Evaluating Australian government policies using administrative microdata

Abstract

This thesis presents three Chapters using Australian administrative microdata to evaluate a range of government policy settings.

The first Chapter investigates the effectiveness of seven business investment tax breaks under a range of different macroeconomic conditions using Australian tax and survey microdata over a 15-year period, providing the first comprehensive evaluation of business investment tax incentives in Australia. We find that policies implemented during the 2009 Global Financial Crisis increased investment. Policies implemented during normal economic times mostly have no effect. Where present, responses are larger for unincorporated businesses, likely reflecting reduced efficacy of investment stimulus under Australia's dividend imputation system. We find no evidence that policies enacted to address COVID-19 had any effect on investment, perhaps because of the unique nature of the economic shock.

The second Chapter presents the first-ever method for tracking mergers and acquisitions in Australia using administrative microdata. Despite the importance of mergers and acquisitions to competition policy, Australia has lacked a mechanism to compel firms to report such activity, thus hindering policy analysis, evaluation and research. Moreover, the disparate methods by which mergers and acquisitions appear in the data is an obstacle for the competition regulator to quantify the impact this activity has on the degree of market concentration. This paper fills a gap in the literature by advancing three methods to identify mergers and acquisitions: first, clusters of employees are tracked between firms in a linked employer-employee database; second, we track companies moving between tax consolidated groups to infer where a merger or acquisition has occurred; lastly we look at merger notification forms submitted by firms to the securities regulator, the Australian Securities and Investment Commission. This yields a total of around 1,500 mergers a year wherein large, complex entities holding trademarks are most likely to be acquirers while smaller firms holding patents and trademarks are more likely to be targets in a merger.

The third Chapter looks at employment policies in the Australian public service, using employee microdata. We investigate whether being female, Indigenous, from a non-English speaking background (NESB) or having a disability affects career advancement in the Australian Public Service. Over the past twenty years, women have become more likely to get promoted at senior ranks than similar men and just as likely at junior ranks. NESB staff have much lower promotion prospects at higher ranks despite being less likely to separate from the public service. Lower promotion prospects for NESB staff are only partially explained by language proficiency or cultural assimilation. The unexplained remainder could reflect tokenism or a penalty for being non-white. Show more