victims of human rights violations; others mentioned government agencies that might benefit from training or advice. Pinpointing specific stakeholders to ­develop participatory indicators may therefore be a challenge. Furthermore, community forums will likely reflect power structures within the community. Disadvantaged or marginalised groups might find they are ­excluded, or may not consider the workshops to be environments in which they can safely express themselves. Another obvious drawback is that developing participatory indicators can be lengthy, cumbersome and expensive. But this is true of all indicators: the quest for perfect ones may be impossible. Their purpose is to help institutions to conduct their work more effectively – to identify trends and problems, and chart progress. They should not lead to prolonged and costly processes. The ­compromise means that indicators will sometimes be rough and ready. In summary, if NHRIs want to devise measures of their own ­effectiveness, they should consult those with whom they work. The starting point should be the existing contacts that NHRIs have with those who use their services. National institutions with complaints procedures already have regular contact with the public, as do those engaged in human rights education. It would not be difficult to devise methods of consultation – individual or ­collective – to determine what are their partners’ expectations of the institution and of its programmes. Later they should be consulted as to whether those indicators have been fulfilled. NHRIs should consult and work regularly with partner organisations – especially human rights NGOs and community bodies – and government branches. They too should be engaged in a continuing dialogue with them about what the NHRI should be doing and how its performance and impact might be measured. One outcome of this dialogue would be the development of more sensitive ­indicators of effectiveness. At the same time, it would extend the participation of various groups in the NHRI’s work. The limitations of participatory indicators should be seen as challenges to ­overcome, rather than arguments against using them. They should not cause the authority and identity of national institutions to be subordinated to the aims and interests – potentially contradictory – of stakeholders. NHRIs will only be able to evaluate their progress if they take into account the expectations of beneficiaries and partners, as well as their institutional targets. 3. Reliable and valid indicators To be useful, indicators must be both reliable and valid. To be reliable, data must first be accurate. In addition, results must be ­replicable, regardless of who gathers the data. For example, the results of collecting ­information through monitoring should not depend on who the monitor is. To be valid, indicators must genuinely show what they are supposed to show. A relevant example for NHRIs is the use of data from complaints ­procedures Assessing the Effectiveness of National Human Rights Institutions 27