fence and ‘Operation Shield’. This may indicate that the reduction of entries in 2012-2013 was not necessarily the output of the technical barrier and the presence of police officers but may also be attributable to a combination of other factors. 69 Figure 4. Apprehensions at the Greek-Turkish Land and Sea Borders Note: data refer to apprehensions, not to people. Hence the same person if apprehended twice counts twice. Emphasis is added by the authors to signal peaks of apprehensions at different borders. Source: Greek police data, www.astynomia.gr For the period 2012-2013, the rapid decrease in arrivals might have also been the result of the wide circulation that Operation Shield and the Evros fence had received, which could have acted as a deterrent. By 2014, when the wait-and-see period started fading off and conflict in the Middle East intensified, irregular entries started increasing. In general, erecting a wall and investing in particular entry points, can be effective in blocking the entry through that specific gate. However if the aforementioned policies seek to reduce irregular migration throughout the country then their effectiveness is limited, since migrants appear to simply be re-directed to other entry points. An effective policy is also not necessarily a cost-effective one. The main weakness of the aforementioned policy is that it seeks to curb irregular arrivals, only after the migrants have reached the gates of the country. As a result, authorities seek to identify or speculate on possible entry points and ‘barricade’ them. This approach is not cost-effective for two reasons: on the one hand it invests in areas which may lose completely in significance in the course of 69 It should be stressed that a technical barrier alone may not always be the sole reason for a change of route, as shifts of this kind do in general take place without the reasons being always so straightforward. Page 32