An initiative of the Woodrow Wilson School of Public and International Affairs and the Bobst Center for Peace and Justice, 
Princeton University

Series: 	Civil Service
Interview no.:	A8

Interviewee: 		Howard Tytherleigh			
Interviewer:		Summer Lopez
Date of Interview:	5 August 2008
Location:		Public Service Reform Unit
      			Freetown
      			Sierra Leone			


Innovations for Successful Societies, Bobst Center for Peace and Justice
Princeton University, 83 Prospect Avenue, Princeton, New Jersey, 08544, USA www.princeton.edu/successfulsocieties
______________________________________________________________________

LOPEZ:	This is Part One of the Civil Service Interview with Howard Tytherleigh, IT Consultant for the Civil Service Verification Project, the Public Service Reform Unit, in Sierra Leone. Interviewer is Summer Lopez. Thank you for talking with us. 

	I'd like to being the conversation just by talking about the role you have played in public sector reform here, as well as elsewhere if you've done similar work in other places. If you could start out by briefly describing your current position and responsibilities, and any related posts you've held over the past several years.

TYTHERLEIGH:	My current position is the IT consultant with the Records Management Improvement Team who is doing the payroll verification for the government of Sierra Leone. We are interviewing every single civil servant on the payroll in Sierra Leone, which is about 14,000 people.

	My background before this was-I was a VSO (Voluntary Service Overseas) volunteer. I came out here to do a different project, which then stalled, and I was adopted by this project. I worked on this project for nine months on the pilot scheme. There was a pilot done of 2,000 civil servants who interviewed and published the results, and now the government wants to go ahead with the main project. I was a volunteer on the pilot and now I'm a paid consultant on the main project. 

LOPEZ:	 Great. Can you just describe the project a little bit for us?

TYTHERLEIGH:	It's evolved from records management, which is an unusual place to start, but a great place to start for payroll verification. The civil service records were in a very poor state. They burnt down the records office a couple of times in deliberate attempts to destroy the records. People weren't keeping records as a sort of a culture, and so, there was a program set up to try to just get the basic information about every single civil servant on the payroll. So, you're just talking name, date of birth, date of first appointment, and a few sort of key details like maybe the letter of first appointment. Some key technical documentation. Not talking sort of Western standards, just enough to prove they exist.

	So, started off doing this exhaustive search from room to room trying to rebuild the records. The team did exceptionally well at actually getting some records on almost everyone on the payroll. That worked quite well and we got that, that's about where I joined the project. They were trying to measure sort of how well they were doing. They wanted a database of a number of anomalies they had found. An anomaly can be where the file data differs from the payroll data, or you can have internal inconsistency.

	I set up a simple database to record all of the anomalies we discovered. That project went very well, finished about Christmas last year. We worked out actually the best-the only other source of information for a lot of these discrepancies and omissions in the data was an interview. Interviewing people when you already have the best possible record you can get is a lot more powerful than just doing a head count. So, we designed a pilot scheme with DFID's (UK Department for International Development) assistance. Well, not assistance, they paid-then-enthusiastic support, and we'd then put 2,000 people, sort of drawn at random, from the Ministry of Health through an interview where we actually present them with the information we have on file, ask them to provide any omissions and to verify any data that we do have. The end result of which is that the data you have has been verified. 

	An awful lot of material came in. The pilot interview bought in thousands of extra pieces of information about years of-but in that interview we photocopy everything. If someone comes along and they've got a copy of a particular form or letter of appointment or something, we record the details and we photocopy it, and we then file it. So, the filing system sort of suddenly took a massive hit and it took us weeks to file it after the end of the program.

	With suddenly with all the filings system up to date, we've got a good computerized record, which we build as we go through, and better still, if you didn't show up for interview, we removed you. Non-attendance of this interview was mandated by the Establishment Secretary as being a sort of-basically, you don't exist, we are going to remove you. We'll probably come back to that in a bit more detail later.

	But the pilot was outrageously successful, interviewed 2,000 people-well, invited 2,000 people for interview, 234 did not attend, and despite massive attempts, so we had an executive order from the Cabinet and we removed 234 names from the payroll. But far, far better than that was that we-with the improved data that we now hold on the people, 100 people were due for retirement. They were properly pensioned off and received their due, you know, pensions and things. So, that's 334 people gone.

	The biggest one for me, the biggest win for us out the pilot was that we then knew who was supposed to be working in each location. We literally stuck it to the wall. We'd go around to the hospitals and we'd pin a list of all the nurses who are supposed to be working there, and believe me, matron loved that. A couple of weeks later we went back, in fact. I learned about this from talking to a guy in a bar, as you do, and we went back and asked the matron how she was getting on, and she loved it. Because she had a list of who was supposed to be working there and all the people who said, "Oh, I'm off on long-term training", "I've been transferred", who hadn't, suddenly were back on the matron's thumb. They reopened two new wards in Connaught Hospital because of the increased attendance of nurses. You don't get much better than that.

	We saved-knocked 234 ghosts off the payroll, 100 people were retired, and of the ones who returned, who stayed and who were valid, more showed up for work. You can see that people were quite chuffed, quite pleased with how the pilot went. So, the government of Sierra Leone sort of-you know, everyone from the President down really, he just said "You're doing it again." It was almost a race of could the government pay before DFID. DFID was saying, "No, no, we like it", you save your money for something else. 

	It was very good. We've set the bar quite high for ourselves, and literally yesterday, we started interviewing every single person. So, we've done a good pilot and now we are going to do everybody else.

LOPEZ:	How do you get people to come to the interviews? You said there was a massive to make sure people attend.

TYTHERLEIGH:	Communication, communication, communication. We got on the radio, and we put posters everywhere. All the posters are, you know, if you know a colleague who hasn't seen this, tell them, because this colleague's out in the wilds. We advertised on the TV and on the radio, and newspapers. I think I've repeated radio a couple of times there, it's quite important in the rural regions, it's about the only quick communications mechanism. So, yes, the second advantage of good communication is everybody knows you're coming and why. And so, if anyone starts kind of pushing back, their motives are instantly suspect. The more public, the more song and dance you make about it, the harder you are to thwart. 

	When we arrive and a Paramount Chief sees us, we are welcomed. Without because any assertions, the people who are in power are the ones who have the most to lose from us sort of taking out ghost workers. So, they are the ones who will be least pleased to see us. But similarly, with the right amount of publicity, they are the ones who also have to help. It's quite an important part of that, if you don't mind me rambling on.

	Okay, where do I start? I've slightly lost my thread there, so I'll come back to that if you ask the next question.

LOPEZ:	I did have a next question, yes. What happens to people who did not show up for their interviews but had a valid reason or would like to redress being removed from the payroll after the fact?

TYTHERLEIGH:	There is an appeals process. Essentially, you have to get your manager to write to the Establishment Secretary-head of the civil service there, to explain why you were unable to attend interview at an actual place of work for three months. And why your ministry was similarly unable to say why you weren't able to attend. Of 234 people, I think around about 30 have appealed so far, and so far the Minister has allowed one back on. No one has been able to say that they've been away for three months.

	I mean, we are not trying to catch the ghosts. Our goal is to get a payroll that is clean, that every single person on it who's being paid exists and has good records. So, the communications helped drive them away. If they know you are coming, they know the questions you are going to ask, and they know they are going to be asked. Publishing the list in the pilot because-of who we were going to interview, but rather than everyone was-you'd have thought it was the black spot of death. The number of people who suddenly moved away-died-but the payroll does a lot of cleaning of itself if you let it. If you make a song and dance and give notice. We said, "Oh, we're gonna take your photograph, we're gonna use the biometrics." You know, almost as a threat, yes, "Get out while you still can. Find another way of ripping the government off because this door is closing." 

	So, you communicate that before hand, because if you corner someone, if you surround the building and come in and say, "We are checking your records and doing a head count", then you are going to get a lot of push back. If you let them go, you end up with a clean payroll and the ones who are, quite frankly, dumb enough to get caught, you refer to the Anti-Corruption Commission. All our results go to the ACC (Anti-Corruption Commission) and are public. So anyone-if they want to start investigating, they are more than welcome, that quite frankly our attitude is that the ghosts stop here.

	We are not the police force. We are not trying to catch them; we're trying to drive them away. That's-took us a little while to work out of "Oh, what do we do if we find evidence of corruption?" We just record it and report it. We don't cry in somebody's ear and sort of say, "oi." That improves the safety of our interviewers as well, and it gives yet fewer reasons why we get pushback. So far, everyone has welcomed us.

LOPEZ:	That was going to be my other question, was if there was much pushback to the trial?

TYTHERLEIGH:	No one has pushed back at all. Some people have said, "Oh, could you do this?" We've had some doctors who've been upset when we interviewed them for fifteen minutes when they got people bleeding on the floor, which is understandable. But there's been no pushback other than the sheer practical "Oh, could you do in this room, you cause a bit of disruption here", just the practicalities of getting every member staffed through a fifteen-minute interview. Technically, fourteen minutes thirty-seven seconds was our average. (laughter) I do the stats.

LOPEZ:	Is there any preparation in place for if there is stronger push back while you're out and doing this more in the field?

TYTHERLEIGH:	Yes, when we're in the field and literally, remember they started yesterday, the first thing they do is visit the Paramount Chief, Chief of Police and the regional-or is it the most senior civil servant in that region or department or city, and introduce themselves to get the support. We've never needed it so far, touch wood, but you know, we will call on the police for security. We do have a small budget to have security if we need it. So far, everyone has been absolutely fine. The worst we've had is someone storm out because he refused to be interviewed by a girl. As it turned out, he was actually not a valid worker. So, he didn't appeal when we knocked him off the payroll. When we invited him to be interviewed by the Establishment Secretary and to appeal, he didn't. That was his bluff and bravado. So, so far, no security issues on a personal level. Obviously, yes, we have to lock our kit up and bag it up in five or six different bays because the laptops will go missing at regular intervals.

LOPEZ:	If you are doing this on laptops now, is it all going to be digital records at this point?	

TYTHERLEIGH:	It's digital records of the interview, so we can maintain a mixed mode. We have-the digital record tells us what we have on file. So, it will have a checklist-yes, we have the original, called a PSC, Public Service Commission Eight Form. So, you have original appointment letter, yes. We have an original medical form, yes. We have the ... etc., etc. So, yes we have the key documents ticked across. When someone arrives with a form we don't have, we photocopy it, so we have the hard copy, and we tick the box. We keep a digital record of what we have and a-but hard copy is still the master.

	We eventually propose an improved version where we are entering in there now, once we've got all this data, start a document management system. Eventually it will recall it slightly quicker and improve security, because files go missing at an alarming rate, even from a locked filing room, but that's part two of our program-three, technically.

LOPEZ:	As far as the trial, have you seen a change in the way that the payroll actually worked? Does it increase reliability of pay for people? Have there been any pay increases?			 

TYTHERLEIGH:	They had it fairly well sorted. There's been a lot of automation of the civil service payroll beforehand. They have a system called IFMIS (Integrated Financial Management Information System) and an awful lot of people are paid by bank transfer. I wouldn't say we've made it any more efficient; we just stop paying quite so many people into spurious bank accounts and things. I don't think we've changed how it's paid, that's sort of the AG's (Accountant General) job. We've just changed the quality of records about who's being paid.

	Interesting one on the payment into bank accounts. The managers, paid up of various hospitals we were talking to, hated it because they no longer had the list of pay-slips. They no longer had to make someone show up and hand over the pay-slip. They felt disenfranchised by the staff being paid directly into a bank account, because they didn't know who was supposed to work there, or whatever. The communication of the payroll sort of had failed, sort of you might say-but it wasn't getting down to the actual managers at the sites. They didn't know-they didn't have an up to date staff list. That's why they loved it when we came along with one. They hated it, but also it meant that the money went straight to the person without being interfered with on the way. So, I would never advocate changing it. If you can pay by direct transfer, do it. But, you've got to have the piece of procedure well established that gets the information down to the right level of management, and that's very low, so that they know who's still working there.

	The old pay-slips they understood and they could manage, but it meant that a nurse could nip-off and work for an NGO (non-governmental organization) up-country and still claim money here. She just had to say, "Oh, I've been transferred." No one had any way of checking. If you wrote to the ministry, the ministry would take six weeks to reply and it would all be forgotten about. So, the nurses just swarmed off to do whatever they like, carried on receiving their pay. We had a nurse who'd been taking her pay for three years and hadn't shown up. Everyone thought she'd gone to the UK (United Kingdom).

LOPEZ:	What are some of the lessons that you have learned from this so far that you think could possibly be transferred to other places? 

TYTHERLEIGH:	Good question. I'm trying to remember the lessons we took out of the final report. But, get the record straight first. Doing a head count gives you a head count. But, if you get your-you know, actually yes, the team will wear facemasks and gloves and go in for your old records, finding every piece of information it possibly could about every single person, and then take it out and use that as the basis for payroll verification. 

	It's really only two questions you're asking. If you got the piece of information, is this right? If you haven't got it, can you give it to us? You agree what information you want to get out with your stakeholders, but pretty much-there's a few which everyone's going to want, which is going to be sort of a consistent mapping of whatever digit or ID to a consistent name, to a consistent date of birth, and first appointment, because those are the two that have a big fiscal head. Then name, grades, designations, place of work. A few of those, get them sorted. Then get as much data as possible first, then computerize, then verify.

	In the verification exercise, we were swamped by the amount of paper. People love keeping their own records. They're better at it than most ministries. People would come in with folders full of almost every communication they've ever received. You can rebuild the file by taking a photocopier in the back of a car. Photocopiers use a huge amount of power, so take a bigger generator than they say you need.

	On the technical side of things, the biometrics work, but they work more effectively almost as a threat to remove the ghosts than the actual need to take photos and thumbprints. Photos are superb because they allow you to review the data with non-technical staff later, much more easily. Once you've got your list of people who showed up for interview, you can sit down with matron and say, "Okay, these are the people who said they work for you." And the matron will say, "Oh, no, I haven't seen her for months, she hasn't showed, she's supposed to be here." So that-a picture does speak a thousand words, but a simple web-cam will do. 

	Just I think-there's also some technical side of things, but travel light. We've just seen one of the other organizations who were trying to do some payroll verification, and the amount of gear they want to install-you can't travel with that, you can't run it out of the back of a car with a small generator. You won't get power, you won't get-we're doing it in the wet season, if you can't carry it and run, it will be soaked. So yes, travel light.

LOPEZ:	What equipment are you using?	 

TYTHERLEIGH:	Every single interviewer has a laptop running Microsoft Access, which has a database containing the entire civil service records, which is encrypted--probably not to UK government standards, but at least it is password protected to a reasonable degree. Then each one has a fingerprint reader, and in the pilot we had a digital camera there to take peoples' pictures as well. 

	As it is, in the main system, the Accountant General is taking with them a digital camera and ID card printing system, so that once we have verified the information, it's printed on a card with a photo and a thumbprint, and they actually the-yes, sort of the 2008 civil service ID card.

	That's the technical equipment. Then there's all sorts of other things; generators, surge protections, UPS's (universal power supplies). The power supplies are dreadful. They'll kill your machinery in seconds. Then there's chains and padlocks for your generator, which you are going to leave outside, which will go missing quicker than--. We actually carry all the laptops in a sort of converted metal box, properly padded. Then there's a small portable photocopier. We tried a scanner/printer combination, and it didn't work. A small portable photocopier, because the volume-as I said, if each person comes in and brings half a dozen documents, that copier works hard. You then pack loads of spare-every spare you can think of, from AA battery chargers for the cameras to spare printer ink. 

LOPEZ:	That's very helpful.

TYTHERLEIGH:	The rest of the kit obviously-a decent 4 x 4, which can take-our interview team is a team leader, three interviewers, and a driver. That's basically two people in the front seats, three in the middle seat, and the back is just full of kit, because by the time you've got a generator, a photocopier, three laptops, and overnight bags for four people, it's a full car. But, it fits in one car. We put a lot of effort into finding and specifying the kit to make sure it did.

LOPEZ:	I assume there are going to be some added challenges by doing this in the rainy season?

TYTHERLEIGH:	Yes, the planning becomes very interesting. You have to give a day to go 50 miles, at times. Yes, you just have to make sure you do your difficult transfers over the weekend, for example. We will do a Monday to Friday in one location, and then finish Friday at lunch. Also, then the team stays on and moves 30 miles up-country over the-this is, actually the transfer is happening this weekend. They're transferring-it's only 30 miles, but it's about the worst road that we're going to travel. They have the whole weekend to do it, so they can be ready on Monday morning at the next location. We're paying all of their per diems and everything to keep them over the weekend. 

	Okay. So, yes, the wet season makes life tough because planning for transport and time taken to get around the place is huge. The capability of your cars has to be much improved. The contingency of having cars breakdown, get bogged in, stuff, etc. is also quite high. Contingency planning is awful. All your kit gets wet at a regular basis. You have to have waterproof metal boxes to put everything in so that you don't lose kit. Your staff gets wet too. 

LOPEZ:	Is there concern about reaching people who might be outside of the capital, in the districts? Or are most people located in the main cities? 

TYTHERLEIGH:	We had a great conversation about that. We worked out that if we applied the pilot mechanism where we went to every location and still wanted to do it in the four months that the government was desperate for us to get results in, we would need seven vehicles. It was actually then going to be cheaper, rather than buy seven vehicles, to pay a taxi to leave Freetown, collect each individual person, bring them here, interview them, and take them back.

	A compromise was struck. We went to every major location, so we're going to every major town in Sierra Leone. Half of the people are actually in the Freetown area, in the peninsula, so we have two interview teams permanently in East and West Freetown. Then we have two teams who basically travel. They travel hard. We are going to every single major location. We stay there for at least a week. We go there at least twice. We expect people to be able to visit their place-these are all the main administration centers. We publish where we are going to be on the main sides, Bo and Makeni we go to three times. You have at least two opportunities to go to your nearest town. That's two weeklong opportunities, plus, you have a three-month opportunity, if you're ever coming into Freetown, to be interviewed. That schedule is in the papers and it's announced on the local radio, when we're going to be here and as we leave a location we say when we're going to come back.

	It's a case of yes, there are people who will have to travel from agricultural research centers and such and such, but we hope that they can time that travel with some other work. If they know we're going to be there that week, they can say, "Oh, well I'm going in on Tuesday, I'll be interviewed then." Because we don't give people interviews if they do not show up. We are there for two weeks in every single location and you come at a time convenient to you. If you miss us in one town, you can come to another. Hopefully we get everybody that way. It's a sort of-it really is a compromise.

LOPEZ:	Excellent. Can I ask you a little bit about relationships between the organization here and the work that you are doing and the donors that are sort of behind this project? And how well that collaboration, I guess particularly with DFID in the case, has gone?

TYTHERLEIGH:	It's interesting, very well in this case. The initial projects had some World Bank funding, but I think they bought a big photocopier/printer, they didn't even-the photocopier didn't show up. The funding took so long and then it comes though sort of a particular government agency, which then took so long. A photocopier didn't show up in the duration of project.

	However, this one-I haven't been on a project that has had such good stakeholder backing. That's one of the reasons we're successful. I mentioned earlier that we got an executive order to get rid of everybody. The ghosts who didn't show up, well yes, the Establishment Secretary said, "Well, I'm not sure what the legal situation is about getting rid of people." So, we talked it through and said, "We need an executive order from Cabinet to back this up." We were working out this in the car on the way home, so came back, and Muniru (Kawa), the project manager, went out and asked for this. That was in the morning, by mid-afternoon we had an executive order signed by every single member of the Cabinet, including the President, on the Establishment Secretary's desk saying, "Get rid of them." That's stakeholder backing.

	When the Minister of Health sees you waiting outside his door, leaves a meeting to come out and sign a press release because it's got to go-yes, sort of to say, "These members of staff have not attended and have got two weeks left", you know you've got stakeholder backing. 

	Our stakeholders are great. DFID coughed up the cash. We were very careful to report back to them and because it went well, it was very easy for them to manage, and for us to do. If it goes well, yes, it's an easy project to manage and the relationship went well. Couldn't ask for much more. Now they are one of our biggest sort of-what's the word? They sing our praises all over. In fact, they are doing a case study on us as well.

LOPEZ:	That's great. Do you have a sense of what has sort of created the environment for the political will to be behind this so strongly?

TYTHERLEIGH:	Political will is useful and having a project manager who just knows everybody. I can do the technical innovation, I can manage a project, but Muniru taught at Fourah Bay College and knows pretty much every major civil servant and an awful lot of the ministers. The ministers have got a real boot behind them. The civil servants sort of will open the door to our program manager and between-and I think everybody's-yes, so you can guess that if you give the dog a good name, and it's everybody's pet.

LOPEZ:	In your experience here or elsewhere, are there several other mistakes that you have commonly observed that donors or international organizations make in their relationships with host country personnel or--? 

TYTHERLEIGH:	I don't think my experience is wide enough to answer that, so, no. 

LOPEZ:	As I said, I think that this is an excellent opportunity for people to learn from a program like this. What is some key advice that you would give to other people?

TYTHERLEIGH:	Start with your stakeholders and make sure they seriously are on board, involve them often. Then, technically follow-the methodology, start with the records, then sort of get-the computerization helps only at the interview stage. Plan your interviews carefully. Verify the information at the interview. Then the interview data is recommended. It comes back as recommendations, which you will view against the evidence. If the evidence doesn't support something, don't change it. It's an evidence-based methodology. You're collecting evidence at the interview, and if someone says, "Oh, my date of birth is the first of the first, 1970", don't believe them until they come up with the evidence and the evidence collected by the interviewer is reviewed by another random member of the team, and on the way-if it leaks, it's a common sense of nope, that's a copy. Or, you know, they signed an original piece of-contract. The original contract of employment says this, so we are going to stick with it. You review the evidence and then you take the data. Other than that--.

LOPEZ:	Are there other challenges that you've faced in working on the project that you've had to work around?

TYTHERLEIGH:	Malaria? (laughter) Typhoid? I haven't had Typhoid personally, but yes. Sierra Leone is not an easy working environment. Your cars break down at awesome rates of repair. Then your staff, almost more so.

LOPEZ:	How have you dealt with that?

TYTHERLEIGH:	Money. You budget to repair-to service your cars monthly. You have a budget to send your staff to doctors and pharmacists, and you take care. If you're lucky, you get Malaria over a weekend. It can be cured--the right drugs take it out in three days, so you're fine.

LOPZ:	Excellent. Is there anything else about the project that you'd like speak about that I haven't asked? Or about any other experience you've had with civil service reform in general?

TYTHERLEIGH:	I think you've pretty much covered it. 

LOPEZ:	Okay, great. Well, thank you very much. I really appreciate you speaking with us. 
 
______________________________________________________________________
Institutions for Fragile States					
Series: Civil Service	
Oral History Program						
Interview number: A-8
______________________________________________________________________

Use of this transcript is governed by ISS Terms of Use, available at www.princeton.edu/successfulsocieties