The purpose of this study is to analyze perceptions of the policy for improving the treatment of contingent workers in Seoul metropolitan city. A policy for contingent workers was introduced in the capital in 2012, and the core provisions relate to their welfare, along with mechanisms to upgrade their status from irregular (indirect-hired) to regular (direct-hired) workers. However, several issues have arisen in the process: implications for the retirement system, policy cost, and the spread of policy. This study uses Q-methodology in order to analyze stakeholder perceptions. The Q-sample was based on a review of related academic literature and newspapers, as well as interviews, while the P-sample consisted of public servants, public firm employees, contingent workers, and experts. Analysis of the results reveals five categories of perceptions towards the policy: (1) policy-supporting, (2) policy-distrust, (3) society-distrust, (4) market-supporting, (5) worker-supporting. Types 2 and 4 are primarily concerned with efficiency, while on the other hand, types 3 and 5 seek fairness. Types 4 and 5 can be understood as representing more extreme positions. The highest degree of correlation was found between the (1) policy-supporting and (3) society-distrust types (0.3714), and the explanatory power of the analysis was 67 percent. The implications of the analysis are that the improvement policy for contingent workers in Seoul needs to achieve a balance between efficiency and fairness, the policy design needs to include mechanisms for the representation of various stakeholder perceptions, and that the introduction of a participation system will lead to a more successful policy result.