David Argente, Munseob Lee, and Sara Moreira write,
In this paper, we exploit detailed product- and firm-level data to study the sources of innovation and the patterns of productivity growth in the consumer goods sector over the period 2006Q3–2014Q2. Using a dataset that contains information on the products of each firm and the characteristics of each product, we document new facts on product reallocation. First, we find that an important component of reallocation of products happens within the boundaries of the firm. Second, the largest changes in product quality come from new firms launching new varieties and from small firms expanding to other product lines. Third, we document that product reallocation within firms is procyclical. Fourth, we find that within-firm product reallocation is larger in high productive firms and firms that invest more in R&D. Finally, we quantify how important how product reallocation affects firm-level productivity growth and and innovation as reflected by changes in their total factor productivity.
NOTE: I am quoting a version that is labeled VERY PRELIMINARY AND INCOMPLETE. The published version is gated. Pointer to the published version from Tyler Cowen.
Apparently, during the recession, firms reduced the pace at which they added new products and retired existing products. I interpret this as a slowdown in investment.
As a first approximation, this is not supportive of my view of a recession as a breakdown of existing patterns of specialization and trade. One would expect to see an increase in the retirement of existing products if my view were correct.
One way to rescue my view would be to say that firms respond to a deterioration in the sustainability of existing patterns of specialization and trade by reducing their investment in creating new patterns. This seems like a counterproductive response, except that it does conserve cash in the short run.
Essentially, there’s a substitution and an income effect in investigating new patterns. You might wish to substitute more exploration of new pattern for existing patterns, but you can’t afford to. In fact, this might be why you want NGDPLT even in a PSST world. By decreasing uncertainty about the future level of NGDP, it minimizes the income effect.
This isn’t what the abstract says, it specifically states that “Second, the largest changes in product quality come from new firms launching new varieties and from small firms expanding to other product lines.” Small and new companies don’t thrive in predictable markets, they thrive in markets where outcomes are less certain. Large companies with well developed lines are ones that want long range, stable forecasting.
I was talking about income and substitution effects for the large firms only.
According to the paper large firms only see intra firm changes, advocating for NGDPLT in this case would be for advocating only for (near) zero growth targeting. You only get the stability by stripping out the portions of the economy associated with growth. This is not PSST, this is arguing for a static equilibrium.
“As a first approximation, this is not supportive of my view of a recession as a breakdown of existing patterns of specialization and trade. One would expect to see an increase in the retirement of existing products if my view were correct.”
I don’t agree, if companies are consistently releasing new products every few months then that is part of the pattern that satisfies consumer demand.
In a recession we end up with piles of old products. Transactions slow, inventory passively grows relative to flow.
As a first approximation, this is not supportive of my view of a recession as a breakdown of existing patterns of specialization and trade. One would expect to see an increase in the retirement of existing products if my view were correct.
I disagree as well.
On average new product introductions may improve revenues, but with the larger average comes a larger variance. The strategic reaction in a recession for a profitable company is to lower the variance to conserve the longest runway.
A venture-backed company, on the other hand, has only high variance plays to make, so they may get even more benefit from their new products in such an environment, where the established players are conserving cash.