• 검색 결과가 없습니다.

Quantitative analysis

문서에서 Impact of (페이지 58-61)

combined models

3.1 Quantitative analysis

3.1.1 Floods

Individual event analysis

The two major floods which took place during the period the NSP was implemented (1998 and 2004) were tested individually using the DiD protocol. The nutrition and food security indicators were computed before and after the events for communities that had been affected by the flood events, and the changes in these indicators were compared to the changes in the same indicators for control communities which had not been affected by the flood12. The analysis shows the following results:

Findings

12 There were two methods through which these comparisons were done due to data limitations. For the 1998 flood, as only a few rounds of data collection took place before event onset, this period was selected as the before period compared to the times after the event without controlling for seasonal variation. This method assumes that differences existing in the early part of 1998 were consistent with that experienced in other seasons. For the 2004 flood, since a full year of data collection was available before the event, the same season in the year prior to event onset was used as the before period enabling us to control for seasonal variation. This method implies that the seasonal variation experienced the year before onset was typical. This is a stronger assumption so whenever possible the second type of analysis will be presented though both methods were undertaken and largely consistent.

13 Maternal BMI was not included in the NSP system until 2000. Dietary diversity was not included until 1999.

14 It should be noted that a smaller sample of households were obtained from Shadjadpur (74 out of a usual 350) and Serajdikhan (180 out of a usual 350) during the first lagged period in 1998 (sampling round 51). These two areas were among the most severely affected by this flood. This it likely related to accessibility of this area at this time, and as such the estimates above may be an underestimate of the severity of the problem.

Nutrition indicators

For the 1998 flood, only the child’s zwfl and zlen could be used for the test as the mother BMI, the dd and DD had not been measured13. For the 2004 event, all five nutrition indicators (zwfl, zlen, BMI, dd and DD) had been measured and could be used for the test. Results for both events are shown in Table 9.

For the 1998 flood event, the DiD test shows that the zwfl (wasting) is negatively affected by the 1998 flood event: all zwfl indicators show negative coefficients for the period directly following the event and the five lagged periods (i.e. for three, five, seven, and nine months following the flood). This suggests that the flood had a negative impact on children’s

nutritional status. The difference is statistically significant (p=0.004;

p=0.006; p<0.0001) for three, five, and seven-month lagged periods.

The coefficient for zlen (stunting) was also negative, but the difference is only statistically significant (p=0.035) for the seven-month lagged period14. The 2004 event shows some degrees of consistency with this pattern. The coefficient for zwfl displays negative sign for zero, three, five, seven and nine-month lags, although none are statistically significant. The coefficient for zlen is also consistently negative for the three, five, and seven month lagged periods, but here again none of them are statistically significant.

Table 9. 1998 and 2004 individual flood event models(a) – nutrition indicators

zlen zwfl BMI child dd Maternal DD

coef p-value coef p-value coef p-value coef p-value coef p-value event

+3m +5m +7m +9m event +3m +5m +7m +9m

-0.044 -0.111 -0.032 -0.119 -0.023 0.033 -0.039 -0.014 -0.029 0.001

0.009 -0.136 -0.125 -0.164 -0.079 -0.044 -0.061 -0.037 -0.058 -0.049

-0.190 0.107 0.050 -0.131 0.093

-0.212 -0.269 0.142 -0.776 0.326

-0.327 -0.663 -0.587 -0.744 0.240 0.035

0.004 0.006 0.000

0.012

0.013

0.010

Legend: Only statistically significant p-values are indicated (in bold), other non-significant p-values are omitted. Coefficient values highlighted in light red indicate cases for which the DiD test suggests a worsening situation in relation to nutrition, that is, a lower zlen, zwfl, BMI, child dd or maternal DD in the affected (treatment) communities, compared to the non-affected (control) communities. Dark red values indicate cases where the worsening of the nutritional situation is statistically significant.

zlen: Length/Height for age z-score; zwfl: weight for Length/Height z-score; BMI: Maternal body mass index; dd: Child dietary diversity; DD: Maternal dietary diversity. The notation +3m; +5m, +7m, and +9m indicate 3-month, 5-month, 7-month and 9-month lagged periods respectively, while

‘event’ refers to the period during or just after the event.

(a) The following variables were also included in the models (but not shown in the table) to control for individual and household effects: age and sex of child; mother age and education; livelihood strategies (farmer; labour; transport; salary; business); family size; birth order; source for drinking; use of latrine –see definitions in section 2.5.2.

19982004

Not collected in 1998

For the 2004 event the three other nutrition indicators (BMI, dd, DD) were also monitored. The DiD test show that the mothers’ BMI does not show any statistically significant change. The child dd shows negative correlations

immediately after the event and for three and seven-month lagged periods, with the seven-month lagged period being statistically significant (p=0.012). This pattern is even more severe for the mother DD: negative

correlation for the period straight after the flood and for the three, five and seven-month lagged periods.

The three and seven-months lagged periods are statistically significant, with p=0.013 and p=0.01 respectively.

Food security and food price indicators

For both the 1998 and 2004 flood events, the five food security and food price indicators (FL, FE, PR, PO, PFB) were recorded and could be used for DiD15. Results for both events are shown in Table 10. No clear patterns emerge for most of these indicators. For FL (food loan) for instance, while the period following the 1998 event and the following five months were negative (with the three-month lag being statistically significant (p=0.024)), the same indicator over the same periods was positive following the 2004

flood, where the three month and five month periods were statistically highly significant (p=0.014 and p<0.001). Another example of lack of consistency in these indicators is the FE indicator (Food expenditure). While in 1998, the indicator was positive for the seven months following the flood event with the first three periods (i.e. for t=0, 3, 5) highly statistically significant (p=0.037; p<0.001, and p=0.018 respectively), suggesting a significant increase in household expenditures on food (usually a sign of degrading food security condition);

the same indicator in 2004 was positive for the three month period

and then negative for up to nine months after the flood event16, with statistically significant negative values for the five month and nine-month lags (p=0.045 and p=0.001).

The only indicator which seems to show a consistent trend is the PFB (price of food basket). Both in 1998 and 2004 this indicator was consistently positive over the nine-month period following the flood events, with statistically significant values for the seven-month lagged period in 1998 and 2004 and for the 9-month lagged period in 1998 (p<0.001; p=0.001, and p=0.04).

15 It should be noted that the price indicators (PR, PO, & PFB), are missing from Shahjadpur, the most severely affected upazila in the 1998 flood for the first lagged period after the flood. This is in addition to the smaller sample of households obtained (see footnote 13). As such the estimates above are likely an underestimation of the severity of the impact.

16 As the 2004 flood was shorter, one could hypothesize that household have moved cash away from just food, to other budget lines, such as replanting/etc. –thus explaining the progressive decrease in the FE coefficient observed after 2 months.

Table 10. 1994 and 2004 individual flood event models(a) – food security and food price indicators

Table 11. Flood event joint models(a) – nutrition indicators FL

zlen all

FE

zlen 10 days

PR

zlen 15 days

PO

zlen 20 days

PFB

zlen 30 days zlen 30 days

문서에서 Impact of (페이지 58-61)