Jaws of the Snake

by Andrew McAfee on March 22, 2012

Smart guy Jared Bernstein came to MIT yesterday and gave a talk to the seminar Erik Brynjolfsson and I are organizing on how technology is affecting the economy. Bernstein is a senior fellow at the Center on Budget Policies and Priorities, the former chief economist to Vice President Joe Biden, a sharp thinker, great speaker, and prolific blogger.

I met Jared last year when we were on a panel at the Aspen Ideas Festival on America’s manufacturing future, and have been following his work ever since. I suggest you do the same.

He talked to us yesterday about the big questions where Washington needs help from academia. One of them was on the causes and effects of America’s large and growing inequality. He said some nice things about the work Erik and I did in Race Against the Machine to highlight technology’s role in this phenomenon (he blurbed our book), and showed a very clever graph.

Courtesy of Jared Bernstein,

It’s a graph of American labor productivity growth and employment growth for the whole post-war period. It shows that these two grew together in lockstep until the turn of the century. After that point they started to diverge, with labor productivity continuing to grow nicely while employment growth leveled off, then turned negative during the Great Recession.

This divergence is particularly puzzling because economic growth was pretty healthy in the 21st century, up until the recession hit. So anemic economic growth isn’t responsible for the anemic employment growth.

So what is? As Bernstein and others have noted, outsourcing and offshoring are probably part of the explanation. I think that technology is another big part of it. The impact of the Web and commercial enterprise software really started to become clear after the turn of the century.

These digital advances let companies grow and do more without employing lots more people. They were great for the productivity statistics, but not for the employment ones. As we discuss in Race Against the Machine, I expect the digital advances to continue and even accelerate. If this is the case, the gap between the two lines in the above graph will continue to widen.

Bernstein calls the right side of this graph — the time of increasing divergence between productivity and employment growth – the ‘jaws of the snake.’ It’s easy to see why —  flat or declining employment is venomous for an ec0nomy, especially 0ne like ours with a growing population.

I predict that the jaws of the snake will become an urgent policy matter for America, even more urgent than it is now. Bernstein advocates a large program to upgrade America’s public infrastructure – its schools, roads, ports, railroads, and so on. I completely agree that this will help a the country’s competitiveness and create a lot of jobs. After all, robots aren’t yet any good at installing a new boiler in a school.

What are your favorite ideas for closing the jaws of the snake as we move deeper into the digital era? Leave a comment, please, and let us know.


Tim O'Reilly March 22, 2012 at 1:11 pm

I love the notion of portable internet-connected devices being used to add skill to currently low-skill jobs. I first noticed this in thinking about the way that the Apple Store swarms with clerks helping customers – and each of them is turbocharged with the phone as a cash register.  Todd Park (now US CTO) clicked when I told him this and noted that Walgreens is exploring adding the equivalent of a genius bar for prescriptions and medications, and that he could imagine the Apple Store model being used to empower home health care aides to help manage outcomes (e.g. reminding elderly patients to take their meds.)  I think there’s a lot here.

Another really interesting trend is our trend to user-generated entertainment, and bit by bit, new ways to monetize that. Etsy is invigorating a craft economy, YouTube, Kickstarter, and Amazon Kindle are enabling new kinds of self-financing of creative work.  It’s small stuff yet, but getting more interesting.

And of course there’s the maker movement, with new kinds of local manufacturing, and new kinds of industries growing up under our noses.  Who would have thought that DIY Drones might be the front end of a new industry?

davey jose March 22, 2012 at 4:09 pm

old jobs are never coming bk right? 
 look at the various revolutions, such as industrial etc.  forward not back.

tim is right, updating physical infrastructure is one way of creating jobs.  will this be a stop gap? how can we create sustainable workforce in the digital era?

how about trying to leverage the digital age and create jobs like Amazon Turk? but struggling to see how it would work on a large scale with varied level of skilled workforce.

Anonymous March 22, 2012 at 4:52 pm

No mystery at all, and pretty obvious, really: this shows the effect of firms’ substitution of capital investment for investment in people, ie hiring and expansion, during a period when the real cost of capital tumbled precipitously. Since the mid-1990s, the real cost of hiring the marginal ie additional worker did not fall — not because of real wage growth, which has been modest, but because of increased healthcare and, to a lesser extent, regulatory costs and our screwed-up tax code. 

Put another way, for US firms especially in their non-digital operations, the ROI on a $50m capex program that creates few net new jobs is much higher than the ROI on a business expansion program program that creates 1,000 jobs, each of which carries with it a 9% payroll tax and healthcare costs of similar magnitude.The less obvious insight – this is from Gary Becker – is that most of the productivity growth has been in the area of retailing and supply chain management, the poster child for which is Walmart. Becker estimated that fully one-third of the entire US productivity gain between 1995-2000 was due to Walmart and its sharp gains in operating efficiencies. Not exactly a beacon on the hill, so to speak, for broad-based prosperity. It’s a hard problem, but not a complex one: whatever course we pursue must have as its chief goal the lowering of the cost of hiring the marginal worker RELATIVE to the cost of capital.Here are ways to do that – they involve changes to the tax code, employer-based insurance, and the schools:a) reduce payroll taxesb) reduce taxes on income, both personal and corporatec) increase taxes on debt financing and PE firms’ “carried interest”d) sever the link between employment and health insurance. Give everyone a voucher, eliminate insurers’ ability to deny benefits to those with pre-existing conditions (Obamacare, btw, doesn’t do this – he carved it out until 2014…)e) bring back serious and well-staffed vocational education at the high school level, on the German model. Involve corporations and local employers in the design of the program. Give tax breaks to companies that hire vocational track grads. Enough of the silliness about everyone attending college – that’s a prescription for yet another housing-style bubble and crash. f) related to e), do whatever it takes to turn around the truly terrifying underperformance by our fastest-growing minority student population. This last point is the most neglected and most important one of all. Because of this alarming phenomenon that no one will discuss honestly or openly, California, for example, will soon face a shortage of college grads – yes, that’s right. Unbelievable? Believe it. Our schools demographic is rapidly becoming dominated by a group whose dropout rate is near 50%, whose STAR test failure rate is north of 70%, and whose share of the schools population will reach 70% by 2030.We don’t have underperforming _schools_; we have an underperforming _demographic. We’re looking at Mexican-style inequality in future decades if we don’t turn around this group’s performance, starting now. The Rocketship Charter School movement in east San Jose is a good start. More, please – thousands more, as fast as we can. And bring back vocational ed while you’re at it. 

davey jose March 22, 2012 at 5:12 pm

very interesting point you make on walmart and retail productivity/efficiency…  also interesting you suggest to create more jobs is to transform the “system” so makes sense to hire more….

i also believe the problem is linked to innovation.  we think we live in an era of great leaps and bound. yes we do. and we don’t. for example, the technological progress we have is vertical.  meaning through software on the computer. which came from the transistor invention.  so essentially without going to into much detail, the transistor created this vertical computer/internet/wired society we live in. and its created jobs. look at all the start ups. however they are all specialised roles.  what we need is to create new technology/innovation which creates different, im looking for the right word here, horizontal sectors, which in turn create “new” kinds of jobs….? eg. some form of invention for transport… this would mean we need to build new infrastructure and all the goods and services around that?

Anonymous March 22, 2012 at 6:12 pm

You’re talking about leaps forward comparable to the last century’s advances in areas like rural electrification, the automobile and the interstate highway system etc. The clear candidate here is energy, but it seems very unlikely that we will have such a breakthrough in our lifetimes. 

I think a much more fruitful prospect would be increasing older workers’ productivity and labor force participation.

To achieve this would require huge changes to the way work is done now – start with much more telecommuting – plus changes in organization to accommodate colleagues who have less stamina and less ambition but more wisdom and perhaps emotional intelligence. 

I have no idea how this would be done, but it seems essential given the actuarial challenges we face, the high unemployment rate and the advanced economies’ sharp declines in fertility to below-replacement rates.

davey jose March 23, 2012 at 4:53 am

yes, you’re defo right in what you’re saying, what we ideally need is to make leaps forward to get out of this spiral (in the west) to create new verticals. but also understand that it is not an easy thing…. perhaps the reason we are not making these breakthroughs is because the talent is being drained by technology/finance and not enough bell labs type setups? a structural issue? can it be addressed?

100% and beyond agree with what you say about increasing older workers’ participation, and technology via telecommunication (remote working), this would  be totally possible… i think technology wise, it’s cheap enough to implement these changes but its trying to get companies to see that this way of working is productive, and has utility is the shift forward we need. also so its changing perceptions that’s key driver to get these things rolling? can governments  “incentivise” this setup somehow?

Alex April 2, 2012 at 2:28 am

nice article Mr. Andrew thats my first visit on the blog and i am started liking it the blog is awesome

Trevor Miles April 2, 2012 at 9:13 am

Productivity improvements have always lead to drops in employment, and concepts such as BPR are proof positive of this.  My take is tthibaud is bang on.  Efficiency is due to a lot more than technology, but undoubtedly technology is an enabler.

My take is here –

slowblogger February 28, 2015 at 6:35 pm

Your link to Jared Bernstein’s blog is incorrect. I think it should be

Comments on this entry are closed.

Previous post:

Next post: