Monday, September 01, 2008

Housing IDPs | Lessons Learnt

In the current situation, one issue facing Georgia is what to do with and for the Internally Displaced People (IDPs) that are now coming from South Ossetia and the Kodori Gorge, in Upper Abkhazia. A friend with significant experience in development work posted a candid analysis in an e-mail group. We thought that the analysis is worth sharing since this issue affects the region.

So here is what he had to say:


"I was speaking about this issue with some Georgian friends just yesterday. They have been involved in IDP shelter work with people from Abkhazia since the mid-1990s. I've done a bit of work with those shelters myself, and would STRONGLY encourage anybody getting involved with shelter issues now to visit some of these old shelters if they have not already done so, and not just in Tbilisi. Some of the shelters I saw in Kutaisi and Samegrelo were horrific. Families have been vegetating in them for fifteen years now, with every politician promising them return next year.

In our discussion, we came up with two big lessons learnt:

1. LESSON ONE: Don't pretend it's temporary

Yes, thankfully, most new IDPs - especially from Gori town - will go home soon. However, many IDPs in Georgia will not. It's not going to happen. Live with it. They weren't ethnically cleansed (houses burnt etc) to pave the way for their return, and Georgia lost the war. The Palestinians aren't returning any time soon, either. No idea how many people this is - let's say Kodori and South Ossetia. [...] We're prob talking 20,000 people minimum.

We should seek to identify those IDPs who will probably not go home soon, and seek durable, humane living conditions for them. Most of Georgia's old IDP shelters are a patchwork of small interventions - an NGO puts in windows, a year later, another puts in floors, two years later there's a new door. Look for yourself - this approach leads to crap results, not only in human terms, but also in terms of value for money. (Repeat assessments are one example - IDPs from Abkhazia have told me how sick they are of assessments by now.)

The key point to remember is that there will be loads of money coming in soon. After that, funds will dry up. By 2006, there was hardly any funding to be had for shelter work. Let's spend it well NOW, while there is money to be spent.

The main prob is that the GoG will probably not look favourably on this approach. IDPs in misery desperate to go home bolster Georgia's claim to these territories. One way to get round this would be to point out that should the IDPs return, the municipalities would get buildings in good shape, ready for privatization or social housing needs. Be diplomatic.

Urban Institute's voucher project was good, but the Georgian housing market was a lot less elastic than they thought. Handing out vouchers to buy flats worth 5000 dollars each to a few hundred families sent housing prices rocketing in Kutaisi (it wasn't enough to money to buy in Tbilisi or Batumi). Vouchers are not feasible on a large scale.

Privatization doesn't solve anything either. One day, your family lives in a crap shelter. The next day, it lives in a crap shelter you own. Without money to renovate, it improves security of residence, but living conditions are exactly the same. How will 50 broke families ever be able to repair their communal roof?

The way I see it, the only humane option is to (1) do large-scale thorough rehabilitation of existing buildings, or (2) to build new housing from scratch (NRC has done the latter on a small scale). A nice side effect is that such large-scale, donor-funded construction efforts will generate a lot of employment in a sector that will be badly affected by the probable fall in investment after this war.

2. LESSON TWO: Don't dump IDPs in depressed or remote areas

Placing people in "no future zones" like the town of Vani, where there is zero employment even for locals, is a bad idea. Also, some shelters in Samegrelo are in the middle of nowhere. Those IDPs who cannot escape get stuck there with nothing to do except to cry, to drink, or to do both. Some have been doing just that for over a decade now. Think poverty trap, think psychosocial problems, think domestic abuse. Also, microcredit or employment generation will not work in the back of beyond.

Maybe they should all be housed in Tbilisi?


That's just my take, and that of some friends of mine. The history of sheltering IDPs from Abkhazia has created lots of "lessons learnt". Ask your more experienced Georgian colleagues and LNGO partners what they think, and please do visit a few shelters from the 1990s outside Tbilisi before you start knocking out funky proposals. Let's use the money that's gonna pour in now wisely, because we are not going to get a second shot at this.

I would be extremely grateful if we could hear the thoughts of more people on this."


This, then, the comment of said person (who works in this field and therefore, for now, preferred not to be named). Any views?


Anonymous said...

Indeed, investing in the “temporary” IDP housing – collective centers/public buildings – is a very unsustainable undertaking. Our follow-up assessments in Azerbaijan has demonstrated that although such projects make a difference for the inhabitants of these buildings, this effect is highly unsustainable. No matter how much community work you do, once you’re out, there is no up-keep and the deterioration accelerates.

While investing in individual, durable housing is much more expensive, the pay-off is so much higher. As for the government’s sensitivity to the return issue, the ownership can formally stay with the state and IDP get the right of use “until the time they get a chance to return”. That’s how NRC did it in Azerbaijan. We also located the new settlement for IDPs not far from their previous place of temporary residence – a Baku suburb. What we learned through our temporary housing projects is here:

Paul C said...

Two options suggested:

"(1) do large-scale thorough rehabilitation of existing buildings," was something the government was looking at as a temporary shelter solution. Unfortunately the buildings identified by the government were 80% unsuitable, leading to questions about whether there are in fact suitable buildings.

"(2) to build new housing from scratch (NRC has done the latter on a small scale)" - and this was the government's preferred option for the longer-term, using prefabricated housing. The problem is the scale and the timing - the critical factor is getting people into shelter for the winter. The danger is that, once they're in those temporary shelters, they're going to have to stay there for another 16 years - while the old caseload of IDPs continues to rot in the pits they live in.

The other proposals seem more viable - rental subsidies and payments to host families. Both would draw on existing stock, would inject funds into the economy in a more distributed way and would not force the growth of a bubble economy around construction (which would mainly benefit the political mafia and NGOs).

HansG said...

a thoughtful response, received in that email discussion on IDPS:

Thank you [...] for summarizing a number of the points in the
discussion. I just wanted to add a couple of points that might seem
inconsequential, but I think that they are important.

1 - 20,000 might not be going home from this conflict, but there are
130,000 +/- that were already living in collective centers and
weren't/aren't likely to go home either. On top of this there are
local populations living in collective centers and/or worse housing
conditions. There are people renting out their space in collective
centers to migrant laborers from Samegrelo... It is important that
any approach keeps these facts in mind. This is one other reason why
I think (and agree with you) that local governments might be a way
in. They are interested in having a functioning local community,
they can cross-cut across these issues.

2 - GoG does have an IDP strategy - although it was developed in 2006-
2007, it contains a lot of good ideas that will be very relevant when
the funding starts coming through. I agree the question will be how
to get the GoG to put the funding to back this document.

3 - To say a few words about the Urban Institute program (full
disclosure - I managed the program from 2006-2007). The housing
market in Kutaisi did not skyrocket because of the program. The
vouchers were priced for the outskirts of the town (meaning a lower
than average price for Kutaisi as a whole). Additionally in the
second year of the project there were 81 families that purchased
housing. Although I don't have the number of housing sales in
Kutaisi for 2006-2007 program year (Sept - Sept), the overall 2007
numbers were 1,357 housing sales in Kutaisi. 81 sales are a drop in
the bucket and did not substantially contribute to the price of
housing rising. The Georgian market did this, and it is still up
195% compared to when we started the program. Do deal with this, we
developed a recommendation for the government of Georgia of the
fundamentals of a flexible voucher pricing system that would be more
in tune to the housing market flucuations than what we used in
implementing the project. The analysis of the two years of the
program showed that income of the IDPs and hence cost of the voucher
was the largest (and only significant) contributing factor to IDPs
being able to purchase housing. So...I would argue vouchers might
still be relevant (say starting early next year) and that they would
be able to address the housing needs of 70-80% of the displaced
population. Subject to a few caveats of course.

4 - A combination approach to shelter is needed. This is painfully
obvious, with so many pilot programs that were going on (CC
rehabilitation, new construction, vouchers, social housing models,
condominium management models) that there needed to be a model that
provided choice to IDPs while being cost effective and providing a
safety net to the most vulnerable IDPs. I would argue that
privatization isn't necessarily bad in some cases because a) some
CC's don't really need substantial renovation to become decent living
spaces (these are mainly ex-hotels I would say), b) privatization
doesn't necessarily mean just privatizing, altough MRA I think was
seeing it that way. Privatization could be used in combination with
renovation which would provide IDPs the same thing vouchers did - a
regular home that they owned (and presumably because they owned it
they would invest in it and maintain it - this theory has a lot of
academic back-up from a host of situations).

Okay - at this point I'm beginning to ramble :-) And yes - there will be funding for the next few seasons that
definitely need to be taken advantage of.

Andy [coordinates available on request]

Paul C said...

20,000 might not be going home from this conflict, but there are 130,000 +/- that were already living in collective centers and
weren't/aren't likely to go home either. On top of this there are local populations living in collective centers and/or worse housing conditions.

The figures I saw showed that there are around 220,000 "old caseload" IDPs living in collective centres right now. This is what really shocked me when I arrived in Tbilisi, and made me worry about the fate of the "new caseload" much more. Once I then realised that many local populations, particularly outside Tbilisi are actually living in similar conditions to the IDPs, I wasn't sure whether to worry less or worry more... Certainly Georgia needs some radical solutions to their housing crisis, but - despite the IDP strategy approved by the government in April - it's likely that the government's fixation on privatization will prevent them from realizing those solutions.

Anonymous said...

You've touched upon a big issue. No one knows the numbers. ICRC did a database of collective centers and came up with a number of approximately 98,000 that wasn't scientific and didn't include very small collective centers (less than 2 families?). MRA has a much larger number based on the number of people registered. In my work I came across some collective centers that are full of registered people, collective centers that are full on non-registered people, and collective centers that are full on paper with registered people, but are empty.

Anywhere between 100,000 and 180,000 I think is a realistic number to program by.