The Stationarity Properties of Software Piracy: Example of the OECD Countries1

Year 2016, Volume: 5 Issue: 2, 80 - 94, 01.06.2016

Selim Yıldırım S Fatih Kostakoğlu Ethem Esen

Abstract

Software piracy refers to the unauthorized copying, distribution and selling of software in copyright. The unlicensed software use and distribution has been important issue for the industry. It leads to tremendous monetary losses in the software sector. According to BSA the commercial value of unlicensed software in 2013 is $13.5 billion in European Union and $17.2 billion in BRIC countries. Moreover, many countries that have high software piracy rate, have shown a predilection towards pirating software in the past. Georgia, Moldova, Zambia and Zimbabwe which has the four highest software piracy rates and all above 90% in BSA’s (2014) study have continuously exhibit rates above 90% since 2005. Furthermore, the commercial value persistently climbs since BSA has first began publishing studies on software piracy. This brings forth the question whether software piracy has a characteristic inertia. Due to the size of loss it creates software piracy has many consequences. This study aims to investigate this question for OECD countries. Two different variables related to software, namely the monetary value of software firms’ losses and the index of software piracy for thirty OECD countries are investigated using Pesaran’s (2007) CIPS (cross-sectionally augmented IPS) and Hadri and Kurozumi’s (2012) panel unit root tests. CIPS test indicates that for developing OECD countries software piracy index is stationary and for developing OECD countries monetary value of software firms’ losses are trend stationary and for developing OECD countries both variables are stationary. HK test on the other hand indicates that only piracy index for developing countries are stationary. The conflict arises from the opposite null hypothesis of both panel tests; while CIPS test homogenous unit root in the panel HK tests homogenous stationarity. Since none of the panel variables are composed of strongly stationary processes or conversely unit root processes tests do not reach a consensus for the variable in every case. Although results of the test seem to be conflicting two conclusions emerge from this study the index variable, which is the perceived level of software piracy, has remained similar within time for developing counties whereas it has changed greatly for developed OECD countries. Finally, monetary value of software firms’ losses has very different dynamic properties even among the countries within the same group so that the two tests cannot agree for a single group or model