Network Governance in the E-Lance Economy

摘要

To date, the primary focus of global information technology management has been to examine how technologies facilitate the coordination and completion of work within multi-national organizations and improve efficiencies across global supply chains. However, the true global potential of advanced technologies is just starting to emerge. We have discovered that global connectivity has made the world flat. The future of work, especially knowledge work, might not only occur in large, command-control organizations, but through decentralized, fluid networks where individuals largely self-organize and manage themselves, in an "e-lance " economy. The goal of this paper is to develop and test a theory of network governance specific to this new emerging e-lance economy by integrating transaction cost economics with the concepts of social controls. Transaction costs are assessed by measuring demand uncertainty (# of projects and % of projects that were awarded), human asset specificity (# of bids), task complexity (average $ amount of bids and average time projects are open) and frequency (% of repeat customers and % of repeat suppliers). Social controls are assessed by measuring reputation (average supplier reputation), collective sanctions (% of suppliers getting flamed) and restricted access (projects that restrict access to who can bid). We use three measures of e-lance success: the number of project posted to the network, the average number of bids per project, and the total dollar amount exchanges on the network. We empirically test our hypotheses using data from 14,644 projects across a two-month time frame on the website - separating data collection of the independent variables and dependent variables across time. In terms of social controls, restricting access by sealing bids and notdisclosing budget amounts leads to lower levels oj bidding, but not disclosing budgets is associated with more projects being posted. In terms of transaction costs, higher average project values lead to more projects and more money being exchanged, but more bids leads to less monetary exchange. We find that the best predictor of e-lance success across all measures is the number of projects posted in the prior time period- thus active networks where there are lots of projects are the most successful over time.