Disaster Mitigation a Community Approach

7/28/2019 Disaster Mitigation a Community Approach

1/113

A project of Yt3iunteeq in Asia

. . . . I lQisaster Mtt~m. h Commw iF$ 3y: Andrev Maskrey1 Bevelspment Guidelines No.3

Published by: Cbcfam27A F?gn;bc.lryoadOxiord OX2 7DZU.K.

Available from: QxfamPublications274 BanbuT RoadOxford OX2 7DZU.K.

Reproduced with permission.Reproduction of this microfiche document in any form is subject to the samerestrictions as those of the original document.


2/113

AConununity ased App~oac

Andrew Mass


3/113

I1 IONnunitv Based A roach

Development Guidelines, No 3(Series Editor: Brian Pratt)

Andrew Maskrey


4/113

0 Oxfam, 1989

Briti..h Library Catalog&g in Publication DataMaskrey,Andre!w1957-Disastermitigation: a community basedapproach. (DevelopmentGuidelines) 1. Developing countries.Disaster eliefI. Title II. Series363.34809 724

ISBN0 855981229ISBN0 855981237 pbk

Publishedby Oxfam, 274 Banbury Road,Oxford OX2 7DZDesignc;i by Oxfam Design StudioPrinted by Oxfam Print UnitTypeset n 11 point Caramond


5/113

More people die as adi!a!msnowthaninfor this is that morevuherable situatiofirst-thet7lli!s-vlJl.nera

I: ,_I ,f1*implementationwerepeoplesown ofganisationsprovides n o%TTviewf tIIJmha, ?ncualley.mitigationwork in otherworld The concl~ions reaDWMiti~nbave imp0implicationsForall who work WIfacingthethfeatofnatufal~ebook makesa powehl case oragencies nd governmentswithresponsibility or disaster elimitigation o focus more

is a town planner withrienceof disastermitigation.PeruvianNational nstituteas5esment lan for Lima.or of PREDESDisasterResearch entre),he alsom-supported rojects n

ISBN 8-E23-7


6/113

contentsForewordIntroductionAcknowledgementsPart 1 Vulnerability to disasterChapter One Approaches o vulnerabilityChapter Wo Seismicvulnerability in metropolitan LimaChapter GkweeNuaicosand floods in the RimacValley, PeruChapter Four Vulnerability n other contextsPart 2 Disaster mtt~gationChapter Five Approaches o mitigationChapter Six Limasseismicprotection planChapter Seven RimacValley projectChapter El&t The recoveryof CuyocuyoChapter Nine Mitigation pfogrammes n other contextsChapter Ten Implemcn2ng communit:, based mitigation

VViiix

151325394147636991

Further Reading

iii


7/113

ForewordSome months ago, I had the privilege of reading the final drafti of thisextremely important book - one of the first to my knowledge, to focuson the actual experienceof a community responding to a severe naturalhazard that threatened their lives and property with detailed andambitious mitigation action. As I read the text, I was reminded of a pairof memorable conversations, ach reflecting a very different view of theworld.In 1972, embarked on raearch into post-disaster mergencyshelterneeds and provision. C&:- of the first pieces of advice I received camein Columbus, Ohio, from two eminent professors in the DisasterResearchCentre , Henry Quarantelli and RussellDynes. We must warnyou at the outset of your travels that in our experienceyou will hear agross exaggerationof damage and casualties rom disaster mpact andan even grosserunderestimateof local resources o tackle the problemsof disastervictims. Their wisdom &ted to post-disaster ecovery - butlater it becameclear that their perception was also widely applicable topredisaster preventive measures.In the subsequent years, field study in various places continuallyreaffirmed their experience and there was no better evidence than thisthan at a highly influential conferenceheld in Jamaica n 1981 o discussthe implementation of Disaster Mitigation Measures. What becameclearly apparent as the meeting progressedwas a situation where wewere long in ideas, recommendations and advice whilst practicalexperience remained exceedingly short. Thus case studies ofimplementation were gold dust and one field report that sticks firmly inmy memory is that of Andrew Maskrey,which has been expanded inthis publication.During his presentation sat next to a delegate rom the Middle East.He was in a very senior position in a Government Ministry with aspecific concern for earthquake reconstruction planning thatincorporated mitigation measures. He became increasingly restiveduring Andrews talk of mobilisation of the community to map hazards,constmct protective walls etc. Then as we left the meeting he placedhis hand on my shoulder 3nd shook it vigorously - What on earth, heasked, are these people and that speakers organisztion doing? It isclearly the Peruvian governments job to protect its citizens - not toleave it to the erratic performanceof non-governmentalagencies,wholack the necessary esources. responded with the rather unoriginal


8/113

comment that governments n all countries are overstretched n terms ofcash, manpower skills and expertise and should therefore only committheir resources o what communities are unable to do. Therefore, f theoccupantsof this highly dangerousvalley were both willing and capableof introducing protective measures surely this should be applaudedand not condemned. We went on to debate the side-benefitssuch asthe development of local institutions and the potential value of thisexperience in terms of raising the perception of flood risk within theentire community by this form of social mobilisation. He remainedsingularly unconvinced, rapped by a deep conviction about the virtuesof centralisationof power in governmentsand a paternalisticconfitience(probably based more on hope than experience) in the ability ofgovernments to act effectively and a corresponding mistrust of localcommunities.Therefore this book contains a series of vital lessons that can bedrawn about the opportunities and constraintson local communities toinstigate protective action and as such is a commentary on afundamental issue of development. Disaster Mitigation (as Jo Boydcnfrom OXFAM observed n a thoughtful editorial of a few years ago) isessentially the transfer of power, as it increases he: self-reliance ofpeople in hazard-proneenvironments - to demonstrate hat they havethe resources and organisation to withstand the worst effects of thehazard o which they are vulnerable. In other words, disastermitigation- in contrast o dependence-creatingclicf is empowering. *As disaster risks increase due to urbanisation, deforestation andpopulation growth pressures, concerned officials in government orvoluntary agencieswill be wise to reflect on lessons rom the Peruvianexperiencesdescribed so vividly in this book. The community basedapproach may be the only way forward given the frcqucnt pattern ofgovernmental apathy towards their ?CX ;;;zcns and the limitations ofoverstretchedpublic sectors.

Ian Davis,Chair, DisasterManagementCcntre,Oxford Polytechnic.

BOYDEN, Jo, DAVIS, tan, Editorial: Getting Mitigation on the Agenda. Need this!-lappen?,Spcciai Issue of IMlctin 18, University of Reading Agricultural Extension andRural Development Centre, October 1984,P.2.


9/113

BackgroundThis book recounts the authors experiences of a number of disastermitigation programmes developed in Peru. The author worked as aplanner in a government agency, INADUH, Institute National deDesaro!!o Urban0 (National Institute of Urban Development) between1981 and 1983 and was founder member and co-director of a nationalnon-governmental organisation (NGO), PREDES,Centro de Estudios yPrevention de DesastresDisasterPreventionand ResearchCcntre) until1985. The book argues n favour of a community based approach todisaster mitigation, and is aimed at those who work in NGOs,international developmentagencies r governmentdepartments.Between 1382and 1983, NADUR produced an innovative earthquakevulnerability study and protection @-r for the metropolitan area ofLima-Callao. t was one of the first systematicstudies of this kind to becarsifzdout in a Latin American city. At the same time it adopted analternative approach, looking at vulnerability not as a characteristicofearthquake but of the urbanisation process. The mitigation measure5proposed focused on policy changesand legal and financial measuredesigned to avoid the emergenceof vulnerable conditions rather thanjust on physical measures o reinforce housesand buildings.PREKES~ormed by three members of the INAI)UR team, set up acommunity based nlitigation prog:c-ammen 2983 n the Rimac Mey,east of Lima. The valley is one of the most disasterprone areas n Peru,suffering annual floirds, as wcli as landslides and earthquakes. Theprogrammc enabled communities to irnplcmcnt their own mitigationprogrammes through: a) strengthening community organisation; b)providing technical assistanceand training; c) developing mitigationproposals which communities could negotiate with government orinternational agencies. The experience alloived PREDES o adviseNGOs working in other disasterareas: n Colcabamba, Iuancavclicaandin Cuyocuyo, Puno. LLowcvcr,when the exyericnce was presented ninternational forums (in Bolivia, India and Jamaica) t was realiscd thatthe methodology of community based mitigation could be valuable forother groups working in variouscontexts.The conclusions and lessons frulil the INADUR and PREDESprogrammes have been compared with documented case studies ofdisasters in other countries, mainly from Latin America. The risks of


10/113

analysing material rom secondarysourcesare considerableand are fullyrecognisedhere. Nonetheless, he conclusionsof other programmesareremarkably coherent with those of the Peruviancases and add furtherweight and evidence to the arguments n favour of community baseddisastermitigation.The conclusions reached are necessarilyprovisicnal and need to beverified in the light of further experience. There are still very fewexamples of community based mitigation programmes. It is hoped thatthis book will stimulate new practice, encourage the sharing ofinformation between those involved and lead to the setting up andevaluating of more community based mitigation programmes. In thisway, the roles and responsibilities or mitigating disastersmay come tobe redefined.StrucaUreof the bookThe book is in two parts. Part i looks at vulnerability to disaster.Chapter 1 reviews existing theories to identify two different approachesto interpreting vulnerability. Two in-depth case studies then examinevulnerability to different hazards. Chapter 2 examines seismicvuinerability in Lima and Chapter 3 is concerned with vulnerability tofloods and atluvions in the Rimac \lalley. Chapter 4 consists of acomparative analysis of documented case studies on vulnerability indifferent contexts and presents conclusions and reflections on theevidence accumulatedas to the causesand evolution of vulnerability.Part II looks at ways of mitigating disaster. Again, hypotheses areestablishedand examined through case studies leading to conclusions.Chapter 5 discusses he conventional top down approach to mitigation,its limitations and the case Lot community based protection. Chapter 6gives an account of the SeismicProtectionPlan for Lima. Chapter7 is adetailcci case study of community based m;tigation carried out in theHimacValley, and an evaluation of the project. Chapter8 is another ca,scstudy of successful community basod risk-reduction, in Cuyocuyo.Chapter 9 is a comparativeanalysisof documentedcasestudieson bothtop down and community based prograrnmcs rom other contexts andpresentsconclusions on these two mitigation mcxiels.Chapter 10 givesguidelines for NtiQs wishiilg to encouragecommunity based mitigationand establishes a number of priorities and principles for successfulimplementation. Finally, an agenda or the agencies s suggested.


11/113

AcknowledgementsThis work has been carried out with the financial support of OXFAM,who also sponsored he PREDES rogramme. Ian Davis of the DisasterManagement Centre at Oxford Polytechnic gave considerableencouragement o write up the case studies and both he and YaseminAysan read through the drafts. Both Christine Whitehead and MarcusThompson at OXFAM and Ian Davis gave access o the set of casestudy material, which has been analysedhere. The researchand writingwas ca,rried out in AHAS (Associated Housing Advisory Services) aLo&on based NGO of which the author is a member.John and BerthaTurner of AHAS both commented on the draft. The DisasterMitigationWorl&op h,eldat Habitat Forum Berlin in June 1337provided a furtheroppofiunity to discuss he casesand draw conclusions,with colleaguesfrom Peru, El Salvadorand Mexico.

Josefa Rojas, who worked in PRdDES rom 1983 to 1985 and whojointly organised the DisasterMitigation Workshop n Berlin worked asco-researcheron the case studies from other contexts. Her thesis ElPapel de1Trabajo Socialen la Prevention de 10sDesastres:El Casode1Valle Rimac - The Role of Social Work in DisasterPrevention: he Caseof the RimacValley (UniversidadCatolicade1 Peru, 1936) ogether withthe book Urbanization y Vulnerabilidad Sismicaen Lima Metropolitana- Urbanisationand SeismicVulnerability n Metropolitan Lima (Maskrey,A, Romero,Gp PREDES, 936)provided the starting point for the preseiziPaperFinally, thanks are expressed o community leaders and residentsofthe Rimac Valley, Peru, who are the real authors of the ideas expressedhere as we!! as to ail those who worked in or collaboratedwith PREDESin the developmentof its projects.

ix


12/113

Dave EminsodOxfam

X


13/113


14/113

Chapter OtieAPPROACHES TO VULNERABILITYDisaster and VulnerabilityNatural hazard and natural disaster are two very different terms whichare frequently confused and used interchangeably.Earthquake, lood,and cyclone come to be synonymouswith disasterbut, although naturalhazards like earthquakes can be highly destructive, they do notnecessarily cause disaster. An earthquake in an uninhabited desertcannot be considered a disaster,no matter how strong the intensitiesproduced. An earthquake is only disastrous when it directly orindirectly affectspeople, their activitiesand their property.Natural disastersare generally considered as a coincidence betweennatural hazards (such as flood, cyclone, earthquake and drought) andconditions of vulnerability. There is a high risk of disasterwhen one ormore natural hazardsoccur in a vulnerable situation:

RISK= WMERABILITY+ HAZARDThe United Nations Commission for Human Settlements UNCHS -HABITAT) as defined the three terms n the following way:HAZARD - is the probability that in a given period in a given area,an extreme potentially damaging natural phenomena occurs thatinduces air, earth or water movements,which affect a given zone. Themagnitude of the phenomenon, he probability of its occurrenceand theextent of its impact can vary and, in some cases,be determined.WLNERABILITY - of any physical, structural or socioeconomicelement to a natural hazard is its probability of being damaged,destroyedor lost. Vulnerability is not static but must be consideredas adynamic process, ntegrating changesand developments hat alter andaffect the probability of loss and damageof all the exposedelements.RISK - can be related directly to the concept of disaster,given that it

includes the total lossesand damages hat can be suffered after a naturalhazard: dead and injured people, damage to property and interruptionof activities. Risk mplies a future potential condition, a function of themagnitude of the natural hazard and of the vulnerability of all theexposed elements n a determined moment.* UNCHS, Settlement Phning for Disasters,Nairobi,1981

1


15/113

Two Approaches to vulnerabilityMost work on vulnerability has been done during the last decade,although researchers have been looking at the causes of naturaldisasters for a great deal longer. Two opposing approaches emergefrom the disasterequation presentedabove. Most researchhas taken asits starting point the assumption that disasters are characteristics ofnatural hazards. A small but significant body of research,however, hasargued instead that disastersare characteristicsnot of hazards, but ofsocioeconomic and political structures and processes. These twoapproaches have been labelled as the dominant approach and thealternative political economy approach? The second approach hasemerged to a certain extent as a critique of the first. It seemsuseful toconsider the two approachesas ends of a wide spectrum,containing arange of viewpoints and options, rather than as opposed positions.At one extreme of the spectrum, theperception is that disastersareirrevocably causedby the impact of natural hazardson people and theiractivities. Disaster is perceived as an accident; as an unforeseenconsequence of unpredictable and uncertain natural forces; as aninevitable occurrence. Researchcentres on the characteristics of thehazards themselves and attempts to predict their magnitude andoccurrence hrough sciences uch as seismologyand meteorology.Towards the centre of the spectrum, much researchby the differenttechnical professions shows that different kinds of construction andsettlementpatterns eceivedifferent impacts rom hazardswith the samecharacteristics and of a similar size. Research ocuses on physicalvulnerability to different hazards, dentifying the resistanceof differentstructures and materials in different kinds of locations. Muchvulnerability analysis s of this kind.Hand in hand with this research, he human ecology of hazards hasemerged. It attempts to explain the differential effects of hazards notonly on physical structuresbut on people, their economic activities andsocial relationshipsVulnerability is examined through concepts Fuch as maladaptation,the inability to incorporate hazardswithin living patterns and irrationalhuman response o hazard.We common element, which links all these

a ~r~NCIWXER, P, Vulnerability and Recovery n Hazard Prone Areas, Middle East andMediterranean Regional Conference on Earthen and. Low Strength Masonry Buildings inSeismic Areas: Middle EastTetiical University, Ankara, Turkey, August 1986.

2


16/113

levels of analysisynd which gives the dominant approach ts coherence,is the perception of disasteras a characteristic f hszard. Disaster s .seenas a function of hazard which, as causal agent, acts an passive,vulnerable conditions. Everyday life is considered to be normal apdundisastrous.Oralywith the interruption of an unscheduledhazarddoesdisasteroccur.At the othei end of the spectrum, f vulnerability is seen as the effectof social and economic processes, ome intrinsic flaws in the dominantapproach can be demonstrated. In the dominant approach, it issupposed hat people live in vulnerableconditions becauseof their ownlack of know!edge about hazardsor their crroneouSperceptionsof riskor becauseof inefficient decision-makingand management tructures ntheir society. However, the approach is unable to explain howindividual decisions are affected or inlluenc4 by social and economicconstraints. Numerouseasestudies now shlnw hat individuais or socialgruu?s may have very little freedom to choose how and where theylive. For example, low income groups often have no choice exceptvulnerable ocations,such as fiord plains, for settlement.This is not dueto lack of knowledge or an inefficient land use planning system but tothe control of land by market forces,which does not ?!low low incomegroups access o safe bui!ding land.The starting point of the political economy approach is that hazardsare normal physical characteristics of the areas where they occur.Vulnerability is consequent not on hazard but on particular social,economic and political processes. Disaster is an extreme situationwhich results rom theseprocesses.Unlike the dominant approach, the alternative political economyapproach provides ai, analysis capable of addressing social process,organisation and change Large numbers of people on the social andterritorial periphery of the global economic and political systemare seento be disabled by unequal economic r&!ionships which do not allowthem access o the basic resources, such as land, food and shelter,necessary o stay alive. The empirical evidence, rom a large number ofcase studies, points to the fact that it is these groups who most oftensuffer disaster.Vulnerableconditions are far more prevalent n the ThirdWorld than in the First World.The gradual evolution of thinking and research away from thedominant approach towards the alternativepolitical economy approach,has brought with it a new danger. In the analysisof global social andeconomic process, there is a tendency to lose sight of the localspecificity of vulnerability in areaswhich suffer different hazards.

3


17/113

This book maintains that the analysis of specific risks to a givenhazard and the analysis of socioeconomic processesare not mutuallyincompatible. Indeed, the experienceswhich will be presented showthat both levels of analysis are necessary to explain people3vulnerability and their actions n the context of that vulnerability.The way in which vulnerability arises and is perceived will beexamined through two detailed case studies based on first handexperience: seismic vulnerability in Lima, Peru and floods and hudicosin the Rimac Valley, Peru.mmRoMB1*)

BOLIVIA

Map showing Nmc river and valley, Peru.

4


18/113

Chapter TwoIntroductionOn 24 May 1940 an earthquake hit Lima leaving 179 people dead and3,500 njured. Studiescarried out in 1983showed that if an earthquakeof similar intensity occurred again, there would be far greaterdamage obuildings and many more people would be killed and injured. Thiscase study shows that Limasvulnerability has increasedbecauseof theway the city and its buildings have developed and been transformedaspart of wider social processes.Earthquakes in LimaLima is located on the Pacific coast of Peru, where two tectonic platesknown as the Nazca and Americana Plates come into contact. Themovement of the Nazca Plate pushing underneath the Americana Platecausesconsidexrable eismic activity in the area. Until the end of the19th century no less han 2,500earthquakeswere recorded n Peru. Themost important earthquakes o shake Lima occurred n 1687,1746,1940,1966and 1974.It is still impossible to predict when and where future strongearthquakes will occur on the Peruvian coast; at best, one can onlyspeak of probabilities. Enrique Silgado n a book TBY~O/OS n el Pertc,written together with Albert0 Giesecke n 1980,has estimated hat, overa 100 year period, the probability of an 8.6 Richterearthquakeoccurringis 96%. In other words, being in a zone of high seismic activity, one dayanother strong earthquakewill inevitably occur.Vulnerability to EarthquakesDamage n the earthquakes aried in similar structures n different areas,depending on variations in soils and geology. There are many factorswhich affect the number of deaths and the amount of destructioncaused by an earthquake. n buildings, the materialsused, the structure,the height and the level of deterioration are all important variables.Howdangerousa particular ncighbourhood is will depend on the availabilityof open space, he width of the streets, he number of access outes andbuilding density. Other factors o be considered are population densityand the ability of people to escapeand deal with an emergency,which


19/113

depends n turn on training and organisation. ncome and occupationalstructures also affect vulnerability. Poor people have more difficulty inrecovering from a disaster especially if their livelihood is affected.Seismic vulnerability is determined by a complex interplay of all theabove actors.In Lima, three broad patterns of seismic vulnerability can beidentified. In residential areas developed by the private sector orgovernment agencies or middle and high income groups, vuinerabilityis relatively low. Buildings are generally well constructed,complyingwith building regulations and incorporating antiseismic structures,Population densitiesare low. There s open spaceand broad streets. Ifaffected by an earthquakepeoples ncomes give them a better chanceof recovery rom disaster.In the new settlements (pueblos jovenes) on the city periphery,seismic vulnerability is also fairly !ow. Housing is either built fromlightweight materials such as bamboo, or else has fairly solid brickconstruction with reinforced concrctc columns. Population densitiesagain are low. If people are affected, heir low incomesare a handicap,but on the other hand possibilities of recovery are enhanced by aI I

Penny TweedWOxfamtfm4..sesn thepueblosovenes un !be ciiypcripbq are oftar built ofligbtum$jhl~~XTTMSike ban&m in the exw of an eatlhquake, a&thoughthe th.aes might beibislroyai, the de&h rate wdd be reialivdy tow, in cantrasl b he inner slum artxas.6


20/113

generally high ie*Jelof organisation.It is in the inner city slum areas hat seismic vulnerability is highest. .~Houses are built from adobe, a matenal with little seismic resistance. -Structures have been weakened by deterioration and the effects ofearlier earthquakes. There is an absenceof open spaces and escaperoutes and streetsare narrow. Buildings are multioccupied with severeovercrowding. Organisation is almost nonexistent and most familieshave low and irregular incomes. People in these areasare most likelyto suffer disasterand least able to recover.The situation in some innercity pueblos ovenes & also critical.The most critical areas n Lima are those slum areaswhich, becauseofunderlying soils and geology, experience high seismic intensities.Studieshave predicted that in the event of an earthquakeof magnitude8.2 Richter, over 26,000 dwellings would be destroyed or madeunusable and about 128,000people would be left homeless n theseareasSeismic Vulnerability and UrbanisationUrban systemsas a whole and cities in particular grow up as a result ofcomplex economic and social forces and relationships. The formationof Lima on the Peruvian coast at the mouth of the river Rimaccorresponded to changes n the Peruvian economy brought about bythe Spanishconquest. Under the Incas all the agricultural and mineralsurplus produced was concentrated n highland cities like Cuzco in thesouthern Andes. This urban system was integrated internally by anetwork of roads which linked all parts of the empire to its principalurban centre - the residenceof the Inca and a military and theocraticelite.After the destruction of Inca society by the invaders and theincorporation of Peru nto a new political and economic structure basedon the relationship with Spain, a radically different urban systememerged. New settlements were formed in the interior for colonialadministration and to extract minerals. The most important cities werebuilt on the coast to ship Perusmineral wealth to Spain and to importconsumer products for the new elite. Movement was from the interiorto the coastalcities and there was no need for cities within the countryto interrelate. Principal cities such as Lima, Arequipa and Trujillo werelargely isolated from each other. The decision to locate Lima and otherlarge cities on the Peruviancoast was basedon new economic, politicaland administrativeneedsand earthquake isk was not a consideration.The replacement of Spanish colonialism by British commercial

7


21/113

l

interests after independence n 1832 tended to reinforce this territorialstructure., The successiveexploitation of guano, rubber, sugar, cottonand wool was intensified, while the internal market was flo&eci withcheap manufactured products. Although Lima was the political andadministrative capital its influence extended little beyond its ownhinterland, and it remained unconnected to the other principal urbancentres.The great depression of the 1920sand the Second mrld War wereevents which led to a process of industrialisation in Lima. The warinterrupted the importing of manufactured products and createdconditions for the development of industrial growth based on import-substitution. In the 1350s he manufacturingsector achieved a higherpercentageof the Gross Domestic Product than the agricultural sectorfor the first time. The increase in trade, the growth of publicadministration, the construction of new roads, the enlargement of theconsumer market, and industrial development tself were factors whichdetermined the rapid growth of Lima and Callao, its port, and theirconsolidation as a dominant urban ccntrc which integrated the wholecountry in economic and territorial terms. An enormous wave ofmigration occurred to the major cities and especially o Lima. Between 1940 and 1981, the percentageof the population of Peruliving in cities grew from 34.8% o 65.1%. In that time the population ofLima-Callaogrew from 645,172 o 4,608,010, nd it now has over 27% ofthe countrys population and 70% of the industry. The concentrationofpopulation in one capital city and the impoverishment and relativedepopulation of rural areas has been a feature of the changingeconomic structure in many countries over the same period. Theoccurrence of destructiveearthquakeson the Peruvian coast was not afactor which carried any weight in the growth of Lima. Obviously,vulnerability to earthquakes increases dramatically as population,activities and buildings become concentrated n areaswhich experiencehigh seismic ntensities.Form of Building and SettlementThe form of building and settlement which has accompanied urbangrowth is also an important factor in determining vulnerability. TheSpanish form of building, with adobe walls and !;eavy tiled or mudroofs, was not seismic resistant originating as i; did in a region whichseldom experiences earthquakes. Under Spanish rule, the sameplanning and building forms spread throughout the country. Thesuccessivedestructions of Lima by earthquakes n 1586, 1655, 1687,

8


22/113

1746 and 1828 demonstrated he inherent vulnerability of the buildingform.In terms of building form Lima has probably become less vulnerablesince the 194os,as new building has used brick and reinforced concreteand has taken into account earthquake isk, although, as experiences nMexico have shown, much apparently safe building can prove to bedefective and dangerous n the event of an earthquake.The Thnsfofmation of the CityThe fxpansion of Lima since 1940 can be explained as a result of thevarying objectives and actions of different social groups - a wave ofmigrants hoping to guarantee heir livelihood in the city and a powerfulelite sector made up of landowners, the construction industry andfinanciers, who have controlled and exploited resourcessuch as land,building materials and finance for profit. With varying success, heGovernment has tried to coordinate and complement the workings ofurban capital through special financial mechanisms, and use planningand direct housing construction using public funds and internationalloans. The Government also legalised the occupation of marginal landby low income groups and promoted progressive elf-help building as away for low income groups to solve their housing problems.With the establishmentof industrial, commercial and serviceactivitiesin the traditional residential areas of the centre of Lima and Callao,Limas wealthy families moved out to establish new middle classsuburbs. At the same time, while trying to obtain a foothold in theurban economy, migrants occupied cheap rented rooms in areascloseto concentrationsof urban services. The large housesof the old familieswere subdivided and rented out to migrants. n the same areas,emptylots were used to build workers houses specifically for renting. Otherfamilies built their own shacks n vacant areas.Slum housing helped people to becomeestablished n the city. It wascheap, gave good accessibility o work and servicesand was temporary.While slum dwellers are well aware of earthquakesand even have anearthquake saint, which is carried in procession hrough Limas streetsevery October, earthquake risk was not a factor which influencedmigrants decisions o live in the slums. For families with low incomes,the positive factorswere of more importance.For many, living in a slumwas the only option open.Since the I95Os,organisedgroups of families from the slums invadedmarginal desert and hills at the edge of the city to build pueblosjovenes. In the inner city slums there was a fairly rapid population

9


23/113

turnover as migrants became established n the city and looked to thepueblo juvcnes to improve their housing situation.In the 197Os, ew factors intervened in this process. Marginal landbecame more scarce and, as new pueblos jovenes were further andfurther away from the city, costs ncreasedand advantages ecreased orslum dwellers thinking of moving out. With the economic crisis of the197Os, he number of people working in unstable survival activitiesincreased rapidly and for them the accessibility factor became allimportant. Building costs increased far faster than real incomes, andconsequently the possibilities for self-help building were considerablyreduced.Theseand other factors meant that population turnover in slum areasstagnated. What was temporary housing became permanent. Olderpueblos jovenes with a high population density have become new slumareas. Increaseddemand for cheap rented housing in central areashascaused the saturation of existing slum areas and overcrowding in newareas. Yhe availability of cheap rented housing in the older areas hasactually decreasedas slums have been demolished and residentsevictedto make way for more profitable usesand new bkxzksof flats, offices orshops have been built on the clearedsites.In the slum areas of Lima, the population living in old adobe andquincha buildings has ncreased30 or 40 fold in this century, giving riseto an unprecedented ncrease n the citys vulnerability. The formationof slums cannot be explained in isolation. It is characteristic of anurbanisation process n which land, building materialsand finance areunder the control of a small, powerful group, within the context of apolitical and socioeconomic structure dominated by national andinternational market forces.Deterioration of BuildingsFor many low income families there is no alternative to living inovercrowded conditions in areas which experience high seismicintensities. However, their vulnerability is further increased hrough thedeterioration of the buildings.

Two factors play a key roie n the deterioration of buildings: overuseand lack of maintenance. A building deterioratesmore rapidly if it issubmitted to intensive and inappropriate use and if maintenance s notcarried out. In the slum areas of Lima, many of the large, two storeyhouseswrjre originally occupied by one single family with the resourcesto carry out periodic maintenance. Such houseswere never intended tolo occupied by 20 or 30 different families. Under the circumstances, ll10


24/113

Andrew M~/Juvd MeWPscdcsBudding td as his, otigindy hilr for alto amify,arenowm&i- occupied,ndshow eui&nce oJra/d ddmbratian as maintmance is no kmger carried cwt.there we2 apt eatib~ks the buikling tmdd be a death trap.

I$

the installations,especially water and drainage, stairways and passagesaild the structure n general, have deteriorated apidly.Lack of maintenance s due primarily to the tenure patterns.Low rentsgive owners no incentive to maintain or improve their properties. Rentlegislation effectively froze rents in older properties; many families paytoken sums. Many owners do not bother to collect rent, with the resultthat they start to lose control over their property. Through inconclusivelaw suits over inheritance,many properties no longer have a clear legalowner, who could take responsibility or maintenance.Tenants occupying the building collectively have no clearresponsibility for maintaining it and their low incomes prevent themfrom carrying out repairs, except the most urgent ones, for example tothe water supply. There is little social organisation, a further factorwhich inhibits maintenance. Because of the level of accumulateddeterioration, any attempt to restore buildings would now require anenormous investment - beyond the capacities of tenants and ownersalike.Over the years the condition of housing in the critical areas hasdeteriorateddrastically. Every few months, Limas newspapers eport acase of families escaping death as their house suddenly collapsed

11


25/113

around them. In the event of a destructive earthquake, thousands ofsuch buildings could collapse, threatening he lives of ten,3 f thousandsof families.Seismicdisasters n Lima are not causedsimply by earthquakes,whichare natural events which have always character&d the Peruvian coast,but also by the vulnerability of people living in overcrowded conditionsin dilapidated buildings. The increase n vulnerability of the city hasbeen brought about by a complex processof change in the economyand its urban centres. The location and growth of the city, the rapidindustrialisation and massive population movements, the form ofbuildings and urban spaces,and the ownership of land and houses areall factors affected by that process and together make up an evolvingpattern of vulnerability.The people most vulnerable to the effectsof an earthquake n the cityare those with very limited options in terms of access o housing andemployment. The inhabitantsof critical areaswould not choose to livethere if they had any alternative, nor do they deliberately neglect themaintenance of their overcrowded and deteriorated tenements. Forthem it is the best-of-the-worstof a number of disaster-pronescenariossuch as having nowhere to live, having no way of earning a living andhaving nothing to eat. Given that these other risks have to beconfronted on a daily basis, t is hardly surprising that people give littlepriority to the risk of destruction by earthquake,which majl not occurfor another halfcentury or more.Low income families in Lima only have freedom to choose betweendifferent kinds of disaster.Within the options available, people seek tominimise vulnerability to one kind of hazard even at the cost ofincreasing heir vulnerability to another.

12


26/113

Chapter ThreeIntroductionEvery year between December and April huaicos (a quechua wordmeaning violent mudslides) and floods occur in Perus Rimac Valley,affecting:

- human settlements located in the side valleys and on the floodplains of the Mmac, Santa E&&a and Bkuxa tivers;- the raiiway and Mghway which connect IJma to the centralmountains and jungle and which run parallel to the river;- farmland and crops through erosion, flooding and alluvion;- irrigation channels and systems;- Umas drinking water supply.Between 1980 and 1983, 44 huaicosoccurred; the impact of huaicosand floods in the regional economy is enormous. When the centralhighway and railway are blocked or destroyed production losses arehuge, as a large proportion of Perusmineral and agricultural productionis carried on these two routes. In 1784,both routes were blocked for atotal of 15 days. When the level of solid material entering Limasreservoir passesa given level water cannot be treated,supply has to be

reduced by at least 2596and enormous amounts have to be spent oncleaning the plant. Limaspopulation suffers rom water shortage n thehottest months of the year. There is also loss of human life anddestruction of housing. In 1383,35 people died in huaicos,96 houseswere destroyed and 622 people were left homeless. Iosses in 1787were greaterstill.This case study looks at the way the natural hazard arisesand howdisasters are caused because of the vulnerability of people resultingfrom changes in the regional economy and its urban centres andsettlementpatterns.Geodynamic Activity in the Rimac ValleyThe Rimacbasin is located on the Pacificcoastof Peru. The area of thebasin, including its principal tributaries, s 3517square kilometres,whichrepresentsabout 0.25%of the total arca of Peru. The Rimacssource sat 5000 metres above sea level in the Central Cordillera of the Andes

13


27/113

and the river reaches he Pacific at Callao only 120 kilometres later. Ithas 39 important tributaries.As in other valleys on the western slopes of the PeruvianAndes, theRimac Valley has different ecological zones. The lower valley, up toaround 2000 metres above sea level, is sub-tropical desert, with analmost total absence of rain and little or no vegetation except inirrigated areas. Along the coast and for some 20 kilometres inland ahumid blanket of low cloud hangs over the valley for six months of theyear, from May to November. Above 2000 metres cactusand xerophiticplants appear. Eetween 2800 and 3800 metres, the valley is semi-aridwith a substantialamount of vegetation,especially after the rainy seasonin April and May. Between 3800 and 4300 metres above sea level, theclimate is cold and humid and the landscape characterised by highAndean pasture. Above that level, shattered ock and scree ead up tothe snowfields of the Central Cordillera of the Andes. In the lowervalley the rainfall rarely exceeds50 mm a year while in the upper valleyit is over 1000mm, concentrated n the months December o April.While earthquakesoccur periodically, the most common hazards inthe area are huaicos and floods. Huaico is a quechua term applied to aform of water erosion which, although particularly characteristicof theAndes, is common to other mountainous areasof the world. A huaicois a fast and violent flow of mud which may also contain stones androcks of different sizes. Huaicos only occur under particulartopographical, climatic, geological and ecological conditions. In theRimac Valley between lSO0 mctres and 4000 metres above sea level,conditions are extremely propitious for the formation of huaicos.A huaico works like a funnel. In the headwaters of the tributaryvalleys, extremely heavy rains fall over a very short period. Of the rainthat falls, one part evaporates,one part infiltrates the soil and one partruns off on the surface. If the ground were covered by trees andvegetation a far greater proportion of the water would infiltrate orevaporate,minimising run off. However, n the RimacValley, the run offis very heavy. Vegetation s scarceafter the long dry season, eaving thesoils exposed to the action of water. Due to centuries of deforestation,overgrazing and inappropriate managementof soil and water, man hasmade conditions even more propitious for erosion. Huaicos are notpurely geophysical phenomena; there is a strong human influence intheir formation.Becauseof the steepnessof the slopes, run off is extremely rapid,giving rise to sudden upsurgesof water in the valley streams.Given thelack of vegetation and the presence of loose and fractured rock the

14


28/113

runoff is highly erosive, causing small landslides and the formation ofgullies. By the time the gullies have oined together the runoff is alreadycarrying considerable quantities of earth and rock, which increases tserosive power still more.The steepness f the gradient and the narrownessof the valleys meanthat the huaico descends at great speed towards the valley bottom.Further erosion takes place adding even more soil, rock and vegetationto the water. At the main valley, where the side valleys open out andthe gradient les,sens,he huaico slows down, depositing mud and rockin a broad cone. Although every huaico is different,two principal kindscan be identified - alluvions, which are composed of large massesofrock and stone; and torrents, which are mainly water and mud. Bothkinds of huaico are common in the R;macvalley.On days of intense rainfall, huaicos descend simultaneously downthe side valleys and tributaries of the Rimac, ncreasing he flow of theriver dramatically. n the dry season, he river has a flow of only 10 - 1Scubic metres per second. In the rainy season he flow may increase oas much as 200 - 300 cubic metres per second and the Rimac itselfbecomes a giant huaico. In the lower valley, the river erodes its banksand spills out on to the floodplain. The river bed fills up with rocks andstones, causing the river to change course and rise in level each year.While it is not possible to predict when huaicoswill occur, it is perfectlypossible to determine both critical areasand the probabilities of huaicosoccurring in those areas. It would also be possible to develop down-river warning systems o give time for evacuation,but these would beexpensive o install.Vulnerability to Huaicos and FloodsWhile a knowledge of geodynamics is essential to understand why,when and where huaicos and floods occur, it is not sufficient to explainwhy these hazards cause disasters. To analyse the vulnerability ofpeople, it is equally important to look at their economic activities andsettlements.The Rimac Valley is heavily populated, with Lima and the port ofCallao located on the alluvionic cone where the valley fans out at thecoast. This case study concentrates on two urban districts, whichstraddle the river to the east of the city (Lurigancho-Chosica andChaclacayo)and with the rural province of IIuarochiri, beyond.Although administratively part of metropolitan Lima, Chosica andChaclacayoare still distinct towns, separated rom the city by tracts offarmland. Both were formed in the late 17th century around stationson

15


29/113

Amid view of MAWCA~A, Rimac Viz&, showing the ut.hmabk location of the main road, raihaayand buildings in the/low@&&. Erosion is c&~dy v&i&k? on the steep sides of the surroundingmountains, particularly where wetation ColAer s den&.


30/113

the Centralclass. Bob Railway to the Sierra and as winter resorts or Limasupperstill preserve their original urban nucleus situated on highground away from the river. The population growth of both areassincethe 1940s has been through the formation of pueblos jovenes byinvasion on marginal land. In both districts pueblos jovenes haveoccupied floodplain land along the river, steep rocky hillsides and thealluvionic cones of the side valleys. Private housing developmentshavealso been built on the lower alluvionic cones, though most haveoccupied farm land.Above Chosica ies the Province of Huarochiri. Huaroehiri is rural,with a low population density. In an area of 4200square kilometres hetotal population is only 65zOO0. A third of that population isconcentrated in four urban areas: Santa Euialia and Ricardo Palma(which geographicallyare part of Chosica)and SanMateo and Matucana(which lie above Chosica n the upper Rimacvalley). The population ofHuzrochiri is static, natural growth being siphoned off throughmigration to Lima.The Rimac Valley is the major communication route between Limaand the centrai Sierra and Selvaof Peru. Both the Central Railway andthe Central Highway follow the valley from the coast to the pass ofTiclio at 4800 metres above sea level. A large part of Perus mineralproduction for export passesalong the railway while much of the foodLima consumes s brought along the highway, making both routes ofstrategic importance. The river generates electricity for Lima, therebeing no less han four hydroelectric power stationson the rivers Rimacand Santa Eulaiia. The river also provides Lima with much of itsdrinking water, via a reservoir and treatment plant on the easternoutskirts of the city.Location is the principal factor which conditions vulnerabilit-r tohuaicos and floods in the Rimac Valley. The forms and st.ructuIcsofbuildings are very much secondary factors, especially in the event ofalluvionic huaicoswhich buly or destroy all in their path. In the caseoffloods, however, adobe buildings arc more vulnerable than those ofbrick and concrete.

Vulnerability is also determined by non-locational actors. n the samelocation, residents of a high income urbanisation have a far betterchance to resist he impact of a disasterand to recover han residentsofa low income pueblo joven. In general, the pueblos jovenes and theirinhabitants suffer generalised vulnerability, characteriscd by low andunstable incomes, lack of basic amenities, such as drinking watersupply, and poor housing conditions. Vulnerability to hazards such as17


31/113

huaicos and floods is only one aspectof a permanentemergency.Every year people in some pueblos jovenes ose their belongings ortheir homes, representing almost the whole of their resourcesaccumulatedover many years. Recovery s difficult, or even impossiblein some cases. Whereas residents of high income areas can affordsu*bstantialnvestments o protect their settlements rom hazardssuch asfloods and rockfaIls, residentsof pueblos ovenes have no such surplus.The state invests large sums to protect strategic nfrastructure such asrailways, roads and power stations. Little or nothing is invested n theprotection of low income settlements and their inhabitants or to helpthem recover rom disaster.Different 1eveIsof vulnerability can be identified in the va!!ey. Themost.vulnerable areas are pueblos jovenes and some small towns andvillages in Huarochiri which are located directly in the floodplain oraluvionic cones. Those who lose farm land, irrigation infrastructure,water supply and road links are also vulnerable. Better off families withmore land and a variety of sources of income are fess affected ihanthose with unstabfesubsistcneencomes.The whole metropolitan area of Lima-Callaoand indeed the regionaland national economy is vulnerable to huaicos. When the centralhighway and railway are closed, food increases n price and becomesscarceand mineral exports cease.Urbanisation in the Rimac ValleyAnnual disastersdid not always occur in the Rirnacvalley. In Inca andpre-Inca periods we have no evidence to suggestmajor destruction ofhuman settlements from huaicos or floods nor can we suppose thatgeodynamic activity in the area was less ntense han it is now.The ruins of cajamarquilla in the lower valley show that the area, ikeother coastal valleys in central and northern Peru, was a major centrefor pre-Inca civilisation. The fertile irrigated agricultural land was ableto support large urban centres which were located on higher ground insituations where they were not vulnerable to huaicosand floods.After the Spanish nvasion in 1532 new settlementssprung up alongthe coast. Lima, the capital, was located in the lower Rimac Valley.Unlike the pre-Colombians, the Spanish built towns and roads in thevalley bottom.The main period of urbanisation has been over the last 50 years.Since the I94Os,both Chosica and Chaclacayohave experienced rapidurban expansion. Whereasearlier settlementswere located in secureareas the recent urbanisation process has concentrated people and

18


32/113

Andrew Maskrcy/Juvcnal McdiruRrcdcsAerial uietu of San Mateo, Xitnac Vaky. 7Be side ualky in the centre of thephotograph shows bow erosion can be controikd iythe vegetation cover ismaintained and agricultural land is ttwaced

19


33/11320


34/113

buildings where geodynamic activity is most intense, causing annualdisasters o occur. Urbanisation n the Rimac valley has been extensivebecauseof its closeness o Lima and the presenceof two main transportroutes linking Lima to the interior of the country.The only land available for low income families has been invulnerable locations. In both Chosica and Chaclacayopueblos joveneswere formed on the floodplain of the river Rimac, at the end of sidevalleys and on the steep valley sides - land of no commercial value.Better quality land in most but not all caseshas been used for privatehousing development. n the alluvionic cones, n the lower valley, largehuaicos only occur every 50 years or so and hazard risk was not takeninto consideration n peoples decision to settle in the area. The samecannot be said of settlements on the floodplain, where floods occurregularly. In reality, the shortageof land to invade and the need to takeinto account other locational factors, such as closeness o work andavailability of transport, means hat for most people hazard risk assumesa secondary importance. People face a series of problems. Eventhough there may be full awareness of the potential damage to becaused by huaicos and floods, everyday problems such as lack ofdrinking water and insecurity of tenure can assume far greaterimportance for many poor people.Increased ErosionThe occurrence of huaicos and floods can be explained by thegeological and climatic conditions existing in the valley. But if there islittle which can be done to affect either climate or geology, it is possibleto manage both soil and water in ways that either accelerateor retardthe formation of huaicos. There are a seriesof measureswhich can beused to control water erosion:

I mechanical methods (the constructionof terracingon hillsides,drainage channels,etc.);- agronomic measures (crop rotation, contour planting etc.);- agrostological measures (control of pastures);- forestiy measures (conservationand reproduction of treesandforests).All the evidence suggests hat the Incas and their predecessorswereexpert geotechnicians, managing resources n a way that minimisedgeodynamic activity. In the upper Rimac valley large areasof terracingstill exist on the steep hillsides. Although now abandoned and rapidly

21


35/113

deteriorating, this terracing would once have controlled soil erosioneffectively, maximising rainwater infiltration and minimising run off. Inareaswhere terracing has been well maintained there is little evidenceof erosion, demonstrating that the formation of huaicos can to someextent be controlled In addition, the small patchesof native woodlandwhich dot those slopes which are inaccessible o cattle are only vestigesof the natural forests which centuries ago would have affordedadditional protection to the soil in the headwaters of the valleys.Obviously huaicos and floods did occur in pre-Colombian times, butevidently there was not the uncontrolled erosion which is taking placein the valley today.The control was possible becauseof the way in which society wasorganised and resources were used. The extremely sophisticatedtechniques or managing soil and water - building terraces, illing themwith fertile soil brought from the valley bottoms and irrigating themthrough a complex systemof channels- depended on a tight social andterritorial organisation.Under the Incas, he population of Peru was roughly the same size asit is today. Although it is important not to idealise precolonial societies,as far as we know everyone was well fed and food reserves weremaintained to allow for variations in production through drought orclimatic change. Security rather than vulnerability characterised thesociety.After the Spanish nvasion, this security was systematicallydestroyed.The indigenous population becamea resource o be exploited and wasdecimated through European disease and ill-treatment, being used asforced labour in the mines and in the large ranchesor haciendas,whichthe colonists set up. Indigenous farmers were driven on to marginalland where subsistence was only just possible. The Spanish wereinterestednot in maintaining a high level of agricultural production butin extracting gold and silver for transshipment to Spain. Existingindigenous crops and farming systems were not respected. Instead,crops, animals and farming practiceswere brought in which were aliento the Andes and not ecologically suited to the region. Because thesocial organisation necessary to build and maintain enormousengineering works such as terraces and irrigation systems wasdestroyed, it is not surprising that since the 16th century these havesteadily deteriorated. Once maintenanceof the terraceswas neglectedand cattle and goats started to destroy the vegetation cover, erosionprogressively ncreased.In the present century, the rural areas in the upper valley have

22


36/113

become steadily depopulated through migration to nearby towns or toLima. The system of subsistence agriculture based on the peasantcommunity and largely outside of the money economy has beenreplaced by production governed by market orces.Becauseof low prices paid for agricultural products, n the interestsofthe urban market, producers earn barely enough to cover their costs.Peasant armers are forced to migrate and their land is put to moreprofitable use. Areas n the Rimac valley formerly used for cultivatingpotatoesand maize are now grazed by cattle and goats. Meat and dairyproduction is far more profitable than potato growing and employsfewer people. With the exodus of the population and the change overfrom subsistence crop production to commercial cattle raising, theremaining vestigesof traditional social organisation have broken down,accelerating he deterioration of terracesand irrigation works.New employment opportunities in mining, commerce and servicesalso encourage the move away from agriculture. Wages,although loware considerably higher than can be gained through crop growing.Agriculture becomes a secondary or tertiary activity and populationshifts from the small villages on high ground to large.:stablished ownson communication routes in the valley bottoms. This has been veryevident in the Rimac valley. Limas concentration of industrial andcommercial activities has also been a strong catalyst for migration andthe change in the employment structure. The availability of alternativeemployment opportunities and easy accessibi!ity o urban markets hasincreased migration, destroyed agricultural production and urbanisedthe population. In Peru t is only in peripheral areaswhere accessibilityto markets s more difficult and other employment opportunities do notexist that traditional agricultural patternsand social organisationpersist.The result of these changes, in the upper Rimac valley, can besummarised n terms of overgrazing,deforestationand the abandonmentof terracesand water management. This accelerates rosion which inturn increases eodynamicactivity.ConclusionsThe RimacValley presentsvisible symptomsof progressive isaster. Thevulnerability of the regional economy and its urban centres is aconsequenceof socioeconomic, political and territorial changes whichbegan in the 16th century. The modcrnisation of the urban economy,the destruction of the rural economy, the closeness to Lima, theexistenceof major communication routes to the interior and vi Blatare inany case highly propitious ecological conditions for huaicos and floods,


37/113

have led to the progressiveaccelerationof geodynamic activity and theurbanisation of vulnerable locations - creating conditions for repeateddisaster. t is equally clear that the specific factors causing vulnerabilityin the Rimac Valley are in turn influenced by the wider forces ofnational and internationaleconomicsand politics.Huaicos and floods in the Rimac Valley affect predominantly lowincome families in pueblos jovenes,who have no option but to occupyvulnerable land. Living conditions in general n the pueblos jovenes arein a permanent state of emergency, character&d by lack of drinkingwater, poor accessibilityand inadequatehousing. The effects of huaicosand floods are only one aspectof this continuous disaster.h the Rimac Valley, it is clear that not only do natural hazardsaffectsocial and economic structures, but social and economic processesaffect ecological conditions, accelerating nd magnifying hazards.The disaster is a downward spiral. Destruction of houses andsettlementsmakes people even more vulnerable. Unable to find newand better locations to live in people are forced back into the same oreven more vulnerable conditions. Vulnerability character&es ife in the pueblos jovenes. The periodicoccurrence of huaicos and floods only serves to exacerbate thesituation, making poor people poorer and creating conditions for newdisasters.

24


38/113

Chapter FourVUUERAB~IN OTHER CONTEXTSA Comparative Andy&The two Peruvian case studies have shown how, in two contexts andfaced with different hazards, vulnerability to disaster evolves withinhistorical processes of economic transformation and urbanisation.Although the hazards are different the causesof the vulnerability areremarkably similar. The way in which people assign priority to risk ofnatural hazard, is complex. From the case studies, it is clear thatpeoples actions are not irrational. When daily survival itself is fragile,people are forced to adopt settlement patterns, housing solutions andeconomic activitieswhich make them vulnerable o hazards.Vulnerability to hazard may be chosen as a lesser evil, because hehazard is infrequent, compared to the more pressing day to dayproblems of homelessness, ack of income or inaccessibility. Poorpeople are vulnerable to hazard as a result of processes that havedeprived them of any power to affect their own physical, social oreconomic environment. Their only freedom of choice is to choosebetween different disasters.These conclusions are not Uque to central Peru. Nine documentedcase studies from other contexts have also been examined. These arenot meant to be a conclusive and representativesample but serve toprovide additional evidence of the ways in which vulnerability arisesand manifests tself.Casestudies on VulnerabiMty1. ARGEiWTNA:El Gran Resistettcia, Cbaco.wnd of Hazard - FloodThe vulnerability of the area o floods increaseddue to two factors:The regulatory efficiency of the water-soil-vegetation system of theChaquenaregion in the basin of the river Plate has deterioratedrapidlyowing to changes in land use and an inadequate management ofrenewable natural resources. ?he deforestation n the upper reachesofthe basin produced a reduction in the interception of rain water and inthe infiltration of rain water in the ground.

25


39/113

The process of urbanisation of Gran Resistenciahas increasedoverthe last few years. Urban growth has overcrowded existing urban areasand occupied floodplain and low lying areas.Source:GWJTO, HARDOY,HERZER,Lz Inundation en el Gmn Resistench Provinchdel Chaco, Aqp*Wt) 19% 1983, n &sastres Natural& y Saciedaden AmericaLatina, Crupo Editora &noamericana, BuenosAires, %j.

2.lUZA2!Ik Sertao (artbeast)Kind of Hazard - DroughtResearch arried out since the 1970sshows clearly that the effect of thedrought is concentrated selectively in a homogeneous social group:those who do not have access o land and who suffer all the symptomsof rural poverty. This part of the population depends, or its subsistence,on the annual food crops which are most vulnerable to the drought, incontrast o large landowr,ers,whose basic activities are cattle raising andcotton growing.The problem of the drought is not just a problem of lack of water.Nor is it just a problem of the ecological and economic adaptation ofeconomic activities. It is fundamentally the problem of the poverty ofcertain social groups.According to the perception of the victims of the drought: theproblem could not be reduced to the simple natural and technicaldimension of the lack of water and the loss of crops. Among theprincipal problems mentioned, two thirds refer to a wider and morecritical social dimension: poverty, hunger, llness, ow salary.Source:PESSOA, irceu, Sequias n el Nordcstedel Bras& de la CatastrofeNaturala laFragi.lidad ocial, n Lkmstm Naturales Socikabd n AmericaLatina, Grupo EditoraIatinoamericana,BuenosAires,1985.

3. DOMINION REPUBLIC Monte Platq &ryaguana,- SabanaGrande de BoyaKind of Hazard - Cyclone.Three kinds of specific vulnerability are identified: vulnerability of rural

26


40/113

communities to flooding in low-lying areas; erosion and falling cropyields in upland areas; vulnerability of urban shanty settlements onmarginal and to hurricanes:The vulnerability is created by certain social and economic processesimplicit in state and private development of commercial agriculturalproduction at the expense of small scale producers. Such processesinclude the concentration of land into larger units for the purposes ofmechanisation,entailing the dismion of existing producers eitherby coercion or by the play of market forces, such as the rise in landprices which (ii the context of inflation, rising prices in basic oodstuffsand consumer goods, combined with the small-scaleproducers imitedaccess o credit in order to increase their production) encouraged thesmall producers to sell their land and seek alternative sources oflivelihood.In the cases presented, the commercialisation of agriculture (inparticular the extension of sugar cane production, sensitive tofluctuations in prices in the world commodity market) has not onlyentailed widespreaddeforestationof fertile lowlands but has also forcedsmall-scale armers into marginal areas either to farm patches of landaround the sugar cane or to invade the poorer hill areas (leading toover-farming and erosion). At the same time it has also been the causeof migration to the urban areasand the growth of shanty settlemenWSource: JEFFERY, usan E, The Creation of Vulnerability to Natural Disaster:CaseStudies rom the Domirkan Republic,Disasters ol.6No.1, 1982.

4. EWADORz QuitaKind of Hazard - Floods, mudslides, andslides,earthquakes.This case analyses the vulnerability of the inhabitants of peri-urbansettlements n Quito. The vulnerability hasvarious dimensions:Popular settlements present problems of insecurity and fragility,both in physical terms, with geographically inadequate sites, withgradientsoften of 600/a nd easily subject o landslidesand in legal termsgiven that most settlementshave not obtained egal recognition.the concentrationof a deficit of infrastructure,access, ransport andservices n the popular settlements.The location of popular settlements in areas which are prone tofloods, landslidesand mudslidesdue to:

27


41/113

- the increase n urban population and its land needs;- sudden ncrease n land prices n Quito due to speculation;- the processof urban renovation n the centre of the city, whichpushes ow-income residents o the periphery.The population of the settlements live in a situation of permanentemergency, which acquires dramatic characteristicswhen affected bynatural hazards.The overall deterioration of the real incomes of the peri-urbanpopulation due to the economic crisis of the country, at the same timeas a rise in food prices becauseof the loss of harvestsand destructionofagricultural zones n rural areas.Source: GARCIA,Jorge, Los DcsastresNaturales Afectan a 10sMas Pobres, n Lksastre sNatural&s n AmericaLatina, Crupo Editora Latinoamericana,1985.

5 INDU: Attdbra PradeshKind of Hazard - CycloneDifferential vulnerability to cyclones n the Krishna Delta area is not somuch a function of location as much as of assets nd resources:Size of family and consequent increase n labour power is closelycorrelated with an increase n assets nd income opportunities.Illness plays a major role in the decline of households economicposition, with subsequent orced sale of assets o meet medical costs.Small householdsseem to be particularly vulnerable.The vulnerability of a household is determined by the interactionbetween its resources and assetsand external forces acting on themeither singly or in combination. Ihc key household characteristics rcits size, structure and the age and health of its members. The key assetsare those from which income is earned. f Iouschold characteristics ndtheir resource and asset position are partly determined biologically butmainly by the income opportunities in the environment. These ncomeopportunities are largely determined by the historical interrelationshipsbetween people and their cnvironmcnt and government policies and theenvironment itself which is subject to the climate and exposure tonatural phenomena.The relationship between pcoplc and their environment and thehistorical social relations of production were the result of individual and

28


42/113

collective decisions made about income opportunities, on the spot,whereas todays land use and resource allocation policy decisions, bytheir complex nature, are made by people who live outside the area inquestion.It was also of interest to find that perceptions of risk, in terms of theprincipal threat to livelihood, had changed and were more stronglycorrelated with socioeconomic status, than with housetype and/orlocation.Source: WINCHESTER, Peter, Community Vulnerability in a Cyclone Prone Area ofSouth India; Housing Provision and other Government Pohcies, CIEW73 conference,India, 1984.

6 PMISTIXNz Bakdti Village, YasitrKind of Hazard - Floods, mudslides, andslidesand earthquakes.Vuhrerability was summarisedas a combination of the following factors:1. A lack of surplus produce for trade and a lack of cash capital forentrepreneurialactivity.2. Increasing pressureon land resulting in le.ssproduce per hectare.This mainly results from abandoning the traditional practice of leavingfields fallow for between one and three years. The potential for seriousfood shortages s considerable and in addition, there have been recentsevere government restrictions on hunting activities, previously avaluable protein source .3. A lack of adequate communication, especially of vehicles androads to facilitate marketing of any surplus.4. A lack of employment opportunities. Road construction work isseasonal, in the Gummer, which is precisely the time of heaviestagricultural commitment.5. Increasinguse of land which is subject to flooding.6. Little real opportunity to cultivate virgin land, mainly because t isnot available; irrigation is an extremely tricky process and it is simplynot cost effective to construct channels on unstable slopes where theywill be destroyed by even a minor landslip.7. In the longer term, extensive logging of the hills, which in turnmakes them more unstable and thus vulnerable to land slip, and aconcomitant shortageof wood plus a lack of cash to buy kerosene orfuel.


43/113

The ability to respond to vulnerability depends largely on theavailability of surplus economic resources....Thepoorest of the poorvery often dont survive in the community and become migrants tourban areas,part of the squatterpopulation living in the slums of largertowns.YIhe location of the poorer members of the community along themargins of the rivers is additional evidence or the notion that obtaininga livelihood today is more important than some possible future dangerthat may or may not happen; precautions are only taken for thosehazardswithin the economic meansof those under threat, all other fearsare suppressedn order that day to day existencecan continue.Source: DSOUZA,Fran&, The SocimEconomicCost of Planning for Hazards AnAnalysisof Barkulti Village,Yasin,NorthernPakhn, XntenwtitmalKardxwwmProject,CambridgeUniversityPress,1984.MOUGHTIN,C, Bark&i in the YasinWey: A Studyof Traditional SettIement orm as a Responseo EnvironmentalHazard, ntemdonalKunakotum fiojecf, CambridgeUniversityPress,1984,

7. PERU a&scaKind of mzard - DroughtThis case analyses the severe drought which affected peasantcommunities in Perus altiplano in 1983. The vulnerability of thecommunities o the drought was due to three principal factors:

1. The amount of cultivated land per person, shows crudely theextreme shortageof land in the southern Andean region.2. The gradual processof exhaustion of agricultural activity over thelast two decadesdue to growing indices of overgrazing.3. The Agrarian Reform of the 1970sdid not manage o transform hestructure of land ownership. Rather han redistributing, t concentratedland in an associative orm.Source:LOVONZ. Cerardo, El SWAndino Peruano la Coyuntun de Sequia:19821983, n fksasttes y Sociedaden America Latina, Grupo Editora Latinoamericana,BuenosAires, 1985.

30


44/113

& PERU PftiraKind of H&d - Fmd

the economy of the zone and which pershted unt.l thecentury, by all economy Of large Planhtions p&u

nsive flooding whichMnrcture (irrigation;---I ?od supply. Thefarmers, wit h aIn well suites toend of the I19th.cing cotton forexporLTht disapp opriation of land began iqcentur the demand for cotton on the beginning of the presentt rYtWhe%ustry, attracted foreign c- wor1dmarket generatedbyL, *-.IIC textile ifl - - Pita1 o the area. Local land)wners built flew irrlgatons forced peasaQs ff the land and used newI echnology to transform the Piura Valley,1 ermanent Or seasonal worker depend The smallholder became aent on a salary. The new

SuuLlr---Je ($ ja d Ownership broke the elationship between he localPoPulation $0d the natural resources,b hich no longer belonged tothem (not oply land; also forest, Pastures ater and villages)m-l7ne intro dUctiondemographic increase. Of the Plantations coincided with a rapidBoth smallholding8 and settlementswere forcedout of fertile areas to *e limits Of *e de~rt. For the smallholders hismeant povert$ social dbintqption, pmqous living conditions and thedestruction 0f a productive em wehadaptedo the ec&gicalconditions of *e area.The Agra~~~rn~~~~t~oor,~9~~a~~~~~ed he ownership of theplantations, red the same.people depend However$ the productivesystem emaifl . .conditionsidfl mPmve ed on a salary and their livingOn the coPtrarybefore the plantations:"A first and 5uperficial investigation, heh relation with rivers and r\0, discovers n the culture ofthe zone a ric Qods,which couldnt be anydifferent whefl life in the region always debendedon them I).

Source:FRAN&4 Eduardo, El DesastreNaturalenAmeicaLatina, GflPo Editora atinoamcricana, u ha, in Lksastm y Sociedaden%osAires, 1985.31


45/113

9. TURKEY=GedizKind of hazard - EarthquakeThe way communities cope with vulnerability is not a simple learningfrom experience process but a complicated series of adjustments tochange. Where there are many forms of vulnerability and the resourcesto avoid them are limited, optimisation of these resources n relation tothe IMa1 and household priorities become the key measure. Thisprocess is an intuitive and experimental one and arrives at certainsolutions over a long period of time.[These f2w examples]... outline the delicate balance that the localcommunities have to maintain for economic, physical and cuhuralsurvival at the cost of incr2asingone type of vulnerability for the sake ofavoiding the others. One can say that some of these problem could besolved if there was a better understanding of safe building, or if thelocal materials and techniques could be improved. However,improvement for the communitieswith limited economiesand sourcewas a matter of choice, each time optimising these resourcesaccordingto their priorities.Source:AYSAN,asemin, Community Vulnerability in Rural heas; Gedit; Turkey,CIB/'W3Conference,ndia, 1984.

The Evolution of VulnerabilitySpecific characteristics f vulnerability evidently vary from area to areaand context to context. The cases, however, show up a number ofrecurring elements.In rural areas these include unequal access to land, inability toproduce a surplus, absenceof other income earning opportunities, noaccess to credit, forced migration to urban areas, ecological damagecaused by excessive pressure on natural resources and shortage ofmaterials or housing.These characteristics have emerged historically through thetransformation of local, regional, national and international economiesthrough mechanisms such as unequal distribution of land betweensocial groups (2. Brazil, 5. India, 8. Peru); &appropriation ofagricultural land from small farmers by national and foreign capital (2.Brazil, 3. Dominican Republic, 8. Peru); displacement of subsistence32


46/113

production by cash crops or cattle raising for external markets 2. Brazil,3. Dominican Republic, 8. Peru); introduction of new technologiesandcrops (8. Peru) and the failure of agrarian reforms to affect the landownership (7. Peru and 8. Peru).In urban settlements, he elements nclude the siting of low incomesettlements in vulnerable locations &Dominican Republic and 4.Ecuador); ncome levels which barely cover subsistence, eficiencies nservices, infrastructure, and transport in low income peri-urbansettlements, ack of access o income earning opportunities (4. Ecuador);poor housing conditions and deterioration of existing housing in oldinner city areas 4. Ecuadorand 9. Turkey).Urban settlements have also been affected by population increasesthrough natural growth and rural migration (3. Dominican Republic and4. Ecuador); and speculation and urban renewal in inner city areas (4.Ecuador) and rebuilding following major disasters 9. Turkey).The ertidence from the ~VVOeruvian cases ogether with the othercase-studies shows that vulnerability is a consequence of particularsocioeconomic processes nd structures ather than being characteristicof hazards.Vulnerability to any kind of hazard s essentiallydeterminedbY poverty-The effects of the natural hazards and the ecological and economicmaladaptation of activities, settlements and building types were alsoevident in the cases analysed. However, rather than being causes nthemselves, hey were manifestationsand characteristics f patterns ofsocial and economic inequalities. n most eases,people were living in astate of permanent disasteranyway, even without natural hazard. Theconcentration of economic power in a dominant minority; theintroduction of centralising echnologieswhich disappropriate esourcesfrom the majority and the subsequent social, economic, political andterritorial marginalisation of this majority are the mechanisms hroughwhich vulnerability emerges.These mechanisms re characteristics f aninternational process of the accumulation and concentration of wealthand power.This process involves a global division of labour and a resultingevolution of vulnerability at the social and territorial periphery of theworld economic system.Vulnerability at the social and economic centresof power has been virtually eliminated by being exported to theperiphery. Underdevelopment is not a stage on the way todevelopment but an inevitable consequenceof development.The morethe centre overdevelops, the more the periphery underdevelops andbecomesvulnerable.

33


47/113

There is, however, an integral contradiction and instability built in tothe current world economic order. The increasing ncidence of disastersand the destruction of the ecosphere consequent uponoverdevelopment and the exploitation of natural resourcesmenace thestability of the system. Extending the limits of exploitation of marginalgroups and of the environment and thereby increasingvulnerability maybe essential to continued accumulation of wealth and power and theovercoming of successive crises, in ihe short term. However thedevelopment of self-generatedand self-perpetuatedcrises in differentmodes of production may eventually lead to their renovation.Vulnerability as itself a cause of the crisis will eventually createconditions for change and transformation.Response to VuherabilityIn some traditional societies,where the social organisationguaranteedafairly equitable access o resources, people, their activities and theirenvironment were in a good ecological balance. Societies andcommunities were able to adjust and adapt to hazardswithin the limitsimposed by knowledge, resource availability and technology. But inother traditional societies life was far from secure. Profound socialinequalities existed, eading to the marginalisationof certain groups whothen become vulnerable to different hazards. As traditional societiesbecame incorporated or coerced into market economies andincreasingly subject to external pressures and constraints which theycould not control, vulnerability increased, while the availability ofresources and the possibility of avoiding hazards decreased. Peoplecannot adjust or adapt XI hazardsbecause hey no longer control accessto resources,production, exchangeand distribution mechanisms. Theseare controlled by market forces, governmentsor other external agents.Daily life becomesvulnerable and fragile for many social groups.Vulnerability can best be understood as a relationship between theforms, meansand relations of production. Attention focusedon specific,usually physical, aspectsof vulnerability is leading research urther andfurther away from a real understandingof why people suffer disaster. fccintinued researchon hazardsand Iheir effects on specific forms is tobe useful, it must form part of a wider historical analysison the evolvingrelationshipsbetween people and the forms, meansand relationshipsofproduction.Peoples Perception of VulnerabilityThe actions and decisions of people and communities show that they

34


48/113

always try to minimise vulnerability when faced with a range of hazards- an attitude which has been referred to as survival strategy.The poorerpeople become, the more their vulnerability to a variety of hazardsincreases,and it becomesmore difficult to play off one against anotherto achieve security. People have to balance extremely limited resourc-to deal with threats like homelessness, landlessness, illness,unemployment, persecution and so on. Faced with such permanentthreats people may give little priority to occasionalnatural hazards.Theless frequent and intense the hazard, the less mportance is attached oit. Floods which occur every year are more important than anearthquakewhich occurs once a century. In general,people are unlikelyto change or adapt their living patterns and activities to reduce theirvulnerability to natural hazard, f it increases heir vulnerability to othermore pressing hreats.Response to hazard is complicated because destructive naturalhazardsoften compound other kinds of vulnerability. Peoplewho haveno choice but to live in highly vulnerable settlementsand who lose theirhomes due to flood and earthquake find an already bad housingsituation made worse and may face severe difficulties in rehousingthemselves. People are often forced in to even more vulnerablelocations.The rilost important conclusion is that people themselves do notperceive vulnerability as characteristic f natural hazard or limited to thespecific time period when hazardsstrike. People often perceive naturalhazardsas perfectly ordinary characteristics f the areaswhere they live.For most people, the separationof natural disaster rom the permanentdisaster n which they live is not common sense. t is that permanentdisasterwhich explains the impact of natural hazardsand not vice versa.Reducing vulnerabilityIf the analysis of vulnerability given above is accepted, t follow thatreducing vulnerability is a complex and far-reachingprocess. Action isnecessary at different levels - individual, local, national andinternational. If the origins and causesof vulnerability lie in social andeconomic and political forces beyond peoplescontrol, then any attemptat reducing vulnerability must involve empowering people, if it is to betruly effective. In the second part of the book, different approaches oreducing vulnerability will be considered n the light of the analysis nthe first part of the book.

35


49/113

Andrew Maskrey/Juvcnal Medina/Prcdes

36


50/113

PARTTWODISASTERMITlGAmON


51/113

Chapter FiveWhat is Mitigation?Mitigation refers to measures which can be taken to minimise thedestructive and disruptive effects of hazards and thus lessen themagnitude of a disaster.Mitigation measures an be of different kinds,ranting from physical measures uch as flood defencesor safe buildingdesign to legislation, training and public awareness. Mitigation is anactivity which can take place at any time: before a disaster occurs,during an emergency, or after disaster, during recovery orreconstruction.Since the 197Os, here has been a growing interest in disastermitigation programmes rom governments, nternational relief agenciesand NGOs. There is a large body of literature which argues he case formitigation and for agenciesas well as governments o reallocateat leastpart of their budgets rom relief to mitigation. It is generally recognised,however, that mitigation still has a very low priority on their agenda.Critique of MitigationThere is, however, growing criticism of mitigation programmes.Most ofthese have been top down in approach, managed by large centralisedagencies, without any real participation in decision-making by thoseaffected by disaster. They have tended to be large-scale, hightechnology solutions, reinforcing the conditions of underdevelopmentand increasing the sense of helplessness of those affected by thedisaster. The emphasis has been on physical measures o address theimmediate threat of a natural hazard and not on social changes toaddress he problems and build up the resourcesof vulnerable groupsof people. As Ian Davis pointed out:Many programmes treat the symptom and not the cause. Thesymptom may be unsafe buildings or vulnerable cropping patterns, butthe causesmay include all or some of the following: underdevelopmentand poverty; control of land by absentee andlords; corruption; lack ofeducation; . .risk reduction policies must rely on both technicalmeasuresand on political intervention(Davis,1984).It is argued hat top down mitigation:- only deals with mitigating the risks of specific hazardsand not with

39


52/113

I reducing vulnerability. As such, it attacksonly the symptomsof disasterand not the underlying causes.- does not take into account the real needs and demands of thoseaffected by disaster. Because they ignore the complexity of mostdisasters, mitigation measures are frequently irrelevant or evencounterproductive in many local situations and programmes rarelyachieve heir goals.

I

- is political and generally favours the rich and powerful at theexpense of the most vulnerable. Risks are mitigated to avoid politicalunrest or economic loss and not to reduce the vulnerability of the poor.Some programmes actually reinforce the underlying causes ofvulnerability.An Alternative Approach to MitigationAs an alternative it is proposed that instead of dealing only with theeffects of hazards,mitigation must also address he underlying causesofvulnerability. In addition to physical measures such as reinforcingbuilding

Documents

Disaster Mitigation a Community Approach