After smart bombs needlepointed down onto
government buildings in Baghdad, it became quite clear that our skill
at destruction has reached unrivaled heights. At the same time, our
skill at reconstruction remains rudimentary at best. A post-conflict
(or post-disaster) refugee -- if he receives U.N. attention -- may be
given only a blue plastic sheet, about 13 feet by 16 feet, which is
meant to serve as a home. An entire family may live under one of these
sheets for weeks, or months.

A handful of architects
have worked to improve this approach and to provide better housing for
the temporarily displaced. Of the few major names toiling in this
rather unsexy niche, Shigeru Ban, from Japan, is perhaps the biggest
star. He gained widespread notice for the ''paper log houses'' he built
in 1995, after a massive earthquake hit Kobe and left thousands
homeless. The logs were vertical cardboard tubes -- supersize versions
of the thing you're left with when you've used the last paper towel.
The tubes are structurally strong (he threads iron rods through a few
key points to connect them), and the cardboard can be dipped in
polyurethane for waterproofing (waterproof sponge seals the cracks).
Ban also built paper houses in Rwanda and Turkey, where he stuffed his
tubes with crumpled paper, for insulation. His foundations are plastic
beer cases filled with sandbags. It's an ingeniously low-impact
solution: the tubes themselves are available locally all across the
world, can be easily formed from recycled paper if they aren't already
lying around and are again entirely recyclable when the shelter is torn
down.

The other big name in emergency shelter -- an Iranian named Nader
Khalili -- works with the shell structure of domes, vaults and apses.
Khalili's work with temporary shelter has its roots on the face of the
moon. In the mid-1980's, Khalili was one of a group of architects and
engineers invited by NASA to offer new ideas for building lunar bases.
Naming his technique Superadobe, Khalili suggested that astronauts
carry up empty sacks, pack them full of lunar dust and then Velcro them
together into a climbing, narrowing spiral. The end shape would look
like a child's stacking-rings toy. When Khalili brought Superadobe back
down to earth (he now runs an institute in California teaching the
technique), he switched to standard sandbags -- or alternatively, long
hollow tubes pumped full of sand. The idea is that your base material,
be it moon dust or just plain dirt, is almost always right under your
feet, wherever you are.

While Velcro might work for the moon, where there's no threat of
hurricanes, on earth Khalili decided that four-point barbed wire was
the best (and cheapest) option. The barbs act as a mortar between
sandbag layers and grip with a tensile strength good enough to pass
California seismic codes. The shelter's parabolic dome shape deflects
rain and snow, its dirt walls provide excellent insulation and the form
adapts to virtually any scale, from hut to warehouse.

''Sandbags and barbed-wire,'' Khalili says. ''The materials of war now
used as shelter.'' In a post-conflict zone, he theorizes, the homeless
could convert a battle's detritus into a new neighborhood. He even
envisions some therapeutic value in the act. Refugees would require a
bit of training and perhaps a few weeks' worth of labor.

Younger architects have followed in the idealistic footsteps of Ban and
Khalili. Architecture for Humanity, a nonprofit group run by Cameron
Sinclair, 29, has held design competitions for temporary shelters and
mobile medical clinics. Mike Lawless, a British architect, with his
firm, LA Architects, and the help of engineer Mark Whitby, came up with
a plan to use bombed-out rubble as a building material. Packed into
gabions, those wire cages you sometimes see shoring up a highway
embankment, chunks of rubble could replace a toppled wall. By allowing
families to stay in place, instead of shuttling off to a camp, the
scheme may also prevent looting and squatting.

A second entry in Sinclair's contest came from the TechnoCraft group in
Japan. It proposes inflating a large hemp tarp into a bubble shape,
then spraying it with liquid concrete or other hardening agents. It's a
quick, cheap way to form a sturdy dome shell.

One class at Harvard's Graduate School of Design has stripped shelter
to its most basic form -- the clothes we wear on our backs. Toshiko
Mori, chairwoman of the school's architecture department, has been
teaching a course called Weaving Material and Habitation. The class has
been experimenting with Lycra-spandex type materials that can stretch
out to four times their original size. A poncho or sweater might expand
into a 7-foot-by-4-foot tarp.

Of course, all these ideas sound wonderful in concept. And the
designers are heartbreakingly well intentioned. Yet with rare
exceptions, these shelters almost never end up stretching over the head
of a needy refugee. When I asked U.N. workers why all these concepts
remain drawings on a shelf, instead of real improvements in refugee
lives, the general response I got equated to: ''Uh-huh, sounds nice. We
use tents.''

Indeed, thousands of tents. The U.N. High Commissioner for Refugees had
stockpiled something near 100,000 tents in preparation for a
post-conflict Iraq. As U.N workers see it, a crisis isn't time for
cutesy theories, or lovely parabolic arches. It's a time for a quick,
easy, on-hand solution that fits any situation. It's a time for tried,
true and tested. In short, it's a time for tents. And maybe not even
for that. Often, it's a time for a blue plastic sheet you prop up with
a pair of sticks. Increasingly, the U.N. hands these out in lieu of
tents.

The architects argue that architecture can offer better options, but
the U.N. guys rarely get that far in the conversation. They're too busy
making calculations: canvas, double-fly ridge tents cost $80; you can
fit 350 of them, rolled up, on a Boeing 737; plastic sheets cost $6
apiece and use almost no space at all.

Several industrial designers have come up with prefab shelters for use
in disasters. These are mass-produced, quick-assembly huts. One of the
most promising models, called the Global Village Shelter, uses
corrugated-cardboard walls, goes from zero to inhabitable in 15 to 20
minutes and costs $400. But it still won't turn many heads at the U.N.
It's five times the price of a tent. Perhaps more important, the U.N.
guys fear that the first time 200,000 families need instant shelter,
such prefab units won't arrive on time and will take longer than
advertised to assemble. That would mean first setting up a tent camp,
then switching to a prefab camp after a day or two -- the U.N. hates
multiple moves, as refugees relive the trauma of displacement. In
Montenegro, one U.N. worker told me, they used prefab units. When the
fixtures broke, they couldn't be replaced locally, and the whole thing
had to be scrapped. Tents have been around for thousands of years, and
the U.N. has used them since it first started handling refugees in
post-World War II Germany. Institutional inertia is a powerful force.

''The U.N. is crying out for new options,'' Cameron Sinclair says.
''They can't live with risk, in such magnitudes of disaster, so they
say, 'Let's buy tents.' But architects are problem solvers. The U.N.
has actually been inspired by some of the designs from Architecture for
Humanity, but the red tape gets in the way.''

That term -- ''red tape'' -- popped up in nearly every conversation I
had with architects, including Nader Khalili, who has spent more than a
decade meeting with the U.N., offering proposals and seeing almost
nothing come of them. Shigeru Ban spent several years as a U.N.
consultant but eventually left and now runs his own nongovernmental
organization, the Voluntary Architects Network. He first proposed his
paper houses in Rwanda because refugees were cutting down trees to
build shelters and whole forests were being lost. But Ban says the U.N.
told him that his houses were ''too nice.'' ''The refugees would stay
longer, instead of returning home,'' he said. ''I was told I couldn't
provide them something too good.'' One U.N. worker I spoke with readily
admitted this: ''We don't really want to do better than tents for
refugees,'' he said. ''If you improve on what they had before, there's
less incentive to leave a host country and go back home.'' Here's the
sad truth: the kind of prefab, fully plumbed and wired hut we might
provide for a hurricane victim in Florida would be so unimaginably more
luxurious than what a Rwandan refugee had ever known that it would, in
fact, be culturally inappropriate -- an uncomfortable reality in itself.