Ten Basic Guidelines for applied researchers on organised crime.
Below we have set out ten ‘guidelines’ to think about when planning, conducting and compiling your research for analysis. They roughly proceed from considerations the researcher should bear in mind at the debut, to those she should remember as the research gathering stage ends. They may seem very basic, but in our experience, first principles are the most important ones.
These guidelines are not exhaustive or prescriptive, and research should ideally be designed and executed with input, formal or informal, from other experienced colleagues. Nonetheless, we hope to sharpen your perception, increase your scepticism, and encourage you to remain alert when sifting unreliable from reliable information. These are the ‘arts’ of empirical research, rather than the scientific method.
1. Identity and the research process
Qualitative research always involves, at base, a conversation between two or more people, and one that proceeds from pleasantries to the subjects at the centre of the researcher’s inquiry. The success of this basic social interaction will depend not just on how much your interlocutor knows, but how much they are willing to share with you, and how they express that knowledge, and this will, to some degree, depend on your rapport. Rapport in this case doesn’t mean that your conversation is friendly or even serious, light hearted or confessional – it simply means there is enough trust and comfort that your respondent will tell you some or most of what you want to know.
Contrary to what many people believe, that rapport is neither automatic nor entirely predictable. It is worth planning for, however, and thinking through how it might affect one’s approach.
The starting point is to understand what your own qualities are and how other people might ‘read’ your identity. Might the fact that people you perceive you as male or female affect how they talk to you? Will your interview subjects interpret your background and position in society as similar or different to theirs? Might you be seen as privileged, or inferior? And is that likely to make them tell you less – or more?
Your identity will certainly affect how your interview subjects receive you – but not necessarily how you might expect. While in some contexts it may be vital that people are approached by researchers who share their appearance or accent or other social markers, in other places people may actually be more trusting of an outsider. Sometimes it is advantageous to be seen not as a highly respected rival, but as a non-threatening and nondescript individual.
Know that every constellation of social signifiers – gender, class, ethnicity – come with both disadvantages and advantages, and also that the balance of these will change according to both context and the individual personality of the person you are trying to convince to trust you with their information.
Sometimes the ‘baggage’ that one’s identity comes with can be mitigated by how you design the research. If it not possible for an outsider to get access, then you may have to hire local consultants to do the bulk of the interviews with which you build up a descriptive picture of a criminal market or network – you could then re-interview a selection of these sources, if you can meet them safely, in order to ask questions that refine your analysis. You might also be able to put respondents more at east by choosing the right environment for the interview to take place – this is why it almost always better to let an interviewee choose a location that they are familiar with and comfortable in, rather than arranging interviews in your office, for example. Mitigating lack of trust may be as simple (or as difficult) as setting up the right kind of introductions to the respondent.
Also reflect on why a source might want to meet you and grant you an interview. An interview is invariably an transaction of sorts in which both sides are giving and getting something. Some interviewees might want to use the research process to spread disinformation, but others might genuinely want to contribute to the knowledge on a subject, they may want to set the record straight on some points, or they might want to obtain information from you in the process. The information requested may vary from innocuous contextual analysis which the interviewee may feel you are qualified to provide in a specific area of mutual interest to more detailed information about law enforcement, intelligence services, or even the competition in the case of a source involved in criminal activity.
Understanding this dynamic can help you recognise how your interviewee might try to manipulate you. It will help you reflect in advance on what you should not be sharing with the for ethical or other reasons and to think of tactful strategies to refuse such requests. Ultimately, reflecting on the transnationality of this part of the research process will help you understand what your leverage in the exchange. This is useful both in terms of potentially improving your information extraction techniques but also as a security strategy in its own right.
In the intense, snake-pit encounter sort of fieldwork that entails access to dangerous criminals and their environments, journalists’ or researchers’ understanding of their value to the source can make the difference between life and death and informs the sort of security plans they put in place. When German writer Jürgen Todenhöfer spent ten days in ISIS-controlled territory in Syria and Iraq in 2014, he and his cameraman were completely at the mercy of his hosts. However, their gamble paid off because it was predicated on months of research and negotiation. Todenhöfer access had been secured by gatekeepers with a real channel to the then leadership of the Islamic State. More importantly, Todenhöfer assessed correctly the Islamic State’s need at the time to project to a western audience – and any potential new affiliates within those societies – that life thrived in their Caliphate.
The vast majority of fieldwork related to organised crime is nowhere near as extreme as Todenhöfer’s example. But the same basic principles apply. Understanding your value to a source in a risky encounter is not only useful in the initial assessment of whether it is safe to engage with a source and eventually arrange a face-to-face encounter but should be taken into account in the formulation of an interview schedule, for instance, and the identification of questions that may imperil this basic understanding upon which access is granted. It should also inform decisions such as whether a meeting should take place at a neutral location or one chosen by the interviewee.
2. Mind how you enter
Be clear about who you are, what your purpose is, what kind of organisation you represent and what the output will be. Be clear about what you do not do – if it is clear you do not do exposés there is less to gain by giving you disinformation (or reliable information, but at least there will be no confusion about anyone’s incentives). Encourage people to provide information for the sake of adding to the historical record, or the creation of an accurate record about current events – incredibly, these reasons can be persuasive.
Always phone and ask for an interview. If you find that your calls are not answered, and the person you are calling is embedded in the underworld, it may be they don’t answer unknown numbers. In this case, send a message to introduce yourself, but follow up with a call if they do not respond. Email is the last resort of the desperate, unless your relationship is highly collegial.
Pay attention to who agrees to speak to you and who doesn’t – this itself may tell you something about who fears exposure. Know that as you begin reaching out to people, news about your inquiries might spread, that then you enter the information ecosystem yourself.
3. Records and locations
It is always great to be able to make audio recording of interviews for you to transcribe, or even selectively consult, at a later point. For background or expert interviews this is often possible, though you should never take it for granted – always ask permission.
With some sources, recording is always going to be denied – it may even set them on edge to ask. This is a very subjective determination. One can usually take hand written or typed notes instead (one should do this anyway, if recording, as transcribing is a very lengthy process), though again, it is often wise to ask. At the very least, write your own notes on the interview as soon as possible after it has concluded. Remember the (supposedly) Chinese idiom – even the faintest ink lasts longer than the best memory.
If you are part of a research team you should also discuss interviews with your colleagues that day, if possible. Check you drew the same conclusions and interpretations of any ambiguous statements.
Do not hold consecutive interviews in the same location, if it is at all avoidable. Respondents could run late or arrive early and then end up crossing paths – revealing who you are speaking to each other, something that can put respondents at risk. If are able to, take the time to scope out several locations in which you can meet respondents that provide varying levels of comfort, discretion, quiet or noise, depending on what will suit them better.
As mentioned, if you are able to meet them at a location of their choosing, this is best (provided you do not have safety concerns about doing so). It may however not always be feasible when conducting multiple interviews in the course of the day in towns or cities which are complicated to navigate. If you or your respondent are worried about surveillance turn your phones off and remove the batteries, and check that your meeting is not recorded by closed circuit TV cameras.
4. Conducting the interview: getting started.
Make sure to begin the interview with a clear statement of what the purpose of your conversation is and what use the information will be put to. Do not expect people to remember anything you said on the phone or when setting the interview up. Clarify whether the interview will be attributed or not, and if the source does not wish to be named, then consider discussing what general description will be adequately anonymising.
Never record an interview without permission from the source, and never pay them for their time – in addition to the usual journalistic qualms about this biasing the information itself, it also blurs the line between you and intelligence agents. It is, however, generally considered acceptable for researchers to buy drinks, coffee or food for your interview respondents, within reason, or cover transport fare when they travelled to reach you.
It may also be useful to think through specific words or questions that might put your respondents on edge, or how you can ask ‘around’ a topic, not raising it directly but still eliciting information about it. Also consider where to place sensitive questions. Often organised crime does not need to be mentioned until late in the interview.
Throughout the interview you will probably draw on a range of different types of questions:
- Open questions – interviews should start with these. They are best for getting new and unbiased information, though responses can be unstructured and irrelevant. A recalcitrant source may need more prodding or direction.
- Leading questions – in most cases to be avoided, but where languages barriers or hostility are present may be the only way to generate a dialogue.
- Option questions – is it this or is this that? These should never be a starting as they limit the way a source will consider an issue, but these are useful later in the interview to nail down details.
- Testing hypothesis – it is very important to ask test your hypotheses – if you have developed them yet – only at the end of the interview, when you have already extracted unbiased reactions. Testing hypotheses with respondents helps to refine your analysis, and practise your own understanding.
Lastly, it helps to be show focus and curiosity in your respondent. As Ken Metzler advised journalists, people will quickly perceive the level of interest you have in a topic: “And when they perceive a lack of interest, they sometimes turn to short, terse answers to avoid the pain of talking to anaesthetised ears.”
5. Assess your source
As a starting point, know who you are dealing with – sources can be anonymous in your final product but they should not be anonymous to you. If a source is not telling you their real name, you need to have much higher scepticism about their information.
Then, as the interview progresses, assess and test the information you are being given.
If a question elicits a surprising or sensational answer, ask it again, using different language and see if you get the same reply.
Along the way, act like a news reporter and do basic, ‘who?’, ‘what?’, ‘where?’, ‘when?’ verifications. These can often expose the cracks in a story and flesh out the details which you will need if you later decide to extract narrative examples from the interview.
Make a distinction between information people have from direct observation or interaction with a criminal system, and second-hand information, which is less reliable.
It is not always obvious from the way people naturally narrate stories which is which – you will often need to delicately ask a source ‘How do you know that?’. While it may feel like you risk offending them by such a question, it also holds promise of eliciting good anecdotes. In the end, it may be one of the most important questions you ask.
6. Sourcing and triangulation
Illicit flows are by their nature hidden so your analysis will be usually be built partially on inference rather than direct evidence (for example, verified documents detailing financial exchanges or findings from legal proceedings). For concrete allegations, rather than trends or patterns, you need details and certainty that your information is reliable. You also need to build up an array of sources which will give readers confidence in your conclusions.
Your starting point will most likely be verbal information passed on by a source – your end point must be corroboration from other, independent, sources. The term ‘triangulation’ suggest how many is usually the minimum: three, unless one of your sources has proof of direct knowledge, or you have a strong documentary source to back it up. You cannot make fundamental claims about a criminal economy on the basis of a single source.
The most revelatory information will unfortunately often come from the least reliable sources. This only underscores the need to protect yourself from misinformation by having as diverse as possible a range of sources.
Even when you consider your best sources trustworthy, maintaining a wide network of sources remains necessary – you do not want to be dependent on an information broker for all your information from a certain government agency or criminal underworld.
7. The tricky issue of referrals.
To speak to people you have to get hold of them, and to convince them to speak to you once you have their attention, referrals are often crucial. In some projects it may be advisable to ask almost every source who you should speak to next, and to gather as many contacts as possible (without revealing if you have already spoken to someone they suggest).
Yet all of this can be quite tricky. Joe Soap may not want you to reveal that they gave you Jane Sponge’s number. And if they have had animosity in the past, mentioning that you spoke to Joe might be the worst thing you could do when contacting Jane! You always have to ask the person who is referring an interview subject to you how you should approach these issues.
Most situations fall into four categories:
- First prize: an introduction.
- Second prize: the person who gave you the contact is happy for you to mention them when you make contact.
- Here be monsters: The person is happy to give you the number, but not to be named as the source.
- More legwork needed: The person will tell you who should speak to, but not otherwise facilitate contact.
As always, there is also a lot to be gleaned about the criminal information ecosystem by paying attention to who will introduce you and who will not – and why.
8. Bedside manner: after the chat
When interviews have ranged into sensitive areas, or involved a tiring process of extracting fine or numerous details from a source, find an easy and agreeable question to end the encounter. This is particularly important when you believe you will be coming back to your interlocutor for more information or another project. In the exchange between researcher and subject no material goods change hands, but the source shouldn’t leave feeling exhausted or exposed, but, hopefully, buoyed by the feeling that they have made a meaningful contribution to your understanding. This goodwill keeps information flowing towards you in the ecosystem.
At the same time, do not feel pressurised to give into a respondent’s post-interview requests as a way of compensating them for taking up their time. Most especially, don’t offer – and don’t agree, if asked – to send a draft of your findings. This is not an offer in the same vein of saying ‘let’s do dinner sometime’ to a casual acquaintance, an invitation that is meant to demonstrate affection rather than stand as a firm commitment. Often people suffer from post-interview paranoia about what they have revealed and will really take you up on any such offers. They will feel aggrieved if you don’t remember or later renege.
While it is imperative to uphold commitments around anonymising your interview subjects, and prudent to clarify or check factual details, your independence precludes you from altering any of your material findings for any other reasons. Promising advance sighting of a draft raises false expectations that a source can control your message, and may lead them to over expose themselves in the interview. Rather proceed on the clear basis that it will not be possible.
There is one acceptable exception to this rule, and that is when your interview subject has been consulted because they have expert knowledge and are contributing background information. In this case, it will be acceptable to show them a draft if they are not materially engaged with the subject. In this case, they will be providing an informal peer review, rather than vetting the analysis.
In this vein, you also do not have to send interview subjects copies of published reports. In most cases both the interview subject and the interviewer forget about such requests. Rather leave your interview subjects with clear information about where your findings will be published and when they can expect to read them there.
9. Placing interviews
It is important to think also, about the order you will conduct interviews in, as far this is something you are able to plan. Generally, proceed from people able to provide background and context, to people who can inform you about discrete, events, people or phenomena, and then back to people able to verify your developing analysis.
At the same time, while you should understand beforehand what it is you want to get out of a source, you might get out something completely different. Don’t try to force the interview in a certain direction, just because that was what you were expecting.
Also, always leave it open that you may return to a source – whether to ask clarifying questions (which can be done over the phone or email), or to interview the subject a second time.
10. Identify the System
The role of the organised crime researcher is to explore individual cases – of murders, extortion, seizures, etc. – but only in the context of understanding the bigger criminal system. It is important not to get lost in a thicket of details, and to identify, as soon as possible, the system that structures organised crime, and to place individual cases within it. Do so knowing that there is always a criminal system. It may be fragmented, isolated, or weak, but it will be there.
One comes to understand such a system, usually, by generating, testing and combining a series of hypotheses about how certain groups or individuals are acting, or how different phenomena are connected.
You can test hypotheses on research subjects. Often people within the system cannot see it in its entirety themselves – but they may recognise it when you propose such a unifying hypothesis. On the other hand, if your hypothesis is wrong, it will be swiftly and repeatedly rejected. It is important, however, to only do such tests towards the end of an interview, when you feel you have extracted all new material, so as not to bias subsequent responses.
The nature of the information ecosystem itself can reveal things about the system. The difficulty you face getting people to talk – and the specific reasons they give for this – may reflect how individuals’ actions are regulated by those with power, for example, by threatening their livelihoods, or their physical safety if they anger criminal, or political, elements. On the other hand, in cases the fact that people talk openly about criminal activities may indicate there is total impunity for criminals. Many other features of the information ecosystem – referrals, which people lie at highly connected nodes, where information is most scarce – can yield clues as to the nature of the criminal system.
Once the nature of a criminal system is understood, it must be analysed using a conceptual framework. This is a different and more advanced topic that the one under discussion in this brief, but an important one.