Startup R&D has replaced corporate R&D

I have a problem with Yves Smith's and Rob Parenteau's op-ed in the Times:

Over the past decade and a half, corporations have been saving more and investing less in their own businesses... Since 2002, American corporations on average ran a net financial surplus of 1.7 percent of the gross domestic product—a drastic change from the previous 40 years, when they had maintained an average deficit of 1.2 percent of G.D.P. ... To show short-term profits, they avoid investing in future growth...

Smith and Parenteau evidently believe that since 1995 corporations have spent less on R&D, stifling innovation.   But, on the face of it, there has been plenty of innovation since 1995.  What gives?

Steven Kaplan and Josh Lerner say in their recent It Ain't Broke: The Past, Present and Future of Venture Capital:

Beginning in the early 1990s... American corporations began fundamentally rethinking [internal R&D facilities]... relying much more heavily on what has been termed "open innovation," i.e., alliances and acquisitions of smaller firms... venture-backed firms are approximately three times as efficient in generating innovations as corporate research.

Corporations are spending less on internal R&D.  But this is because internal R&D is far less efficient than buying venture-backed companies.  Smith and Parenteau forget that Savings = Investments.  When a corporation saves money, it invests it somewhere else.  Sometimes in something innovative.

Smith and Parenteau recommend

[Creating] incentives for corporations to reinvest their profits in business operations... impose an aggressive tax on retained earnings that are not reinvested within two years...  At the same time, the federal government must continue to encourage investment in the economy—ideally by creating incentives for investments in national priorities, like new energy technologies. 

This is exactly wrong.  Corporations should save money or distribute it to their shareholders to save.  That money should be turned into investment in innovative new firms that the corporations can later acquire.  And as to the government steering investments: if corporate-backed R&D is three times worse than venture-backed R&D, I can't even imagine how bad government-directed R&D is.

More investment does decrease consumption in the short-term*, and this is a problem during a downturn.  But because investment increases income growth, it increases the opportunity to consume longer-term.  It might well be that this is the wrong moment in time to encourage savings over consumption, but to create a policy that would discourage more productive investment for years to come makes absolutely no sense.

-----
* GDP = private consumption + gross investment + government spending + (exports − imports)