Added altogether, including bills, repair and maintenance costs, the current annual estimated cost of running Amygdaleza reaches €10,500,000 per year, i.e. roughly €14 per person per day. This is a 1/3 of what Italy pays to detain a migrant (€45 per day). 41 In practice, the budget for Amygdaleza sum has proven hardly enough to cover round-the-clock the most basic needs of the 2,000 persons held on a daily basis under the exclusive control of the Greek State. Assessing the cost-effectiveness In accordance with Greece's National Action plan on Asylum and Migration Management, the main purpose of this recent detention policy is two-fold: first, with the exception of persons entitled to international protection, to increase the return rate of irregular migrants currently residing in Greece and second, to deter irregular arrivals in the long run. 42 For the purposes of the present case-study, two sets of data were used to assess the costeffectiveness of this policy: first, the statistics that cover the whole period of Amygdaleza's operation as a pre-removal centre since its operation 28 April 2012; second, the statistics that reflect the situation in the camp on the day of the visit on 8 April 2014. These will be then evaluated against the above-described costs. All in all, 9,677 detainees have been registered since 28 April 2012. On the day of the visit on 8 April 2014 the number of detainees was 1,715.43 The overwhelming majority comes from Bangladesh, Pakistan and Afghanistan. On the day of the visit there were also a couple of Somalis, Syrians and Palestinians in detention. All three top nationalities can be deported. Since its opening, a total of 2,700 out of the 9,677 detainees was forcibly or voluntarily returned from within the facility (27,9%) and 2,220 detainees applied for asylum in Amygdaleza (22,9%). Overall 1,195 migrants had consented to be returned to their home countries (12,3%) and 767 detainees had explicitly requested to be returned and were actually returned(7,9%). On the day of the visit on 8 April 2014, 586 had applied for asylum (34,1%) and 145 had agreed to be returned to their home countries (8,4%). In terms of effectiveness, the figures illustrate that in a period of two years the use of detention resulted in an average 27,9% return rate. Detention was purposeless for a 28,5% of detainees, since an application for international protection had been submitted, therefore resulting in a short-term suspension of the return procedures. Finally, detention did not achieve its purpose for 48,6%, because they were neither returned nor had their status been regularised. As far as consent is concerned, an average of 10% had — at an unknown time — agreed to be returned and 7,9% had volunteered to be returned and were actually returned. The figures leave open to what extent the 10% “agree” rate overlaps with the 7,9% “request” rate. The overall consent rate ranges thus between 10%- 17,9%, i.e. an average of 14%. The figures also leave open the question of whether the consent was a result of detention — as this new policy aspires to achieve — or the person had already consented and was placed in prolonged detention nonetheless. Looking at the wider picture, 40 41 42 43 See http://static.diavgeia.gov.gr/doc/Β430Ι-ΟΥΕ . See Global Detention Project “Italy Detention Profile”, available at http://www.globaldetentionproject.org/countries/europe/italy/introduction.html#c2423 Supra fn.11. Data provided by Police during fieldwork visit on 8. April 2014 8