Page 58 - AJWEP-22-4

Basu

Table 6. The impact of Cyclone Bulbul on rice production (estimated using panel fixed effects regression
model)

Variables Model 1 (Khulna) Model 2 (Satkhira) Model 3 (Combined)
Land (acres) −0.0125 (0.0383) 0.0468 (0.0423) −0.0937*** (0.0301)
Labor (man-days) 0.179 (0.142) −0.0431 (0.0487) 0.125** (0.0623)
Seed (kg) 0.0159 (0.0495) 0.0298 (0.0513) −0.00214 (0.0355)
Chemical fertilizer (kg) −0.188** (0.0785) −0.0565 (0.0427) −0.0107 (0.0314)
Pesticide (kg) 0.0305 (0.0393) 0.0862*** (0.0330) 0.0767*** (0.0272)
D during −0.491*** (0.0254) −0.252*** (0.0302) −0.429*** (0.0211)
D after 0.0910*** (0.0290) −0.0623*** (0.0163) −0.0323** (0.0151)
Constant 4.574*** (0.753) 3.783*** (0.262) 3.687*** (0.250)
Number of farmers 200 200 400
R (within) 0.660 0.555 0.570
2
R (between) 0.001 0.176 0.093
2
Number of observations 600 600 1200
Notes: Regression coefficients are expressed in logarithmic form. Robust standard errors clustered at the farmer level are reported in
parentheses. The dependent variable is rice output measured in maund (1 maund=40 kg). D during is the dummy variable=1 during the
cyclone; D is the dummy variable=1 after the cyclone. *p<0.10, **p<0.05, ***p<0.01.
after
insignificant. In contrast, the combined model for increases vulnerability to natural disasters. Graphical
both districts found that land had a negative effect, representations of the regression results are provided in
while labor had a positive influence on rice output. Figure A6. To account for geographical variation, the
The coefficient for the Dduring dummy variable was same models were estimated separately for Khulna and
negative and highly significant across all models, Satkhira. Results are reported in Tables A10 and A11,
indicating that cyclone-affected farmers experienced which indicate that farmers in Satkhira experienced
substantial production losses during the cyclone season: greater relative financial loss compared to those in
49% in Khulna (p<0.001), 25% in Satkhira (p<0.001), Khulna.
and 43% overall (p<0.001) across both regions. Post-
cyclone effects varied: rice production increased slightly 5. Discussion
in Khulna but declined marginally in Satkhira. The
coefficients of Models 1, 2, and 3 are presented visually This study examined the impact of Cyclones Amphan
in Figure A5. The standardized regression coefficients and Bulbul on rice production in the coastal areas of
are provided in Table A9, further confirming the adverse Khulna and Satkhira, Bangladesh. Cyclone Amphan
effects of Cyclone Bulbul during the event. led to a significant reduction in rice yield, with an
average production loss of 38% in Khulna and 26% in
4.5. Determinants of Cyclone Bulbul‑induced Satkhira. Cyclone Bulbul, though slightly less intense,
relative financial loss similarly diminished productivity by approximately
The determinants of the relative financial loss of 45% in Khulna and 38% in Satkhira. These findings are
cyclone Bulbul-affected farmers were estimated using consistent with prior research indicating that climate-
Equation V, with results shown in Table 7. Among induced disasters disproportionately affect agricultural
several sociodemographic variables, the primary outputs in vulnerable coastal regions, where soil
occupation of the farmer emerged as a critical factor salinity, waterlogging, and direct crop damage intensify
influencing financial loss across all farming activities— the effects of climatic shocks. 5,11
rice, fish, vegetables, other crops, and overall farming. The application of the Cobb-Douglas production
Specifically, farmers who relied exclusively on function model and fixed effects panel regression
agriculture as their main occupation suffered significantly underscores the multidimensionality of cyclone impacts
higher relative losses in all farming activities except on agriculture. Results show that key agricultural
fish. These findings highlight a crucial aspect of cyclone inputs, including land, labor, and chemical fertilizers,
risk management: Monocentric livelihood dependence contributed positively to productivity. However, their

Volume 22 Issue 4 (2025) 50 doi: 10.36922/AJWEP025100063