Going (More) public: institutional isomorphism and ownership reform among Chinese firms

Heather A. Haveman*, Yongxiang Wang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

17 Citations (Scopus)


Publicly traded Chinese firms recently reformed their ownership structures by converting non-tradable shares, which constituted two-thirds of shares outstanding and were held largely by the state, into shares that could trade on domestic exchanges. To facilitate this reform, tradable shareholders were compensated with stock grants from non-tradable shareholders. Our analysis focuses on the level of compensation, the compensation ratio, the ratio of new tradable shares granted to tradable shares outstanding before the reform. Contrary to the predictions of asset-pricing models, most firms set the compensation ratio around 0.3. We explain this surprising convergence using institutional theory. In doing so, we analyze the power and interests of all relevant actors - not just owners, but also state regulators, executives, and other agents - and draw on insights from resource-dependence and agency theories. We find strong evidence of coercive and mimetic isomorphism, but no evidence of normative isomorphism. Because our dependent variable is continuous (a ratio), we are able to show that the mimetic effects we observe cannot be attributed to coercion or norms. Thus, we not only explain an empirical puzzle, we also advance institutional analysis of isomorphism by clearly distinguishing three isomorphic forces that have been conflated in much previous research.

Original languageEnglish
Pages (from-to)17-51
Number of pages35
JournalManagement and Organization Review
Issue number1
Publication statusPublished - Mar 2013
Externally publishedYes


  • agency
  • corporate governance
  • isomorphism
  • reform
  • resource dependence


Dive into the research topics of 'Going (More) public: institutional isomorphism and ownership reform among Chinese firms'. Together they form a unique fingerprint.

Cite this