If you buy a goldfish, you shouldn't be disappointed that it doesn't catch mice. It is therefore pointless to criticize Starmer's speech yesterday for being insufficiently critical of capitalism. There is however one curious but important omission from his speech - one which is odd for a technocratic social democrat.

I'm referring to the lack of any discussion of, or remedy for, our productivity problem.

Just before the pandemic, real wages were actually lower than they were in 2007. The average worker has gone 13 years without a pay rise. This is not because capitalists have taken a bigger share of the pie: the share of non-financial profits in GDP was much the same at the start of 2020 as it was in the mid-00s. Instead, it's because productivity has stagnated: in the 12 years to the start of 2020, output per worker-hour grew only 0.2% per year, compared to 2.2% per year in the previous 30. And in fact in the last 20 years productivity growth has been lower than at any time in the last century.

If workers are to get a pay rise, therefore, we need to raise productivity growth.

Doing so isn't just a matter of good economics. It's good politics too. Yes, Starmer's talk of insecurity and poverty spoke to the jobless and economically marginalized, which is good. But Labour also needs the votes of a wider section of the working class. Talk of raising pay speaks to these. Also, we know that a faster-growing economy fosters more liberal, inclusive sentiments and hence support for the left. Grace is bang right to say:

In the long run, a social democratic model would provide a far better terrain on which to advance class struggle than the current neoliberal one.

And yet Starmer's speech was astonishingly light on productivity.

But there is much that he could reasonably have said, even within the confines of centre-left Fabian-style thinking.

We could start from the perspective that competition raises productivity - not just by forcing businesses to up their game if they are to survive, but because lots of productivity gains come from new firms entering (pdf) the market.

Some of Labour's current policies should be presented in this context. Starmer's proposal for state-backed business loans would encourage start-ups, entry and competition. And we need them because there is a market failure here; banks lend eleven times as much in residential mortgages as they do to all SMEs outside the financial and real estate sector. And Rachel Reeves' call for more honest procurement should be seen as a way of using the state's buying power to encourage competition.

We could of course go further. We might call for tougher competition law or a relaxation of restrictive intellectual property laws. And, of course, trade frictions reduce competition and hence productivity, suggesting a need for at least a less stupid type of Brexit.

What's more, Starmer is right to call for more active government. It's not just investment in transport and better broadband that might raise productivity. Stian Westlake and Sam Bowman, two of the few intelligent centrists, propose direct state funding for innovation. And perhaps simply running the economy hot via looser fiscal policy could also raise productivity: there might be some truth (pdf) in Verdoorn's law.

Labour could, therefore, talk about productivity without straying much from centre-leftism.

Of course, there's also a more radical agenda.

One strand of this would be full-on attack on rentiers. It's no accident that the slowdown in productivity has coincided with the rise of rentierism as splendidly documented by Brett Christophers. Capitalists' incentives are skewed towards seeking the quiet life of extracting rents rather than the risky entrepreneurialism that raises productivity and living standards. And the extraction of rents suppresses dynamism by squeezing profits and real incomes: if you have to spend half your income on rent, that doesn't leave much demand for the goods and services that could be produced by entrepreneurs. Ricardo was right: rents are a drag on growth.

Labour has proposed one way to restrain rentierism - bringing the more egregiously rent-seeking government outsourcing back in house. But there are other responses, such as shifting taxes from incomes to land, and for nationalizing monopolies.

Another radical strand starts from the perspective that high inequality depresses growth. One reason for this is that the managerialism unleashed by neoliberalism might have reduced productivity by empowering a "long tail" (pdf) of bad bosses and by entrenching vested interests who want to retard creative destruction. As Joel Mokyr has written:

Every society, when left on its own, will be technologically creative for only short periods. Sooner or later the forces of conservatism, the "if-it-ain't-broke-don't-fix-it," the "if-God-had-wanted-us-to-fly-he-would-have-given-us-wings," and the "not-invented-here-so-it-can't-possibly-work" people take over and manage through a variety of legal and institutional channels to slow down and if possible stop technological creativity altogether.

This points to a need for democratic reform of the economy, and empowering workers rather than bosses: we know that democratic workplaces tend to be more productive. But as Phil says, it is on this point that Starmer breaks with Corbynism.

None of this is to suggest that it is easy to raise productivity. As Dietz Vollrath and John Landon-Lane and Peter Robertson have warned us, the opposite is the case.

This, however, is not a reason to give up. Most policies suggested here have little downside; even if they don't raise productivity they'll do little harm (except to those who need harming). There is, therefore, a big potential agenda here for Labour which could be both popular and economically sensible - if that is, it could get a fair hearing.