Counting Levels or Measuring Distance? Organizational Size and Hierarchy Across and Within Organizations

That larger organizations have more hierarchical levels is among the most replicated findings in organization theory. Yet Cullen, Anderson, and Baker (1986) found that the relationship did not hold within organizations as they changed size, and concluded that size–structure theory explains differences of scale rather than processes of change. We argue that this gap is partly a consequence of measurement. Because span limits force a new level only after a several-fold increase in size, and because the level count is set by the longest chain of command, the count registers growth only at widely spaced thresholds and moves for many reasons unrelated to size. Average geodesic distance, the mean number of reporting steps separating any two members, has neither limitation. Using published organograms of UK central government bodies (a cross-section of 70 organizations, 2011–2013, and a panel of 18 organizations observed 520 times, 2011–2025), we find that levels increase with size at a decreasing rate both across and within organizations. Within organizations, however, the count changes in only a third of periods and does not respond to period-to-period changes in size. Average distance is far more strongly associated with size in both designs and responds to growth and contraction. As predicted, distances also become more uniform as organizations grow, at similar rates across and within organizations, suggesting that differences of scale are the accumulated result of change. Distance-based measures thus give theories of organizational size and structure a more sensitive account of how hierarchies change.