当前位置:首页 >> 农林牧渔 >>

Flood


Flood, Productivity and Wage Rate in Agriculture: The Case of Bangladesh

Lopamudra Banerjee? First Draft: March 2009

Introduction

A recurrent theme in Azizur

Rahman Khan’s work has been the role of ownership of productive resources, especially land, and access to employment, towards poverty alleviation and equity. Entitlement to productive resources and secured employment ensure a stream of income such that a particular level of consumption is maintained not only in the ‘normal’ times, but also in the times of ‘distress’, when the regular economic functionings are disrupted. In the absence of institutional arrangements that ensure the right to asset and the right to job, this flow of income, however, may be disturbed, more so in extreme situations when covariate shocks like disasters occur. In this article I look at the agrarian labor market in Bangladesh to examine how agricultural productivity and the stream of wage income for the agricultural workers are affected in the years of flood disasters. I shall analyze my results in the light of an early paper by Khan (1984), where he explains agricultural wage formation in the country in terms of productivity, trend factors in the labor market, and terms-of-trade in agriculture. I use my results to illustrate how the contested claim over returns in agriculture between the landowners and the workers may become all the more apparent in the times of disasters. The motivation here is to observe what happens in the extreme situations in the hope to get important insights about the customary functioning of social institutions, including the conventional employment and income sharing practices in rural Bangladesh, in the ‘normal’ times. I interpret my results in terms of policy implications not only to reduce flood hazards for the rural poor, but also for long-term agrarian development of the country.

Productivity and wage rate in agriculture

It is argued in the literature that, in a labor surplus economy, the average labor productivity determine agricultural wage rate (Ranis, 2004). In other words, in the situations where man/land ratios are high and marginal productivity of workers is relatively low, wage determination is based on a principle of sharing output and/or income. This argument is

? Department of Economics, The New School for Social Research, 6 East 16th Street, Room 1125, New York, NY 10003, USA; T: +1. 212. 229. 5717 (x 3402); F: +1. 212. 229. 5724; E: banerjel@newschool.edu

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

1

invoked to explain wage formation in the agrarian labor market in South Asia, including Bangladesh (Bardhan, 1984). Average productivity of labor in this region, in turn, has been closely linked to the average productivity of land under the conventional assumption of a fixed-input ratio per unit of output or a Leontief type production function (Khan, 1984). Based on the above theoretical specifications, we can then claim that, if there were any increase in the average productivity of land, there would be an implicit increase in the average productivity of labor, and a corresponding increase in the real agricultural wage rate. Thus, any improvement in the yield rate should be reflected in terms of an improvement in the factor’s share of income, including the worker’s share. In the case of Bangladesh, this result was empirically found to be true in Khan (1984), where an analysis of annual average agricultural wage formation in the country (over 1954-1983) revealed that, in the long run, a rise in agricultural productivity plays a significant role in causing a rise in real wages by stimulating the demand for labor. This result may, however, not hold in the short run, especially when covariate shocks like flood disasters are realized in the labor market. On analyzing the crop productivity data and the real agricultural wage data for the districts in Bangladesh I find that, in the ‘flooddisaster’ years, an improvement in average annual agricultural productivity is actually accompanied by a decline in the average wage rate. I also find that while agricultural productivity in Bangladesh has shown a positive trend, agricultural wage rate has been declining over the long-run. Even before we delve into how this result is obtained, the immediate question that comes to mind is what explains this result. The increase in average annual yield rate observed in the flood years may be explained in the following manner: Freshwater riverine floods act as an open-access resource in supplying irrigational input to agriculture in the country (Paul and Rasid, 1993). The floods, therefore, can capture the effect of a land-augmenting positive technological change, especially in the immediate post-flood crop season. A loss of crops in the flood months is more than compensated by a bumper production of post-flood crops (Banerjee, 2008; Brammer, 1990). This gain in productivity, however, does not reach the agricultural workers. A decline in average wage rates that is observed in the flood months continues in the post-flood seasons. This decline may be explained in terms of the ‘distress sale’ of labor triggered by the disaster. In a competitive framework, this phenomenon can be described in the following manner: A loss of income and decline in consumption in the flood-months may induce the workers to work for more hours once the water recedes from the fields (del Ninno and Roy, 2001). The effects of this increased labor supply on wages can outweigh the effect of the increased labor demand, such that average wages decline even when productivities increase. An alternative explanation for the decline in wage rate may also be in terms of weakened bargaining position of the workers vis-à-vis the landlords after the disaster. Protracted unemployment in the flood months may

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

2

increase the ‘haggling costs’ for the workers, either because they are now more anxious to find employment, or because flood have destroyed their assets, thus undermining their ‘security point’ (Kreps, 1990; Muthoo, 2000). In either case, in the absence of adequate social transfer mechanism, the positive benefits of post-flood increase in productivity will not be realized for the agricultural workers (Montgomery, 1985). The anomalies in fluctuations in productivity and wage rates that are observed in the short-run, during the periods of disaster shocks, however are also present in the long-run. As I shall show later in this article, over the period 1978-2000, while productivity in agriculture shows a positive linear trend, real wage rates in agriculture (money wages deflated by the rural cost of living index) shows a negative linear trend. Trend factors, as determinant of wages capture the long-term determinants in the labor market, including demographic factors and the very-slow growth rate in employment in non-agricultural sectors that determine the supply of labor. The trend factors also captured the slow expansion of land-frontier and the low cropping intensity in Bangladesh, and thus indicating a slow rate of increase in demand for labor in agriculture. I now present the main analysis of this article.

Data

In this essay, a period of ‘flood disaster’ is defined as the month-year when 20% or more of the total area of Bangladesh was inundated with a flood depth of 180 cm or more (‘deep’ flood) (Bangladesh National Water Plan, 1986). Over 1979-2000, such ‘flood disasters’ have occurred in the monsoon months (June to September) of 1980, 1984, 1987, 1988, and 1998 (Banerjee, 2008).1 I introduce a dummy variable F to indicate these periods of disasters, F takes the value of 1 for June-September in 1980, 1984, 1987, 1988, and 1998, and is zero otherwise.2 I measure agricultural productivity in terms of the yield rate per acre of cropped land for rice (the principle food crop) and jute (the principle cash crop jute).3 The series on output produced per unit area under cultivation (in acres) in each district, by month over 1979–2000, is generated for each variety of rice (aus, aman, and boro, in metric tons) and each variety of

1. The data on relative severity of ?ood conditions in the country is brought together from multiple sources. These sources are: Ahmad et al. (2001), Asian Development Bank (ADB) (2004), Bangladesh Poribesh Andolon (BAPA) (2000), Chowdhury et al. (2006), del Ninno and Roy (1999a, 1999b), Hossain et al. (1988), Islam (1999), Khalequzzaman (1994), Shahjahan (1989), United Nations Department of Humanitarian A? airs (DHA) (various years), United Nations Office for Coordination of Humanitarian Affairs (OCHA) (2000). 2. Monsoon floods have also occurred in other years in the period 1978-2000, but these floods were locally concentrated, and are generally considered as ‘normal’ phenomenon for deltaic Bangladesh. 3. In 2000, rice and jute were cultivated in respectively 76 percent and 2.86 percent of the total cultivable area in Bangladesh (BBS, 2002).

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

3

jute (white and tossa, in bales).4 I take a weighted average of the yield rate of the di? erent varieties of rice and generate a continuous series on monthly productivity of rice (all types, in metric tons per acre) for January 1979–December 2000. Similarly, I generate the series on weighted average of monthly productivity of jute (all types, in metric tons per acre). The relevant weights are determined using the crop calendar for the country (BBS, 2002). The data on real wage rate for agricultural workers, by districts, by month, is generated for the period 1979-2000, by deflating the nominal wage series for male agricultural workers (without food by the rural CPI).5 The least-square-with-dummy-variables (LSDV) method is used to pool together the temporal and cross-sectional information (Maddala, 1977). The continuous series on real wage is thus generated across the twenty districts in the country, for the successive month–years January 1979 to December 2000. The series has 3,675 observation points.

Trend, seasonality and the effects of flood shocks on agricultural productivity

The trend and the seasonality in the productivity for rice and jute is estimated in terms of

Qd,t = α + τ 1t + σ 1S1 + σ 2 S2 + ε d,t [1]
where d = 1,2, ..,20 for the twenty districts, and t = 1 for January 1979, 2 for February 1979,.. for the successive month-years over January 1979–December 2000; acre yield rate of the crop,

Qd ,t

is the per

t

is the linear trend.

S1

and

S2

are respectively the dummy the value of 1 for the months

variables indicating summer and winter crop seasons. March-June and is zero otherwise,
6

S1 takes

S2

takes the value of 1 for the months November-February

and is zero otherwise. The effect of seasonality is examined only for rice, and not for jute, which is an annual crop. Table 1 presents the estimation results for equation [1]. Column 2 in the table reports the coefficient estimates for rice productivity, while column 3 reports that for jute productivity.

Source: BBS, various years. Source: BBS, various years. I use the rural CPI data as a proxy for the CPI for agricultural workers, as the latter data series is not available. The series on CPI is available from July 1978, and has missing data for December 1987 to October 1988. Also, while the nominal wage series is available for the districts, the CPI series is available for the four divisions in the country: Dhaka, Chittagong, Khulna, and Rajshahi. 6. In terms of the nature of irrigation, crop calendar in Bangladesh consists of two somewhat overlapping seasons: (a) Wet (monsoon) season or kharif crop season (mid March-early January) and (b) dry-season or rabi crop season (mid-November-August). Aman variety of rice is the principal wet-season (kharif) crop. Jute, the main cash crop (an annual crop), is also harvested in this season. The dry (rabi) season consists of (i) winter (mid November-May) and (ii) summer (mid March-August). Boro variety of rice is the main winter crop, while aus variety of rice is the main summer crop. The common practice for Bangladeshi peasants is to cultivate aman rice in monsoon, followed by boro or aus in dry season (Hossain, 1990; Datta, 1998; BBS [a], various years).
5.

4.

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

4

INSERT TABLE 1 HERE

To examine how productivity is affected in the years of flood, equation [1] is extended to include the dummy variable

F

indicating ‘flood disasters’.

Qd ,t = α 0 + τ 1t + β1S1 + β 2 S 2 + δFd ,t + θ (S 2 * Fd ,t ) + u d ,t [2]
In the ‘extended’ model, while

δ captures the impact of floods on the yield rate in the monsoon,

θ

captures their impact on the yield rate of post-flood winter crops in the same year.

Estimates of equation [2] are reported in columns 4 (for rice) and 5 (for jute) in Table 1. From the Table 1 we find the followings: Over the period 1978-2000, agricultural productivity in Bangladesh showed a positive linear trend. The seasonal yield rate of rice in the winter (the season for HYV boro crop) was higher than that in the monsoon (the season for aman crop), but lower in the summer (the season for aus crop). In the years of disasters, rice yield sharply declined in the monsoon flood-season but significantly increased in the post-flood winter season. The estimated average shortfall of rice productivity in monsoon in these years is almost 70% below the trend. 7 Annual yield rate of jute also increased in the disaster years, though the increase was not statistically significant. The estimated coefficients that are of special interest to us are ( ?

?α ? δ ? ?0 ? ? *100 ) and ? ? α0 ?

(?

? β ?θ ? ? ? ?α ? δ ? ? ? 2 *100 ). ( ? 0 ? *100 ) captures the percentage deviation in the monsoon yield rate ? ? ? ? β2 ? ? α0 ? ? β ?θ ? ? ? 2 *100 ) captures the percentage deviation in the winter yield rate in ? ? ? β ?
2

estimated in the flood years from that in the non-flood years (i.e., the ‘normal’ yield rate in the season), while ( ?

? ? ? ? the flood year from the ‘normal’ yield rate in the season. α 0 , δ , β 2 and θ are respectively the
coefficient estimates for the fixed coefficient of regression ( α 0 ), the ‘flood disaster’ dummy (F), the winter dummy (S2), and the ‘cross-effect’ dummy (S2*Fd,t) that capture the lingering effect of flood on the immediate post-flood winter crops. From Table 1, we find that the yield rate in the monsoon season in flood years fell 50% below the ‘normal’ non-flood estimated monsoon yield. In contrast, productivity increased by an estimated 146% above the ‘normal’ seasonal yield rate in the post-flood winter season.

7.

?τ ? δ ? ? ? ? ? *100 , ? ? τ ?

The percentage deviation in the yield rate in flood years from the trend is calculated as where

? ? τ 1 and δ are

respectively the coefficient estimates for the trend (t) and the ‘flood

disaster’ dummy (Fd,t).

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

5

The fluctuations in yield rate observed in disaster years point to the differing effects of flood on crops produced in the different seasons. While productivity in the inundated fields decline in the monsoon flood months, productivity of the post-flood crops increase considerably enough to result in an ‘above-normal’ annual output in extreme flood years (Banerjee, 2008). The moisture retained in the soil from inundation increase post-flood yield rate (Paul and Rasid, 1993; Brammer, 1988). Evidently, the peasants/landlords who have concentrated more on aman crops are adversely affected by the disasters, while those who can invest on winter HYV boro crops actually benefitted from the floods.

Trend, seasonality and the effects of flood shocks on agricultural wages

To examining the trend and the seasonality in the real wage data, I start by estimating the following equation:

w d ,t = α + τ '1 t + σ '1 S1 + σ ' 2 S 2 + ε ' d ,t [3]
where

wd ,t is

the current real wage for male agricultural workers (without food) (in terms of

rural CPI) in district d in month-year t (d = 1,2, ..,20 for the twenty districts, and t = 1 for January 1979, 2 for February 1979,.. for the successive month-years over January 1979– December 2000). All other symbols have their earlier interpretations. Column 2 in Table 2 presents the estimation results for this model.

INSERT TABLE 2 HERE

To examine the effect of flood disasters in the years 1980, 1984, 1987, 1988, and 1998 on the real wages, I extend equation [3] to build up the following model:

W = α' 0 +τ '1 t + β '1 S1 + β ' 2 S2 + δ ' F + θ ' (S2 * F) + u′ [4]
Estimates of equation [4] are reported in column 3 of Table 2. The percentage deviation in the wage rate in the ‘flood disaster’ year from the ‘normal’ trend is captured by (?

? (τ ' ?δ ' ) ? ?1 ? ? *100) . ?′ ? τ1 ?

The percentage deviation in the monsoon wage rate explained by ‘flood disaster’ is captured by (?

? (α ' ?δ ' ) ? ?0 ? ? *100) , while percentage deviation in the post-flood winter wage rate in ‘flood ? ? α′0 ?
? (β ' ?δ ' ) ? ? ? 2 ? *100) . ? ? β′ ?
2

disaster’ year is captured by (?

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

6

From Table 2 we find the following: In the ‘normal’ years, the seasonal fluctuations in the wages, however, closely match the fluctuations in productivity. Compared to the monsoon crop season, wage rates, on an average, are higher in the winter and lower in the summer. In the disaster years, however, wage rates in the monsoon season decline 76% below the ‘normal’ annual trend, and 34% below the estimated ‘normal’ monsoon average wage rate. In these years, wage rate continue to remain 9% below the estimated ‘normal’ rate in the winter crop season, even though productivity of crops rise in the season. This paper argues that the impact of flood on wages is realized through the impact of flood on demand and supply conditions in agricultural labor market. As floods destroy the monsoon crops, the demand for the workers decline (del Ninno and Roy, 2001a, 2001b). Given the supply of labor, this reduces the real wage. In the post flood season, wages are depressed further with the increase in ‘distress sale’ of labor. These negative effects can continue in the post-flood months through increased household debt (del Ninno and Roy, 2001b). With the destruction of assets (homestead, cattle and so forth) in the flood season, and the consequent reduction in the already meager wealth of the workers, post-flood supply of labor may increase (Islam, 2001). In their effort to deal with the income loss in the monsoon, the workers may be willing to accept lower wages to find employment in the winter season. Table 2 also shows that, in contrast to the trends observed in agricultural productivity, real wage rates in Bangladesh show a declining linear trend over the period 1979-2000. The negative trend in real wages was also a result presented in Khan (1984). In other words, there has been a secular decline in the worker’s share over the total produce. There can be different explanations for this decline. To start with, real wages may fall in the long run if rural cost of living increases at a rate higher than that in money wages. Rural cost of living includes both the food and the non-food prices of consumption. The effect of an increase in price of food on wages, however, can be ambiguous. On one hand, higher food crop prices encourage production and raise labor demand; on the other, these higher prices increase workers’ consumption expenditure and induce higher labor supply. The latter effect is more severe on landless workers who are net-buyers of food (Ravallion, 1990; Boyce and Ravallion, 1991). The ultimate effect of increase in food prices, therefore, depends on the relative strength of the demand-generating effect and the supply-generating effect of price rise. In addition, changes in crop prices may change the terms-of-trade between agriculture and industrial sectors. Agricultural workers consume such non-food industrial products as clothing and footwear, fuel and other household requisites (BBS[b], various years). Khan (1984) has explained that improvement in terms-of-trade in agriculture vis-à-vis industrial sector enables the agricultural sector to absorb higher real wage rates in long-run. In short-run, however, increased crop price reduces the real wages (Boyce and Ravallion, 1991). Once again, the

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

7

ultimate outcome of terms-of-trade effect on wages will depend on the relative strength of the long-run effect and the short-run effect. Real wage rates may also fall in the long run due to secular decline in money wage rates. This is an even more disturbing phenomenon than an increase in the prices for consumption goods, as it indicates a weakening of the bargaining position of the workers with the employers, which in turn, indicates a shift in the institutional arrangements in income sharing against the workers over the long run. An obvious implication of this change would be further increase in the income inequality in rural Bangladesh. In the absence of equitable distribution of productive assets like land, the decline in money wage rate may even lead to an increase in the absolute poverty level.

Conclusion

In this concluding section, I summarize the main results of this paper. I have started by examining the trend and the seasonality in the yield rate of crops in the ‘normal’ or non-flood years, and estimate the fluctuations in the yield in the ‘severe-flood’ years. I carry out a similar exercise for the real wage series, and estimate how wages deviate from their ‘normal’ annual and seasonal trends in the flood years. I find that in the flood years an improvement in agricultural productivity is actually be accompanied by a decline in the wage rate. I also find that while agricultural productivity in Bangladesh has shown a positive trend, agricultural wage rate has been declining over the long-run. In other words, there has been a secular decline in the worker’s share over the total produce. It has been argued in the literature that the factors that cause increase in demand for labor in flood-free periods can mitigate the negative impact of disasters in the flood-months. Increased agricultural productivity, that increases real wages, will reduce the vulnerability of the exposed population to potential short-fall in income level in times of disasters. In this regard, the need for long-term investment in agricultural, that facilitate intensive cultivation of high-yielding-variety and labor-intensive crops in the dry-season and that of flood-resistant variety of rice in the wet-season, has been emphasized (Rasid and Paul, 1987; del Ninno et al, 2003). The positive benefits of increased productivity, however, can be translated to increased welfare of agricultural workers only in presence of adequate social transfer mechanism (Montgomery, 1985). Agricultural wages in Bangladesh are also affected by such factors as land-distribution and bargaining power of workers. In presence of such structural and institutional constraints of wage determination, the issues of disaster responses in the country cannot be separated from the processes of rural development, and have to be embedded in the long-term, ongoing process of poverty reduction and social welfare.

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

8

References

BAPA. (2000). Bangladesh Environment, 2000: An Outcome of ICBEN-2000. M. F. Ahmed (Ed.). Dhaka: Bangladesh Poribesh Andolon (BAPA).

Bangladesh Ministry of Irrigation. (1986). National Water Plan, Vol. I, Sector Analysis; Vol. II. Resources; Vol. III, Alternative Plans; and Summary Reports. Dhaka. BBS[a]. (various years). Yearbook of Agricultural Statistics of Bangladesh. Dhaka: Bangladesh Bureau of Statistics (BS). BBS[b]. (various years). Statistical Yearbook of Bangladesh. Dhaka: Bangladesh Bureau of Statistics (BS). BBS[c]. (various years). Monthly Statistical Bulletin for Bangladesh. Dhaka: Bangladesh Bureau of Statistics (BS). Boyce, J. K. (1989). Agrarian Impasse in Bengal: Institutional constraints to technological constraints. New York: Oxford University Press. Boyce, J. K. (1990). Birth of a Mega project: Political Economy of Flood Control in Bangladesh. Environmental Management, 14 (4), 419-428. Boyce, J. K., & Ravallion, M. (1991). A dynamic econometric model of agricultural wage determination in Bangladesh. Oxford Bulletin of Economics and Statistics, 53 (4), 361376. Brammer, H.(1988). Floods in the agroecology of Bangladesh. Mimeo. Relief and Development Institute, London. Brammer, H. (1990a). Floods in Bangladesh: Geographical background to the 1987 and 1988 floods. Geographical Journal, 156 (1), 12-22. Brammer, H. (1990b). Floods in Bangladesh II: Flood mitigation and environmental aspects. Geographical Journal, 156 (2), 158-65. del Ninno C., Dorosh, P. A., & Smith, L C. (2003). Public policy, markets and household coping strategies in Bangladesh: Avoiding a food security crisis following the 1998 floods. World Development, 31(7), 1221-1238. del Ninno, C., & Roy, D. K. (1999a). Impact of the 1998 flood on labor market and food security and effectiveness of relief operations in Bangladesh. FMRSP Working Paper No.8. Dhaka: IFPRI (International Food Policy Research Institute). del Ninno, C., & Roy, D. K. (1999b). The1998 flood and household food security: Evidence from rural Bangladesh. FMRSP Working Paper No. 9. Dhaka: IFPRI (International Food Policy Research Institute). del Ninno, C., & Roy, D. K. (2001a). Determinants of labor market participation in rural Bangladesh after the 1998 flood. FMRSP Working Paper No. 22. Dhaka: IFPRI (International Food Policy Research Institute).

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

9

del Ninno, C., & Roy, D. K. (2001b). Impact of the 1998 flood on household food security. FMRSP Synthesis Report No. 6. Dhaka: IFPRI (International Food Policy Research Institute). del Ninno, C., Roy, D. K., & Mukherjee, S. (1999). Recovering from the shock of the 1998 flood: Household food security and nutritional status one year later. FMRSP Working Paper No. 23. Dhaka: IFPRI (International Food Policy Research Institute). DHA. (various years). Bangladesh Floods. Situation Reports. United Nations Department of Humanitarian Affairs. http://www.reliefweb.int. Hossain M., Rahman A., & Hossain Z. R. (1988). Economic impact of the 1988 floods: Impression from field visits. Dhaka: BIDS (Bangladesh Institute of Development Studies). Islam, K. M. N. (2001). Flood loss potentials and construction of standard loss data sets: A case study of residential sector of Bangladesh. Dhaka: BIDS (Bangladesh Institute of Development Studies). Khan, A. R (1984). Real Wages of Agricultural Workers in Bangladesh. In AR Khan & E Lee (Eds.) Poverty in Rural Asia (pp. 185-203). Bangkok: ILO (International Labour Organization). Maddala, G.S. (1977). Econometrics. Mc-Graw Hill Book Company. Paul, B. K., & Rasid, H. (1993). Flood damage to rice crop in Bangladesh. Geographical Review, 83(2), 150 -159.

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

10

Tables to be inserted

Table 1. Trend, Seasonality, and the Effects of Flood on Agricultural Productivity, Bangladesh, by districts, by months, January 1978-December 2000Ψ Explanatory Variables Estimated value of the coefficients (Standard error in the parenthesis) Model given by [1] Model given by [2] Agricultural productivity Agricultural productivity explained in terms of trend and explained in terms of trend, seasonality seasonality and flood occurrences Rice Jute Rice Jute 0.084 0.595 0.081 0.573 (0.002) (0.017) (0.002) (0.019) 4.006 10.003 3.822 9.74 (1.44) (2.37)** (1.41)* (2.21)** -0.083 -0.079 (0.13) (0.11) 0.143 1.132 (1.43) (1.41)* -1.145 (2.01)** 2.78 (2.23)** 0.672 -1.25 (1.48)* 0.507 (0.022) 0.76

Regression coefficient Linear trend (t) Summer (S1) Winter (S2) Flood related variables ‘Severe’ flood (Fd,t)[1]

Effect of flood in postflood crop season (S2*Fd,t) R-squared 0.34 0.27 Ψ Data: Notes: * ** *** [1] ‘Severe’ flood years 1980, 1984, 1987, 1988, 1998

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

11

Table 2. Trend, Seasonality, and the Effects of Flood on Agricultural Wage Rate, Bangladesh, by months, by districts, January 1979-December 2000Ψ Explanatory Variables Estimated value of the coefficients Model given by [3] Model given by [4] Agricultural wage rate Agricultural wage rate explained in terms of trend explained in terms of trend, and seasonality seasonality and flood occurrences -3.93 -3.89 (-0.81)*** (-0.83)*** -0.16 -0.12 (-5.22)*** (-5.98)*** -0.47 -0.43 (1.75)*** (1.62)*** 0.68 .58 (2.51)** (2.21)** -0.064 (2.19)** -0.11 (3.36)** 0.351

Regression coefficient Linear trend (t) Summer (S1) Winter (S2) Flood related variables ‘Severe’ flood (F)[1]

Effect of flood in post-flood crop season (S2*F) R-squared 0.264 Ψ Data: Notes: * ** *** [1] ‘Severe’ flood years 1980, 1984, 1987, 1988, 1998

Flood, Productivity and Wage Rate in Agriculture: The case of Bangladesh Banerjee, First Draft March 2009

12


相关文章:
flood 洪水神话
flood 洪水神话_历史学_高等教育_教育专区。Running head: WESTERN AND EASTERN FLOOD MYTH 1 Compare and Contrast the Flood Myth between west an ...
参考资料Flood
参考资料Flood_法学_高等教育_教育专区。学英语报社 http://www.e-l-e.net.cn 金色教案 全新课标理念,优质课程资源 Flood Floods are one of the most ...
HTTP Flood 攻击与防御
HTTP Flood 攻击与防御_互联网_IT/计算机_专业资料。网络攻击防御,云加速防御。防御,DDoS,HTTP Flood攻击,HTTP Flood防御,防火墙,云加速, ...
Flood_Banerjee
Flood, Productivity and Wage Rate in Agriculture: The Case of Bangladesh Lopamudra Banerjee? First Draft: March 2009 Introduction A recurrent theme in ...
UDP Flood的攻击原理及分析
18 网络安全项目实战 2 UDP Flood 的攻击原理及分析 论 文 摘 要 UDP Flood是当前最流行的DoS(拒绝服务攻击)与DDoS(分布式拒绝服务攻击)的 方式之一,这种攻击...
The Flood中英对照歌词
The Flood中英对照歌词。Cheryl Cole最新单曲The Flood中英对照歌词,绝对原创,版权所有,严禁抄袭The Flood 爱如洪水 爱如洪水 Cheryl Cole Turn the lights out, ...
进行UDP-Flood攻击练习
任务一:进行 UDP-Flood 攻击练习 UDP FLOOD 是一种采用 UDP-Flood 攻击方式的 DoS 软件,它可以向特定的 IP 地 址和端口发送 UDP 包。 UDP Flood 界面 .....
udp flood攻击
udp flood攻击_互联网_IT/计算机_专业资料。UDP flood 主要攻击一些服务器原理 UDPFlood 是日渐猖厥的流量型 DoS 攻击,原理也很简单。常见的情况 是利用大量 ...
Flood n洪水
flee 逃走 逃跑 flesh 肉 flexible 易变曲的 柔韧的 可变通的 灵活的 float 漂浮 飘动 飘 飘扬 Flood n 洪水、水灾 层 flour 面粉 v 淹没、使泛滥 fioor ...
MIKE FLOOD 手册
MIKE FLOOD 手册_电力/水利_工程科技_专业资料。MIKE FLOOD 主要应用领域:洪水预报和洪水管理 技术(产品)简介: 动态耦合的一维 — 二维洪水模型 MIKE FLOOD 是...
更多相关标签:
floor | flooding | floodlight | blood | flood攻击 | syn flood | flood滤镜 | the flood |