A typology of unexpected events in complex projects

Abstract

Abstract Purpose – The purpose of this paper is to understand the origins and nature of unexpected events that affect complex projects, by relying on a view of projects as social systems. The authors argue that the project relation to its environment is mediated by a model of this environment that is embedded in the communications between project participants. The adequacy of this model to the causal texture of the environment inspired a first, epistemological, dimension for characterizing events: event predictability. The nature of the boundaries between system and environment inspired the second dimension: locus of generation. Design/methodology/approach – This study followed a multiple‐case study approach. The authors collected data in 17 complex projects, in three types of industries: construction, IT/IS, and pharmaceutical. Findings – In total, nine categories of unexpected events were identified from the intersection of two dimensions: event predictability and locus of generation. Research limitations/implications – The empirically validated two‐dimensional framework sheds new light on the way organizations react to unexpected events and on the reasons for the eventual project performance. Practical implications – The findings show that project managers tend to underestimate certain risks. This research will help managers better predict those types of risks. However, some risks are simply unpredictable, therefore the authors argue for the necessity to prepare projects for the unforeseen. Originality/value – Analyzing the previous literature in unexpected events, the authors identified two main, but opposing, theoretical perspectives: one rooted in decision theory and the other that sees projects as social systems. The value of this paper comes from the original mode in which the authors propose to reconcile these perspectives, by viewing projects as networks of communicative couplings between actors.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 41 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.