Evaluating Digital Pakistan Policy : evidence on firm innovation

Abstract

This thesis offers the first quasi-experimental evaluation of Pakistan's Digital Pakistan Policy 2018 on firm-level innovation, using a matched panel of 200 firms from three waves of the World Bank Enterprise Survey (WBES) (2007, 2013, and 2022). The empirical strategy combines a Difference-in-Differences (DiD) design with an event-study and a placebo extension, comparing high- and medium-technology manufacturing firms (treatment) with the low-technology manufacturing, retail, and services firms (control). Two-way fixed effects absorb time-invariant firm heterogeneity and province-specific shocks; standard errors are clustered at the firm level. The study addresses three questions: (1) Did the Digital Pakistan Policy generate differential changes in product/service and process innovation between treated and control firms? (2) Which firm-level capabilities explain innovation in Pakistan's enterprise sector? (3) Did the policy alter export participation among tradable manufacturing firms? DiD estimates reveal an asymmetric pattern. The differential effect on product/service innovation is small and statistically insignificant. The differential effect on process innovation is positive under the firm- and province-year fixed-effects specification, reaching 10% significance. The differential effect on export participation is statistically insignificant once firm and province-year fixed effects are absorbed. Read together, the evidence is consistent with a policy whose firm-level reach was narrow and concentrated in operational, efficiency-oriented improvements rather than in new product development or export expansion. The mechanism analysis is the strongest empirical component of the study. Formal employee training and access to external finance emerge as the strongest correlates of innovation. Training is associated with both product/service and process innovation, while access to external finance is associated with higher process innovation only. These are within-firm associations rather than identified causal effects. The thesis makes three contributions. It provides the first quasi-experimental evidence on the Digital Pakistan Policy's firm-level effects and demonstrates that careful DiD identification is feasible on the WBES Pakistan matched panel despite a small treated cell and the data constraints of a developing-country survey. It documents that ambitious digital policy in a weak-institution setting did not generate the broad firm-level innovation response its design implied, while identifying training and finance as the firm-level capabilities that explain innovation. Finally, it cautions against treating digital policy announcements as substitutes for firm-level capability building and supports pairing future iterations of the Digital Pakistan Policy with targeted instruments in skills development and SME finance.