Comparative Institutional Ethnography — Multi-Site IE

Multi-site ethnography (following Marcus) traces connections across sites by following people, objects, or discourses wherever they lead. Comparative IE is more structured: it holds the institutional function constant, selects sites to vary the context deliberately, and uses text-mapping as a specific analytic tool to compare how ruling relations operate at each site. The comparison in CIE is always organized by the question of how institutional coordination works, not simply by cultural connectivity.

Two sites is the minimum for comparison; three or four sites are common in published CIE studies. Sample size per site is not fixed by rule — the criterion is whether the key textual sequences and ruling relations at each site have been sufficiently mapped. In practice, 8–20 interviews per site plus systematic document collection is typical, but the analyst continues until the institutional logic of each site is adequately understood.

Yes, and this is one of its most common applied uses. By showing how a policy text enters the work at each site and what it requires of workers and clients, CIE can reveal whether a policy achieves its stated aims and where implementation diverges from intent. The analysis does not render a verdict in the positivist sense but provides a detailed account of how the policy operates in practice across settings.

No. CIE studies are typically conducted sequentially or with staggered overlapping fieldwork periods. The analytic work of cross-site comparison happens after data collection at each site is complete or substantially advanced. Co-investigators or research assistants can conduct fieldwork at one site while the lead researcher works at another, provided all team members share the same IE analytic framework.

Deeply so. IE was developed within feminist sociology precisely to make visible the invisible work of women and marginalized groups within institutions. Comparative IE extends this commitment by showing how institutional ruling relations reproduce or differently configure gender, race, and class inequities across settings. The method's standpoint orientation and its centering of workers' and clients' actual experience align closely with feminist epistemology.