First Draft
Valuing Damages of Flood Induced Sand Deposition: Case of a Rice Bowl in Assam, India
KALYAN DAS
OKD Institute of Social Change and Development
Guwahati, India
January 2011
South Asian Network for Development and Environmental Economics (SANDEE)
PO Box 8975, EPC 1056
Kathmandu, Nepal
SANDEE Working Paper No
TABLE OF CONTENTS
1. INTRODUCTION 1
2. VALUING DAMAGES OF ENVIRONMENTAL IMPACTS 2
3. STUDY AREA AND DATA 4-8
3.1 DESCRIPTION OF THE STUDY AREA 6
3.2 SAMPLING FRAME 6
3.3 TABLES AND DESCRIPTIVE STATISTICS 7
4. REFLECTION ON THE CONDITION OF PADDY PLOTS: 9
RESULTS OF SOIL ANALYSIS
5. VALUING DAMAGES OF SAND DEPOSTION 11-14
5.1 VARIABLES TO EXPLAIN THE DAMAGE 11
5.2 RESULTS OF REGRESSION ANALYSIS 11
5.3 ESTIMATING THE DAMAGE VALUE 12
6. CONCLUSION AND POLICY IMPLICATIONS 13
7. ACKNOWLEDGEMENTS 14
REFERENCES
TABLES
MAPS
FIGURES
QUESTIONNAIRES
LIST OF TABLES
Table 1: Damages caused by flood in Dhemaji district
Table 2: Summary data from the stage I survey
Table 3: Average monthly household income from agriculture in the study villages
Table 4: Average monthly household income from non-agriculture jobs in the study villages
Table 5: Yield of paddy per acre in kg- own land
Table 6: Yield of paddy per acre in kg- leased in land
Table 7: Soil characteristics of the sampled paddy plots
Table 8: Sand magnitude in the study villages
Table 9: Nutrition availability in differently textured soil
Table 10: Required norms of nutrients in soil
Table 11: Nutritional availability in the sampled paddy plots by the type of soil
Table 12: Correlations Coefficients of selected explanatory variables
Table 14: Summary statistics of variables used in regression
Table 15: Regression results of determinants of paddy yield
LIST OF MAPS
Map 1: Sand Deposited Areas of Dhemaji District
Map 2: Sand Deposited Villages of Jiadhal River Basin, Dhemaji
Map 3: Study Villages: Paddy Plots and Embankment Breaching Points 2009
LIST OF FIGURES
Figure 1: Trend of Yield of Paddy in the Selected Districts of Assam
Figure-2 Sand and paddy yield in the sampled plots
Figure 3: Sand as Determinant of Paddy Yield (Lowess smoothing, all paddy plots)
Figure 4: Sand as Determinant of Paddy Yield (Lowess smoothing, paddy plots using HYV seeds)
Abstract
This study measures the value of damages caused by flood induced sand deposition in paddy fields of Jiadhal basin in Dhemaji district, India. The changing texture of soil has created a climate change like situation, with poor water retaining capacity of the soil to sustain crops even in this high rainfall area. The study adopts Ricardian approach but finds certain constrains because of virtual absence of land transactions in the degraded environment and sharing of output in rented plots on equal proportion. This has, however, made no differences on the outcome capturing data on paddy return from 346 agriculture plots and regress the return with the soil characteristics of the plots. Soil test shows high presence of coarse sand, low pH, low organic carbon and subsequent low availability of nutrition. The coefficients of sand, pH and organic carbon show negative and significant associations with paddy yield. The annual damage value is estimated at INR 437,000/ (about $ US 9,300) to INR 1,337,769 (about $ US 28,000) per village, large enough to ruin economy of small villages of about 70 households. In this context of irreversible damage overall strategy now requires rethinking on the embankments, regulation of anthropogenic activities in the upstream and strengthening of social security provisions of the state.
Key words: Valuation of land degradation, Ricardian Approach, Soil Texture Test, Embankment, Adaptation, Social Security Provisions
24
Valuing Damages of Flood Induced Sand Deposition: Case of a Rice Bowl in Assam, India
Kalyan Das
1. Introduction
There are certain positive impacts of flood, if flood water deposits alluvial (silt) nourishing the agriculture fields. Damages caused by flood are, however, enormous with destruction of standing crops, damage of houses and deposition of sandy layers. The damage caused by the deposition of sandy layer is not common (degraded land classification of National Remote Sensing Agency of India; Reddy, 2003) but once done this damage is irreversible[1]. This factor finds little space in the discourses on flood havoc and related policy prescriptions and mitigation plans in the flood ravaged state of Assam in northeastern region of India.
Dhemaji district, the rice bowl[2] in the state of Assam now to a significant extent has transformed to a virtual desert of sand deposition caused by the flood of Himalayan tributaries of the Brahmaputra River (Map 1). Estimate derived from satellite imagery in 1999 reveals that an area of 14,879 hectares has been rendered unproductive by sand deposition in the district (National Resource Management Foundation, 2006). The information derived from near complete enumeration[3] of 1059 households covered for this study in 15 sample villages in the Jiadhal river basin of the district reveal sand deposition to the extent of 83 percent in the total acreages of 3730 acres of their paddy fields. Soil test of 346 agriculture plots in 148 sample household establishes that the soil texture now comprises 54 percent sand and 36 percent silt (measured during 2009-10, but could vary yearly). Secondary data reveals that during 1992 to 2004-05 net sown area of the district has reduced by 7,500 hectares and fallow and uncultivated land has increased by more than 7,800 hectares (Statistical Handbooks of Assam, 1996 and 2008). These figures indicating degradation of land are significant in the district having a total cultivable area of about 63,000 hectares.
The sediments transported by Rivers are of two types – the bed sediment load of coarser particles and wash loads of finer particles (micro nutrients). Generally in high flow conditions, only the finer fractions of bed sediment load are carried by wash load with possibility of depositing in the agriculture field during flood (Weibe, 2006). Flood in the district now carries and deposit relatively heavy bed sediments loads (sand)[4]. Local people and researchers point at the acceleration rates of sedimentation to shifting cultivation, rampant tree felling for commercial purpose and extraction of boulders in the upstream mountain valleys (some indication in Goswami and Das, 2003). Primary observations indicate that some other valley districts in the north bank of Assam though inundated by flood (in many cases by breaching of embankment) of numerous Himalayan tributaries; the damage caused by sand deposition is not common. A World Bank policy paper (Weibe, 2006) on river flooding and erosion in north east India addresses the issues of sedimentation and rise in river beds, channel shifting and erosion; but, the damage caused by sand deposition in the agriculture field attracts no attention. There are, however studies (Caldwell et al., 2002 in the context of Thailand), indicating decline in paddy productivity due anthropogenic factor induced sand deposition. In the Dhemaji district since the 1990s people have started to face the repercussions of land degradation caused by sand deposition.
Assessment of the cost of the damages done in such situation has three dimensions. One- there is cost involved with depleting housing stocks and standing crops; two- the cost of land damage which to an large extent is irreversible; and three- the cost in the mitigation plans, relief operations of the state as well as of the affected people in coping and adaptive process. This is reflected that policy papers, prescription and valuation of flood damages miss out the second dimension, underestimating the damages comprehensively.
The primary goal of the study is to measure the value of the damage caused by flood induced sand deposition in the cultivable land. The study tries to use Ricardian approach establishing the association of return from paddy plots to the changing physical conditions of land.
This study has VI sections. Section II provides a brief review on the approaches and methods in valuing damages of environmental impacts and makes the ground for methodology of the study. Section III discusses about the characteristics of the study area facing the ravages of flood and sand deposition. This section also discusses the sampling frame and shares descriptive statistics on economic conditions of the households. Section IV presents the laboratory results of soil analysis to narrate the present condition of the flood ravaged paddy plots. Section V identifies the determinants influencing the return of paddy and estimates the damage values of declining return of paddy. Section VI, finally discusses the issues relating to adaptation of people in degraded environment.
2. Valuing Damages of Environmental Impacts
Contemporary literatures reveal adoption of two approaches-the Production Functions and the Ricardian Approach in valuation of damages of environmental impacts. On the positive side, the production function is the only approach which estimates in satisfactory manner the potential beneficiary or adverse effects (Kumar and Parikh, 1998). However, this is argued that in damage valuation, the traditional production function approach has certain inherent bias. Production function overestimates the damage as it omits variety of the adaptations that the farmers customarily make in response to changing economic and environmental conditions. Introduction of a new crop in changing physical condition of land, technological changes and changes in land uses are often not considered in contemporary literature (Mendelsohn et al., 1994).
The critic is that the production function approach fails to take into account of economic substitution as conditions change in innumerable farms under study, and this bias or failure in principle can be corrected by Ricardian approach (Mendelsohn et al., 1994). Ricardian approach provides direct evidences on how profit maximizing farmers respond to various climatic or environmental conditions. This approach collects the information on value of land and farm revenues, how conditions in different places affect the net rent or value of land. The arguments of Maddison et al. (2007) are on the same line. The assumption in Ricardian approach is that, under competitive markets, the value of land or land rent will be equal to the net yield of highest and best use of land. In farms land rent tends to be a large fraction of total costs and can be estimated with reasonable precision (Mendelsohn et al., 1994, p 755).
Researchers have used the Ricardian approach to estimate the implicit value of a variety of land characteristics and the impacts of changes in climatic variables on agriculture productions. Kumar and Parikh (1998) have adopted the net revenue approach (land prices presumably based on expected future revenue) incorporating the optimize choices of inputs and outputs. District level estimates of annual net revenues were regressed on climate and other social and economic variables in their study to measure impacts. Maddison et al. (2007) regress the agriculture outcomes as revealed by land values or net revenues on environmental variables to measure the estimated importance of the environmental variables. This study, however, unlike other studies uses farmers’ own perception on value of the land. Research adopting Ricardian approach in general relied on either observed sale prices or net revenue aggregated. Maddison (2007) study has used actual sale price at the level of the individual farm to determine the value of climate. There are, however, criticisms on using sale price in the Ricardian approach (Koundouri and Pashardes, 2003). This is because the farm value may vary for the presence of structural attributes in the land. Furthermore, in any given year only a very small fraction of farmland is likely to be put on the market making the sample unrepresentative.
Ricardian approach requires incorporation of several supportive parameters those could influence the value of land. Availability of water could change the value of land even at adverse environmental conditions (Mendelsohn and Diner, 2003). Further a host of households and farmers’ characteristics, such as human and physical capital, land tenure status could control the land value (Kurukulasuriya and Ajwad, 2004). There are however, certain problems of calculating the net revenue in developing country context like India. The problems are in valuing the family labour inputs and the produce consumed by the subsistence farms can be very difficult (Kumar and Parikh, 1998).
Critic of Ricardian approach
Darwin (1999) comments, the Ricardian approach can provide useful information about economic affect of climate-induced agriculture changes; but the approaches as adopted by Mendelsohn et al. (1994) are not fully consistent with the basic Ricardian principle of land becoming unsuitable with higher level of temperature. Their preferred model (the crop revenue model) greatly overestimates the value of warm-weather agriculture in the US. Darwin tells, the omission of the variable irrigation, is a likely reason. However, in a reply to Cline (1996) that Mendelsohn et al. (1994) assumes the supply of water for irrigation is perfectly elastic; Mendelsohn and Nordhaus (1996) comments that the 1994 study models irrigation in a reduced form manner assuming that the availability of water in a region is a function of its climate and geography, including such factors as mean and seasonal precipitations and temperature, slope, and so forth. Treating irrigation in a reduced-form method, this implicitly assumes that the market or shadow prices are a function of climatic and geological variables. The Ricardian model assumes that, if the climate were to change to hotter and drier conditions, water availability under this new condition would resemble the water availability observed in the hot dry places today.
Making the ground for the methodology of the study
The review reflects that we can measure the affect of sand variable on agriculture outcome, land rentals and land value using the Ricardian approach. The information derived from the enumeration of 1059 households in 15 villages reveal that the villages in this remote, underdeveloped and flood ravaged district of Dhemaji do not have an active agricultural land market. Land transaction is virtually absent in nine of the 15 villages studied. In rest six, land transactions though exist to an extent, the prices fetch are minimal, in between Rs. 18,000/ ($US 375 to Rs. 36,000/ per acre ($US 750).
Contrary to minimal presence of transactions of sell and purchase of land (an indication that people do not value the degraded land), one finds the general practice of leasing in land for agriculture purpose. Poverty, lack of sources of income avenues compels the people in affected villages to look for best possible land to lease in. The rental value of land, here, however may not be the same as the annual net revenue; that at a given environmental condition equals to the maximum amount of rent one willing to pay for a year. This is because, in these villages, the land rent in the context of leasing in land for agriculture purpose is basically sharing the output at 50:50 basis whatever may be the return. Interactions also reveal that in the leased in land the lessee’s effort to invest in inputs and derive higher yield are not there. This is because of the prevailing uncertainty of the occurrences of multiple waves of flood during the year.