3.3.1. Hurricane damage and reconstruction

Accomodation and services

1. Housing

Most houses in Dominica are of wood with galvanized roofs, although cement block construction is guise common in Roseau. Single-store homes are the norm, often with wooden verandahs which in towns sometimes support upstairs quarters. Before the disaster struck over half the housing stock was estimated to be at least twenty-five years old. Three-quarters of the population in our survey lived in their own homes, a fifth were in rented accomodation and a few lived in houses belonging to other members of their family.

The hurricane inflicted severe damage on people's homes. Roofing was torn off, verandahs, doors and windows were blown out and, less often, walls collapsed. (42 % of the population lost their entire roof, while 24 % had the walls of their houses destroyed).

Table 1: The Sample Population

Settlement type	inhabitants census 1970	N° of households foodlist 1979	households surveyed	N° of persons surveyed	average household size (survey results)
Main town Roseau	16,842	4,976	302	1,487	4.92
Small urban settlements	27,660	8,762	530	2,367	4.46
Villages	25,800	6,450	397	2,123	5.35
Total Dominica	70,302	20,188	1,229	5,977	4.86

The difference in average household size is significant between all settlement groups (p < 0.001).

Figure 1: Dominica: Age Distribution of Surveyed Population By 5 Years Age Groups

Figure 2: Dominica: Age Distribution of Surveyed Population by Sex

Once the roof, or part of it was off, the interior was exposed to the elements. Heavy rain, as well as gale-force winds, damaged furniture and personal possessions, as well as soaking the occupants. (Table 2). In the days that followed, inability to secure the home sometimes resulted in further deterioration and losses.

Table 2: Damage and Repair: Housing

% of total population affected

	No damage	Unknown	Damaged*	Total
Roof	10 %	2 %	88 %	100 %
Walls	39 %	5 %	56** %	100 %
Furniture	34 %	3 %	63 %	100 %
Personal	30 %	6 %	64 %	100 %
				n=5977

* includes slight and partial damage as well as total loss.
** Loss of doors, windows ant verandahs was included in the category c slight damage to walls.

In order to have a single indicator of the degree of damage to homes, houses were put in four damage categories according to the extent c' damage to their roof and walls. Damage to the roof was taken as the most significant indicator, since it was considered the most important element in shelter.

Table 3: Houses According to Damage Categories

	% of houses	% of population
1. Little or no damage to house	25.5	27.1
2. Moderate damage	26.7	26.5
3. Severe damage	17.0	17.6
4. Total destruction of house	22.0	23.2
5. Inadequate information	5.9	5.6
Total	99.1 (n=5977)	100.0 (n=5977)

Key :
1. Roof undamaged ; walls undamaged, slightly or partially damaged
2. Roof partially damaged ; walls undamaged, damaged or partially destroyed
3. Roof totally destroyed ; walls undamaged, slightly or partially damaged.
4. Roof and walls totally destroyed.

As shown in Table 4, there was some difference between settlement types as to the degree of damage sustained. The main reason for this is geographical. Roseau, the capital, was in the south of the island, which was most severely hit by the hurricane. In contrast, 5/11 of the urban units sampled were from Portsmouth, the second largest town, situated in the north-west of the island, which largely escaped the worst of the hurricane damage.

Table 4: Damage to Houses According to Type of Settlement

Settlement type	% of population with houses in categories				Total
	1. Undamaged	2.3. Damaged	4. Destroyed	Unknown
Roseau	15.8	47.6	30.6	6.0	100.0
Towns	32.7	45.7	15.4	6.2	100.0
Villages	28.6	39.9	26.7	4.8	100.0
Total (n=5977)	27.1	44.1	23.2	5.6	100.0

Table 5: Housing: Repair After 9 Months

	Repair complete	Temporary repair	Not yet repaired	Total
Roof	26 %	59 %	15 %	100 %
Walls	19 %	45 %	36 %	100 %

Table 5 chows the slow rate at which homes were being restored to normal. After the disaster, repairs to roofs were a priority and over half were at least patched up in the month following the disaster. Nine months later. however, only a third of the population was living in houses with sound roofs and half were in houses where roof repairs were still only of a temporary nature. Similarly, most households had not completely repaired or replaced their furniture by Hay, 1980. A fifth of those who had lost personal effects replaced them in the month following the hurricane but over half had not done so nine months later. With such a high proportion of temporary repairs, there was a risk that many would become semi-permanent. While a few people remained optimistic that the devastation brought by the hurricane would give them a chance to build a better home, it seemed more likely that the general result would be a long-term deterioration of the housing stock.

Underlying reasons for the delay in restoring homes included the generally low levels of savings and the fact that the vast majority has no household insurance. (It was asserted that insurance, would, at a maximum, cover 15 % of the damage to homes). Immediate causes included lack of money, resulting from loss of income in the postdisaster period, scarcity of building materials and, while there was ample informal assistance and self-help, shortage of skilled tradesmen.

2. Utilities

Table 6 shows the proportion of the sample population who had some domestic utilities - water, electricity and toilets - in their homes. It shows how these were damaged in the hurricane and the extent to which this damage had been repaired nine months after the hurricane.

a. Electricity

Before the hurricane, well over half the population sample had electricity. Of the houses that had none, 41 % were in the villages, 35 % in the towns and 24 % in Roseau. The hurricane played havoc with the distribution lines (which were uninsured) and damaged power stations and equipment. Virtually the whole island was without electricity.

Nine months later, less than a quarter of those who had previously had electricity hat their supplies restored. These were mainly people living in one area of Roseau ant about half hat beer. reconnected in December.

b. Private water supplies

Four-fifths of the population of Dominica hat access to piped water supplies, but half used public standpoints. 34 % of our sample had private water supplies in their homes. Of the houses without private supplies, 51 % were in the villages, 38 % in towns and only 11 % were in Roseau.

Table 6: Damage and Repair: Utilities

Utility	Population having utility		Hurricane damage			Repair after 9 months
	No.	% of total	% of population with utility			% of population with damaged utility
			Undamaged	Damaged	total	Repaired	Not yet repaired	Total
Electricity	3521	58.9	0.9	99.1	100.0	23.2	76.8	100.0
Private water supply*	2050	34.3	17.8	82.2	100.0	88.7	11.3	100.0
Private W.C.	1020	17.1	44.6	55.4	100.0	79.5	20.5	100.0
Private pit toilet	2628	44.0	17.2	82.8	100.0	60.8	39.2	100.0
Public toilet	701	11.8	16.8	83.2	100.0	12.5	87.5	100.0

* Private piped supply in house or yard in a low cases, private cisterns.

The hurricane caused considerable disruption to water supplies i pipelines were washed out, inlets blocked ant reservoir roofs and some of the few chlorination plants damaged. 88.7 2 of the sample who had private water supplies found then cut off. However, because of the potential risk to health from contaminated water, the authorities acted swiftly in the wake of the disaster.

Thus, 41.7 % of homes with damaged supplies were reconnected within a month and only 11.3 % had not had their water restored nine months later.

c. Toilet facilities

Just over half the sampled population had private toilets, a relatively small percentage used public toilets and over a quarter had no formal facilities.

Nearly half the flush toilets were undamaged following the disaster. This surprisingly high figure is possibly due to the fact that such toilets were to be found in well-constructed homes that withstood the force of the hurricane. Many were useable again as soon as water supplies were restored. The high proportion of pit latrines damaged was unexpected. But in many cases it was the building around the pit rather than the pit itself which was demolished. Four-fifths of those using public toilets found that their facilities were damaged or destroyed. Unlike toilets in private hones, the vast majority had not been repaired or replaced nine months later.

3. The effect on employment

The occupational structure of the population of Dominica (shown in Table 7) is typical of a young developing country. There is a fair degree of overlap between occupations, particularly those involving the self-employed. (For example, a woman may class herself as a housewife when she is also growing much of the family's food in her garden). Moreover, much of the work available is casual or seasonal and 10 % of people with an employment had secondary Jobs, which were often part-time. The agricultural sector is important. If secondary Jobs are included, 42.2 % of those employed were farmers or workers on the land. Those in the very small industrial sector were almost al; engaged in agricultural processing or packing.

Table 7: Jobs Lost and Gained After the Hurricane

a) Population according to main occupation

	People
	No	%
Employed	1923	32.2
Schoolchildren, students, apprentices	1889	31.6
Housewives	533	8.9
Unemployed	505	8.4
No occupation*	954	16.0
Unknown	173	2.9
TOTAL	5977	100.0

*Includes those under-school age, the elderly, infirm and retired etc.

b) Employed population according to main job

	People
	No	%
Hand labourers	350	18.2
Farmers	336	17.5
Domestics	187	9.7
Building tradesmen	130	6.8
Teachers	79	4.1
Fishermen	54	2.8
Others	787	40.9
TOTAL	1923	100.0

Table 8: Occupation and Employment

	No of people	% of employed
Main job lost	356	18.5
Secondary Job lost	78	4.1
Main Job gained	117	6.1
Secondary job gained	9	0.5

After the hurricane (Table 8), over a fifth lost their Job (18.5 % of main jobs were lost). About half of those who lost Jobs were land labourers. The main reason they gave for losing work was the destruction of crops on the estates. For the others, the main reason for losing their jobs was damage to the place of work and to equipment. The local fishery activities were badly affected as 45 out of 69 boats were totally destroyed and only 20 were undamaged. However, at least some new jobs were created in the wake of the hurricane. Initially there was extra work to be done in clearing agricultural land of fallen trees and in repairing roads. Building tradesmen were in considerable demand.

4. Agriculture

a. Crops

No questions were asked about the extent of crop production. Households were only asked which crops they grew. No distinction can therefore be made between the small-scale cultivator and the large plantation owner. Replies to a question about land ownership (not analyzed here) indicated that almost all those sampled had holdings of a few hectares. (This la in agreement with an official estimate that 75 % of owners have less than 3 hectares, while 1.4 % own over half the land in large estates).

Nearly half the households grew bananas, the main export and a steady year-round source of income to the farmer. Over A third had coconuts and nearly a third had ground provisions, the traditional staple food. Cocoa (together with coffee) and citrus were next in importance (20 % of all families).

Households reported very heavy crop losses caused by the hurricane (Table 9). 90 % of those growing bananas lost all their plants. 72 % of those growing cocoa or coffee and 65 % of those growing coconuts lost all their trees (not merely the current harvest). Only few cultivators had crops which suffered little or no damage. The high proportion of growers who reported total loss of ground provision was unexpected, since public authority reports, written in the first few months after the disaster, had estimated a reduction in production of 50 % or less.

After the disaster, householders were quick to restore their banana plants. Bananas are rehabilitated relatively easily and official aid (fertilizers, pesticides, etc.) was concentrated on this key crop. Other cash crops were replaced much more slowly, although a third of those who lost coconut palms had reestablished them nine months later. In the case of other tree crops there was little replanting ; simple vegetative reproduction was impossible and it was often unclear whether trees which had not been uprooted had any prospect of recovery. As in the case of bananas, about three-quarters of those with ground provisions had rehabilitated their plants nine months later.

b. Domestic animals

Households were asked what types of domestic livestock they owned. Ho distinction could be made between, for example, households with a few hens for family use and those who used poultry to provide a subsidiary income.

Over 40 % of households had poultry, over a fifth kept goats and nearly as many had pigs. Cows, then rabbits, were next in importance. Although over half the households owning livestock were in the countryside, around a third were in the towns (indeed these smaller urban settlements have kept many rural characteristics). Livestock ownership was not unusual even in Roseau; notably, 14 % of poultry-owning households were in the capital. In general households owning animals tended to be larger than average.

Table 9: Crops, Losses and Replacement

Type of crop	Proportion growing crop		% of crop growers with			9 months later. Percentage of those with crop losses who had not replaced crop
	% of households	% of population	Little or no damage	Partial losses*	Total loss
Bananas	47.5	52.1	2.6	7.4	90.0	25.9
Coconuts	37.2	40.6	5.5	29.5	65.0	66.7
Ground provisions	29.4	33.2	8.3	23.0	68.7	27.6
Cocoa or coffee	19.1	21.8	5.9	22.3	71.8	80.7
Citrus	17.8	19.6	10.9	24.1	65.0	73.1
Other fruit trees(mangoes etc)	11.6	12.2	12.2	21.1	66.7	76.2
Other crops	10.9	13.5	8.1	25.9	66.0	80.6

*Particularly in the case of ground provisions, there may have been some confusion over how to record loss of yield as distinct from the destruction of plants. This may have led to some exaggeration of losses.

Animal losses in the hurricane varied greatly between species (Table 10). Fewer than 15 % of cattle owners lost part or all of the stock. This was important since cattle represent considerable capital investment and, as was shown, are cloy to be replaced. A third of pig-owners lost animals. The figure rose to around a half where goat and poultry ownership was concerned and was highest of all for those with rabbits. Not all losses were caused by death. Animals were also blown away or strayed and some of these were not recovered by their owners. Nine months later only a small percentage of owners had replaced all their lost stock. Around 60 % of poultry and pig owners still had few animals and around three quarters of those owning rabbits and goats were in the same position.

Table 10: Domestic Animals, Losses and Replacement

Type of animals	proportion of people with animals		No losses	Partial or total losses	9 months after proportion who hurricane: replaced had not stock
	% of house-holds	% of population
Cattle	10.3	13.0	85.8 %	14.2 %	99.2 %
Pig	18.0	23.7	67.0 %	33.0 %	59.9 %
Goat	21.2	26.5	52.2 %	47.8 %	73.1 %
Rabbits	5.3	8.3	18.5 %	81.5 %	75.5 %

In Dominica, domestic livestock, particularly poultry, pigs and goats, are a useful source of animal protein. Failure to replace stocks must be of concern to nutritionists in an area where starchy roots and tubers, low in protein, are still important as staples. Massive provisions of food aid were brought into the island after the hurricane.

4. The effect of Hurricane David on food supplies

The cyclone resulted in an immediate reduction in the amount of food available locally. Food crops were damaged, mainly through exfoliation. Consequently, green vegetables were lose though ground provisions (root crops) recovered toward the end of the year. Livestock, particularly poultry, were killed and hens moulted and temporarily ceased to lay. Food stores whether in warehouses or in homes were spoilt by rainwater and several commercial premises were looted. Additionally, broken bridges and cut roads disrupted the distribution system.

Not surprisingly four-fifths of the population sampled reported that they were short of food in the week following the disaster. Admittedly, in the first few days after the hurricane, there were often ample supplies of individual foodstuffs, such as fallen fruit or damp grain, which would be wasted if not immediately consumed, but housewives, often in a state of shock, their kitchens damaged, could not make beat use of these gluts.

Shortly after the disaster, massive food aid from abroad began to arrive and continued on a large scale for four to six months. It then dwindled but in some areas ceased only a few weeks before the survey began. During the fires four months the average daily ration was 1565 calories and 80 grams of protein. Since some local foodstuffs were still available, the quantity was more than enough. It is, however, a credit to the distribution system that virtually the whole population sampled (99,2 %) reported having ample food in the months following the disaster.