Mr. John McKeon:

On the long-term unemployed, we do not refer anybody to JobPath who does not have a duration of 12 months on the live register. We just do not. It does not happen. There will be and are clients on other schemes. For example, they might be on another scheme and then go on to the live register. We have always counted them. For example, if they want access to the back to work enterprise allowance or back to education allowance, there are 12 month conditionalities around those. Those conditionalities already existed. The duration was counted for those schemes and is counted for JobPath too, which is of benefit to the people concerned.

I would maybe challenge the thinking on it. The implication is that they should not have to go because they are not at 12 months. For most people, overwhelmingly, it is a positive experience and the outcomes are positive. The Senator needs to think about that.

On the random nature, we do the random selection. We do it from our IT systems and it is entirely random. That is the situation. We have a mix of clients in six different cohorts, of one to two years unemployed, two to three, three to four and so on. We have to make sure that when it is picked, it is reflective of the overall live register, but within each cohort it is entirely random.

I might ask Mr. Christopher Kane to take the question on flexibility. The Senator referred to the question of the correct metric. The correct metric, from my perspective, is whether going on JobPath results in improved employment outcomes for people. Overwhelmingly, the evidence is yes. There is a 59% higher chance of being in employment where a person who has been unemployed for more than three years has been on JobPath compared with him or her not having been on JobPath. That is the most important metric. If the Senator wants to divide that by a cost, she can do that, but then that needs to be done like for like. What is the equivalent metric for the local employment service or for the Intreo service? Around the world, the metric used is the cost per client served because there is recognition of the value of employment advice, personal progression planning and so on.

The primary objective is to move into employment but it is of benefit to that in moving somebody further along a continuum. That might be a continuum into community employment or some other scheme. Around the world, it is the cost per client served. I am quite happy that people want to construct another. Let us do it across all our services. That is what I would say.

On exclusion in the procurement guidelines for health and social services, all I can do is repeat what I said earlier. In procurement of this scale, a contract of €50 million or more a year requires procurement. My understanding and our interpretation of the EU procurement rules is that it is absolutely essential. Leaving that aside for a moment, I have to go into the Committee of Public Accounts and account for spend. The Committee of Public Accounts would challenge me very strongly if I signed off on a contract of €50 million a year where there was no tendering involved. It would correctly challenge me very strongly. The best practice for procurement is to tender. The exceptions are exceptions from best practice, not the rule to be applied for certain services. It is not a case of saying that we like a group, so let us not try to discommode it.