A Model of Service Sector Growth through Creative Destruction and Outsourcing

Over the past few decades, the service economy has expanded dramatically in the industrialised world and, at an accelerating pace, in emerging markets and developing countries. Most empirical studies attribute this secular shift to rising per‑capita incomes and, by extension, to household demand for consumer and financial services. This paper argues that such explanations are incomplete. A sizeable fraction of service‑sector growth is instead fuelled by demand originating in manufacturing and agriculture—a process labelled servicification. 

The ultimate driver is productivity, which links goods‑ and service‑producing industries through two intertwined channels. First, the uptake of specialist services—logistics, design, ICT, finance—raises productivity within manufacturing and agriculture by allowing firms to focus on core competencies. Second, independent productivity improvements within service firms make those inputs cheaper and more effective, inducing still greater downstream demand. These mechanisms operate concurrently, creating a feedback loop in which gains in one sector reinforce gains in the other. The resulting joint productivity surge, mediated by competitive selection pressures in both industries, strengthens incentives for firms to outsource an ever‑wider range of tasks to highly specialised service providers. Deeper structural forces amplify this process. Progressive improvement of market‑supporting institutions lowers transaction costs and contract risks, while creative destruction driven by entrepreneurial entry and exit reallocates resources toward the most efficient producers, further boosting aggregate productivity over the long run.

We evaluate this theoretical framework empirically in three steps. First, we assemble stylised facts that map the co‑evolution of service‑sector size and multifactor productivity across countries and time. Second, using the OECD TiVA input‑output database, we trace the rising value‑added share of services embodied in manufactured exports—our benchmark measure of servicification. Third, we estimate reduced‑form regressions that relate servicification to productivity growth, institutional quality, and indicators of entrepreneurial dynamism. The estimates reveal that joint productivity dynamics and their underlying determinants account for observed increase in servicification. While this outcome may appear intuitive, quantifying the relative strength of each channel allows us to calibrate a structural model capable of replicating cross‑country and intertemporal variation in servicification intensity with considerable precision.