https://macropaperwarehouse.com/authors/christian-vom-lehn/Christian Vom LehnHugo -- gohugo.ioen-ushttps://macropaperwarehouse.com/papers/structural-change-in-production-networks-and-economic-growth/https://macropaperwarehouse.com/papers/structural-change-in-production-networks-and-economic-growth/<p>Gaggl, Gorry, and vom Lehn document that services sectors produce a rising share of output in both of the US production networks they study — the input-output network for intermediate inputs and the investment network for new capital — over 1947-2020, and they construct price series for goods and services separately by final use. Their central measurement finding is that while the relative price of services used for consumption and intermediates is rising, the relative price of services used as investment is falling, a reversal of previous studies that they attribute to aggregation bias: since only around 5% of services output is used for investment, investment prices are averaged out of the sector-level gross output prices those studies rely on. Fitting a multi-sector growth model to these price and expenditure patterns, their calibration implies that goods and services are complements in producing consumption and intermediates but substitutes in producing investment, so that structural change endogenously reallocates resources toward the slowest-growing intermediates producers and the fastest-growing investment producers. In their growth accounting over 1947-2019, investment-specific technical change accounts for 52% of aggregate GDP-per-worker growth, rising to 94% over 2000-2019, while intermediates-specific technical change has stagnated and contributes negatively after 2000; imposing Cobb-Douglas investment aggregation, which shuts down structural change there, lowers cumulative growth by 5% over the full sample and by 20% since 2000. Projecting to 2070 under the assumption that all sectoral TFP series grow at their average 2010-2020 rates, growth in GDP per worker rises from 0.66% to 0.87%, which the authors read as indicating that reallocation within the investment network appears sufficient to offset the growth drag from Baumol&rsquo;s cost disease in consumption and intermediates. The conclusions are conditional on their calibrated elasticities and on the constant-TFP-growth assumption behind the projection, and the authors describe the net impact on aggregate growth as, in principle, a quantitative question.</p>