Reihan Salam, in a characteristically excellent post here at NRO, points to a paper by Michael Mandel, who is one of the most interesting blogospheric commentators on innovation from a “New Democrat” perspective. Mandel makes the basic point that progressives should not be so gung ho about antitrust enforcement, because big organizations like AT&T Bell Labs and Apple are the anchors of business ecosystems that drive innovation.

My experiences lead me to agree with the conclusion – my first job out of grad school was at Bell Labs, and I’ve since started and built a global enterprise software company — but also to be a little more jaded about exactly how big firms contribute to innovation in the kinds of industries he discusses. 

Just as Mandel indicates, there is some straight-up development of new technologies in big labs that is then deployed by the parent company (I was involved in some). Further, consciously planned ecosystems of the type he cites — for example, developers building apps for the iPhone and iPad — can help to identify and then scale successful new technologies efficiently. Both of these things matter a lot. But here’s what I have seen big companies actually do to drive the innovation that I think is most important for overall job creation and long-term growth:

1. Because innovation can only be partially planned, even the best research labs that create enormous value for the parent company also inevitably discover things that cannot be practically exploited by the parent firm. In the more extreme cases, they produce innovations that would threaten the parent company’s business model. Xerox’s comparatively tiny PARC lab invented the laser printer, which Xerox turned into a multibillion-dollar business. It also developed the graphical user interface, Ethernet computer networking, and most of the other elements of the modern personal computer that Steve Jobs famously exploited to make Apple, not Xerox, a leading personal-computer company.
2. Big companies provide a cash exit for successful start-ups, either before or after IPO. In this way they act as informed allocators of capital that intermediate between general investors and the complex technology landscape. Think of most software start-ups and IBM, Oracle, SAP, Microsoft, and HP.
3. Big companies also use partnerships and other vehicles to act as marketing arms and integrators for successful technologies developed by start-ups. Think of biotech and big pharma.

What’s critical about these roles for big companies is that they require that you have lots of entrepreneurial firms to compete with the incumbents. And, in fact, if my characterization is correct, you would expect most of the job creation to happen in the successful new entrants as they grow, which is just what we see. According to the National Venture Capital Association, venture-capital-funded firms employ a majority of all workers across many of the most productive and growing sectors of the economy, including the software, biotech, semiconductor, electronics, telecommunications, and computer industries. 

I’m glad to see somebody on the left arguing for a modernized view of antitrust, but I think that what is essential if we are to do this is to reduce simultaneously the political power of large companies to stifle competition, as manifest in manipulation of patents, financial regulation, safety rules, and the endless list of regulations, subsidies, and tax breaks that govern the modern economy. This is similar to what Reihan called in his post “completing the neoliberal revolution.”

The market process is imperfect and takes time, but in my view is preferable to one in which we allow large companies (which will always have an advantage in lobbying and compliance) to use the political process to protect their position, which we then counter-balance with antitrust regulation. No real system of political economy is ever pure, so we will always have some amount of political jockeying and counter-jockeying; but in general, the more we get government out of the way of innovation, the better off we will be.

I think that “de-politicizing” the economy could be an important and powerful component of a Republican presidential campaign in 2012.