Introduction
In this paper, I would like to address the issue of recording and monitoring of racist and discriminatory incidents in a local setting and a multi agency context. I will do so, in presenting some of the concrete experiences RADAR, as a regional Dutch NGO, has built up in developing a multi agency monitor in dialogue with Rotterdam reality

1. About RADAR
RADAR, the Rotterdam Anti Discrimination Action Council, was one of the first local organisations in this field in the Netherlands. The activities of RADAR concentrate on four main areas.

-RADAR has a complaints office, which assists victims of discrimination through mediation and/or legal procedures. We deal with about 600 cases a year

-Second, RADAR registers, monitors, analyses and reports about discrimination
on a structural level.

-Third, it provides information and training to individuals and organisations. We deal with about 250 requests of this kind pro annum.

-Fourth, RADAR does research and advises on policy matters.

For all of these activities RADAR interacts with a great many individuals and organisations, notably the police, local government, housing associations, education institutes, café and discotheque owners and bouncers, personnel departments etceteras.

My presentation, thus, is one from the perspective of a Dutch, regionally operating NGO, serving a population of appr. 700.000 people with about 40 % inhabitants of foreign origin. RADAR works with a team and board with people both from majority and minority populations. We are not a traditional ethnic (community) organisation, but an organisation with a general task, combating the forms of discrimination prohibited in the first Article of the Dutch Constitution.
I shall not elaborate on the general work of RADAR, but focus of the work in the field of multi agency monitoring and recording of discriminatory incidents Before that, however, some notes on working on the local level

2. Basic work at the local level
For anti-discrimination work, the local level is important in a number of ways:

-Incidents, confrontations, but also most interventions and solutions to problems are practised in a local setting.

-At the local level, equal treatment has to be guaranteed in practice and the local level is the first level for intervention in cases of unequal treatment.

-The quality of local implementation is crucial for national legislation and other measures to combat discriminatory practices.

-The local level is also the level to experiment with new approaches and to find out what works and what does not.

-The systematic documentation of frequent forms of discrimination at the local level can be used to document or prove the existence of structural problems of inequality and to advocate structural (race)equality measures.

-The technicalities of fact-finding illustrate the interconnection and the interdependency of local, national and international policy. International reports on racist incidents are based on national reports, which in turn are based on local observations.

3. Detection of discrimination at the individual and structural level
For a NGO like RADAR, victim support is a key issue.
Not all forms of discrimination, however, can be detected through individual complaints. An individual who is on a waiting list for a house or who applies for a job, often is no position to prove that the denial of the job or the delay in getting the house is caused by the fact that he or she has been discriminated against.
In cases like this, structural action is required.

But on what issues should this kind of action focus? On education, labour market, distribution of houses or preventing conflicts between youngsters? In making these choices, monitoringsystems can be helpful, because they link structural choices to empirical facts. This goes especially in situations where a monitor is developed in a multi agency setting.

The idea of a multi agency monitor is based on the assumption that many agencies are in possession of relevant indications about the occurence and characteristics of racial incidents. Bringing these together in a multi agency monitor should guarantee obtaining a reliable and more realistic picture of cases of discrimination. A multi agency approach should also guarantee mainstream institutional participation in providing solutions to institution related cases of discrimination (e.g. police-service,
housing). RADAR is not only dealing with racist incidents. Because the organisation has a broad scope and combats several types of discrimination, as well as more subtle forms of exclusion, RADAR’s monitor includes the wider area of structural (sometimes indirect and unintentional) forms of discrimination.

4. the RADAR model for monitoring discriminatory incidents; factors and actors
In the model below, you see the way the RADAR Monitor is structured.

In the upper part of the model, the different sources of individual signals RADAR uses are specified. These are complaints, requests for information and reports on right wing activities.
The importance of recording complaints and reports on right wing activities probably speaks for itself. Requests for information, from our experience, are relevant too, because, in a number of cases, they point at problems in a pro-active way, while on the other hand they indicate resistance against racist or discriminatory practices.
At the bottom of the model, on the input side, you see-data being integrated in the monitor with a series of information flows of non-specialist, predominantly mainstream agencies.

In present practice these are the police, public prosecution, the public service in´
charge with graffiti removal and media reports.
I must admit that it’s a permanent struggle to keep the flow of records going, to instruct and assist new liaison-persons, to assist organisations in the optimisation of their internal record keeping procedures and to make new arrangements with other agencies. The number and type of partners involved, of course, will be expanded either on a structural or an ad-hoc basis, as the monitor develops.

Some organisations are involved in more (or even all stages) of the monitoring process. The police are prominently present at the input side, but could also make use of the output, e.g. may ground the decision to patrol certain areas more thoroughly or to devise training policies for their staff on the information acquired through the monitor.
A specialist agency like RADAR carries responsibility for co-ordination, integration and analysis of data, in other words, for input as well as output.
Making use of the output for new policy development or adapting existing policies, is mostly a responsibility of the mainstream organisations. But a NGO like RADAR will also monitor the progress made.

Signals and observations produced at the output side in the long run may also have consequences for the input side. For instance It’s conceivable that on the output side, a signal is given that a police district hardly makes any observations of discrimination. This may be an indicator that this specific district has a problem with taking discrimination complaints seriously, which leads to underreporting or failing recording. A recommendation would be made to the management to look into this problem. Whether or not this recommendation is taken up in a successful way will manifest itself in the changing number of records (inputside) in this specific district in the next monitoring report.

In this way you see that for the application of signals generated by the monitor output, partnerships – again - are vital, especially with mainstream organisations, because they are the ones with the power to change things.
So far RADAR has invested predominantly in starting input and keeping it going. We are now experimenting with producing output. I want to stress, however, that it has cost the agencies involved, years to build up the monitor as a starting to function operational system.

I don’t want to leave unmentioned that we were substantially facilitated to engage in this developing process by the EU Commission, the Dutch Ministry for the Interior and the City of Rotterdam. The model is also being introduced to other local and regional anti-discrimination agencies in the Netherlands. It will also be part of a transnational consultancy project which will be started soon, lead by the NGO Cospe in Bologna, Italy, in co-operation with the REC. Reading (UK) and Dimitra in Greece. The EU Commission, again, supports this project.

5. Examples of products and their effects
Let me now give you some recent examples of concrete products on the output side:
Discrimination in discotheques:
Discrimination at the entrance of discotheques is illegal but frequently practised in the Netherlands. At the beginning of last year, RADAR did a field-experiment together with other parties to document this practice. This experiment alerted public opinion and RADAR received a lot of individual signals. These were put together with police files on this issue and it turned out that an extremely high number of cases regarded one very popular discotheque. Because it was hard to prove discrimination on an individual level, a case was built on the total number of cases and the lord mayor invited the owner to explain why his entrance-policy provoked so many complaints on discrimination. In fact, this meant a reversal of the burden of proof.

When the owner failed to give a satisfactory explanation, the Lord Mayor issued a formal warning and demanded that he would change his door policy and make his rules for admittance transparent. In fact, the warning is the last necessary step preceding a formal sanction, like a (temporary) closing down of the discotheque.

Underreporting and underregistration by the police:
Research shows time and again that cases of discrimination are underreported and that victims often have low expectations of reporting on discrimination. The actual or perceived threshold between victim and police is often high. Two years ago, we started a publication, together with the Rotterdam-Rijnmond police, which compares the number of cases of discriminatory incidents in the police-files and the RADARrecords.
So far we have published two yearly reports on this issue and the third one will be published within weeks. In these reports we show that up to 50 % of the cases the police gets in, can’t be tracked down in the police information system under the heading ‘discrimination’. As far as I know, this is the first time that underregistration by the Dutch police has been documented on an empirical basis like this.

Overview right wing extremism activities:
Last year we produced a first and overall monitor report on the activities of the extreme-right wing and racist political organisations. In the first nine months of 2000 we registered the double amount of manifestations we counted in the preceding year, which is remarkable, considering that local elections are not scheduled before 2002. We found that most propaganda-activities concentrate on the south part of the city and follow a particular pattern. This led to the conclusion that racist parties reckon their chances to be best in this part of town and will probably candidate predominately in these boroughs.

Sector analysis discrimination on housing market:
When a new and more transparent system of distribution of social houses was introduced in Rotterdam, we expected the number of complaints about discrimination in obtaining a house to go down. This actually happened. On the other hand, we see the number of possibly racist motivated incidents in neighbourhoods go up. These two observations may be related because people who were ‘kept out’ in the old situation, gain access to certain estates, but still are not accepted by other tenants.

6. Advantages of a multi agency monitor
-A well functioning monitoring system will give a better and structural overview of the location and occurrence of different forms of discrimination

-The putting together of a multi agency monitor, based on several information flows, contributes to the reliability and credibility of the observations and also provides information on the performance of different agencies involved. Traditionally, NGO’s often are suspected of and blamed for exaggerating the number of discriminatory incidents. GO’s, on the other hand, are liable to play things down. A multi agency monitor can help in objectifying this discussion and help us concentrate on recognising and solving the most severe problems.

-The monitor is a permanent system and will also create possibilities for assessing the development of discriminatory practices and the effectiveness of measures to combat discrimination over a prolonged period of time. When implemented on a comparable geographical scale (e.g. municipalities or regions) it will create possibilities for comparison of developments between these areas.

-Monitoring systems containing a large amount of data will also offer possibilities for specific item, or group-related analyses and reports, like a picture of perpetrators or victims involved.

-Monitoring -reports will function as an early warning system by illustrating the existence of ‘hot spots’ which ask for extra policy arrangements or other measures.

-They can demonstrate the effects of existing policies and practices. Such as: do the number of reports go down after the introduction of a code of conduct. It can also demonstrate what instruments are missing or which are to be improved (e.g. improvement of internal recordkeeping procedures, production of a referencebook and/or manual for police officers). In this way, it also produces compelling arguments for mainstreaming the issue of promoting equal chances.

-Monitor systems also structure the relation- and communication network of antidiscrimination agencies. Statistics tell us what is happening but not why Therefore we discuss the outcomes of monitoring reports with experts and others involved, like people living or working in certain areas. This structures the exchange of information, not only after the publication of a monitoring report, but also on a day to day basis.

7. Recommendations
-The possibilities for fact-finding, recording and monitoring of racist incidents are in the first place determined by the availability of monitoring structures at the local level.Therefore, it’s wise to give priority to the development of these local
structures.

-Developing models in this field should always happen in close contact with local reality, not from a general blue print, but from an analyses of specific chances and threats in the local situation. In this sense it can be wise to copy the method, but certainly not to copy the model.

-The success of monitoring systems is determined by the way the results are applied in the policy and practice of mainstream institutions. From this perspective, public-NGO coalitions are essential, since they bring together the best of both worlds:

-NGO- sensitiveness and low threshold for reporting of discrimination,
combined with the power of force of mainstream organisations

-If mainstreaming processes are successful (e.g. the priorities in the EU action play against racism), this will create more possibilities and a greater need to look at the detection of structural forms of discrimination from the perspective of multi agency monitoring.

-Partnerships mostly – and certainly in the early stages of co-operation – are linked to motivated and alerted individuals, working within institutions which are often not sensitised as a whole. Obtaining reliable, credible and comparable data on a structural and continuous basis, requires ongoing investment in people, procedures and institutions as a whole, to reach and maintain institutional commitment and to keep the information flow going. This also means that monitorbased systems can only be set up and budgeted on a long term or even permanent basis.

-Monitoring forms an apt illustration of the interdependence of local, national and international policy. The work is focused on and practised in the local community, but supports the activities at the national and even the international level. Institutions at the national and international level should therefore not hesitate to support and facilitate local work in this field