WPS4060




         Distributional Effects of WTO Agricultural Reforms
                              in Rich and Poor Countries*

                                                     by

                       Thomas W. Hertel** and Roman Keeney
                                         Purdue University
                                   Maros Ivanic GOIC, Doha
                                                     and
                                           L. Alan Winters
                                           The World Bank




World Bank Policy Research Working Paper 4060, November 2006

The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the
exchange of ideas about development issues. An objective of the series is to get the findings out quickly,
even if the presentations are less than fully polished. The papers carry the names of the authors and should
be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely
those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors,
or the countries they represent. Policy Research Working Papers are available online at
http://econ.worldbank.org.


*Preliminary version of a paper prepared for the 44th Panel Meeting of Economic Policy,
Helsinki, Finland, October 20­21, 2006. The Annexes are available from
https://www.gtap.agecon.purdue.edu/resources/download/2816.pdf.

** Corresponding author: T. W. Hertel, Center for Global Trade Analysis, Purdue
University, 403 West State Street, W. Lafayette, IN 47907; hertel@purdue.edu

  Distributional Effects of WTO Agricultural Reforms in Rich
                                 and Poor Countries


                                         Abstract


        Rich countries' agricultural trade policies are the battleground on which the future

of the WTO's troubled Doha Round will be determined. Subject to widespread criticism,

they nonetheless appear to be almost immune to serious reform, and one of their most

common defenses is that they protect poor farmers. Our findings reject this claim. The

analysis conducted here uses detailed data on farm incomes to show that major

commodity programs are highly regressive in the US, and that the only serious losses

under trade reform are among large, wealthy, farmers in a few heavily protected sub-

sectors. In contrast, analysis using household data from 15 developing countries indicates

that reforming rich countries' agricultural trade policies would lift large numbers of

developing country farm households out of poverty. In the majority of cases these gains

are not outweighed by the poverty-increasing effects of higher food prices among other

households. Agricultural reforms that appear feasible, even under an ambitious Doha

Round, achieve only a fraction of the benefits for developing countries that full

liberalization promises, but protect the wealthiest US farms from most of the rigors of

adjustment. Finally, the analysis conducted here indicates that maximal trade-led poverty

reductions occur when developing countries participate more fully in agricultural trade

liberalization.


Keywords: WTO, Poverty, Trade Liberalization, Distributional impacts of policy




                                              ii

  Distributional Effects of WTO Agricultural Reforms in Rich
                                and Poor Countries

  "Trade theory is about whose hand is in whose pocket and trade policy is about who
                             should take it out." Finger (1980)


        This paper is about some well-known hands in well-known pockets but in new

combinations and at a level of detail that has not previously been possible. For the first

time it considers the trade-offs in global agricultural trade reform between farmers in rich

and poor countries making use of farm-level and household-level data. It delves further

into the distributional consequences of reform than previous research and in doing so lays

bare some of the political economy that has made agricultural trade reform so tortured.

        The Doha Development Agenda (DDA) of the World Trade Organization (WTO),

which is currently suspended, has an explicit mandate to improve welfare and reduce

poverty in developing countries (WTO, 2004). The bulk of the global gains from

merchandise trade reform derive from reforms in agriculture (Hertel and Keeney, 2006;

Anderson and Martin, 2006), and most of these are predicted to accrue to rich countries

as they reduce outlays on farm programs and reduce protection for agricultural products.

But such reforms also benefit many households in developing countries ­ particularly

those in the farm and rural sectors, which comprise a majority of the world's poor ­ so it

would seem that such reforms should be an easy sell to policy makers in rich and poor

countries alike. Experience suggests the opposite, for the failure to agree on a package of

agricultural liberalizations was the principal cause of the suspension of talks in mid-2006.

       While agricultural reforms in industrial countries are indeed likely to benefit large

and diffuse groups of taxpayers and consumers, they will hurt some of the farm sector ­

with the impact concentrated on some of the most powerful and well-organized interest

groups in that sector. By contrast, farmers in developing countries -- the potential

beneficiaries of reform -- have little or no influence in the political process, while their

urban counterparts have some interest in maintaining the status quo.

       The political economy of trade policy has long recognized the greater

effectiveness of concentrated lobbies ­ see, for example, Winters (1987) or Anderson

(1995) on agriculture ­ and 70 years ago Schattschneider (1936) recognized that one

needs to evaluate such concentration at a fine level of disaggregation. Thus, in this paper

we argue that the interesting issue in agricultural reform is not the potential global

welfare gains, although these can be substantial (Anderson, Martin and van der

Mensbrugghe, 2006), but rather, the distribution of the benefits and costs of reform

across households in rich and poor countries. Given the ambition of this exercise, we can

consider only the United States among rich countries, and 15 developing countries for

which we can assemble household survey data on income sources on a relatively uniform

basis.

       We are very interested in the impacts of agricultural reforms previously deemed

possible under the DDA, but to help advance the policy-making process ­ currently at a

stalemate -- we devote considerably attention to those reforms not currently under

consideration. Notably, we consider greater agricultural liberalization by developing

countries, which turns out to be pro-poor ­ and some compensation mechanisms which

might reduce rich country opposition to agricultural trade reforms.



                                              2

        This analysis contains four key steps: the specification of a plausible DDA

agreement including the translation of these into cuts in actual agricultural support;

calculating the impacts of such reforms on global trade, prices and production; tracing

these global impacts back to different classes of farm households within the US; and

tracing them back to households in our focus developing countries. The combination of

these steps into a holistic framework represents a significant contribution of this work,

which brings together data and modeling components to conduct global scale analysis.1

        DDA Specification: There have been many studies of WTO trade reforms in the

context of the DDA, but few of these bear close relationship to the actual negotiations

undertaken in Geneva or to actual trade barriers in the world at the time the DDA will be

implemented. In contrast, recent studies based mostly on the GTAP 6 database, recognize

the significance of trade preferences for developing countries' exports and also that the

DDA will be implemented in a world in where China has acceded to the WTO and the

EU has been enlarged (reference CEPII work here).2 This is the approach taken here. We

build on two recent World Bank projects which begin with tariff line data and specify

agricultural market access scenarios based on detailed analysis of tiered formula cuts in

current levels of tariff bindings (Anderson and Martin, 2006; Hertel and Winters, 2006).

In cases where post-reform bindings fall below currently applied tariff levels,

liberalization is predicted to occur. If this is not the case, no actual liberalization occurs

despite the reduction in tariff bindings. This detailed analysis is particularly critical for

analyzing developing countries, where bound tariffs are high and reductions in these



1We also offer modest methodological advances on the previous literature in two of the four steps.




                                                 3

bindings are modest due to special and differential treatment. Similar detail is necessary

for prospective reductions in domestic support (Jensen and Zobbe, 2006).

        Given a set of plausible liberalizations, we need to translate these into a set of

changes in prices, outputs, inputs, etc. around the world. Since reforms are widespread

sectorally and geographically, this requires a global, multi-sectoral, general equilibrium

approach as epitomized in global computable general equilibrium (CGE) models. Many

such models have been used to analyze trade reforms, each emphasizing different

features according to the authors' purpose. Box 1 offers a brief introduction to the

essential features of CGE analysis.

        Distributional Impacts for US Farm Households: US farm household population

data are taken from the Agricultural Resource Management Survey (ARMS) (USDA-

ERS, 2005). This comprehensive survey of US farm households is conducted over a

sample of around 15,000 households using economic and geographic sampling frames.

Sampled farms include calibrated weights for aggregating the individual data into group-

wise representatives.




2Studies dated prior to 2004 typically miss these features.




                                                      4

Box 1 Computable General Equilibrium Modeling
        General equilibrium, which dates back to Leon Walras (1834-1910), is one of the crowning intellectual
achievements of economics. It recognizes that there are many markets and that they interact in complex ways so that
loosely speaking, everything depends on everything else. Demand for any one good depends on the prices of all other
goods and on income. Income, in turn, depends on wages, profits, and rents, which depend on technology, factor
supplies and production, the last of which, in its turn, depends on sales (i.e., demand). Prices depend on wages and
profits and vice versa.
        To make such an insight useful, economists have to be able to simplify it sufficiently to derive predictions
and conclusions. Theorists typically do this by slashing the dimensionality, say to just two goods, two factors and two
countries, and often focusing on just a few parts of the system. An alternative approach is to keep the complex
structure but to simplify the characterization of economic behavior and solve the whole system numerically rather
than algebraically. This is the approach of Computable General Equilibrium (CGE) modeling.
CGE models specify all their economic relationships in mathematical terms and put them together in a form that
allows the model to predict the change in variables such as prices, output and economic welfare resulting from a
change in economic policies, given information about technology (the inputs required to produce a unit of output),
policies and consumer preferences. They do this by seeking prices at which supply equals demand in every market--
goods, factors, foreign exchange. One of the great strengths of CGE models is that they impose consistency of one's
view of the world, e.g., that all exports are imported by another country, that the sum of sectors' employment does
not exceed the labor force, or that all consumption be covered by production or imports. This consistency can often
generate empirical insights that might otherwise be overlooked in complex policy analysis ­ such as the fact that
import protection gives rise to an implicit tax on exports.
        The mathematical relationships assumed are generally rather simple, and although `many' markets are
recognized, they still have to be very aggregated--particularly for global economic analysis. For example, the global
CGE model used in this paper has 31 sectors, so, for example, `transport and communications services' appear as a
single industry. In principle all the relationships in a model could be estimated from detailed data on the economy
over many years. In practice, however, their number and parameterization generally outweigh the data available. In
the model used for this paper, only the most important relationships have been econometrically estimated. These
include the international trade elasticities (Hertel et al., 2005), the agricultural factor supply and demand elasticities
(OECD, 2001), and consumer preferences (estimated specially for this paper, based on the methods outlined in
Cranfield et al., and Reimer and Hertel). The remaining economic relationships are based on literature reviews, with
a healthy dose of theory and intuition. An important limitation of CGE models is that very few of them are tested as a
whole against historical experience--although ours is one such (Valenzuela et al., 2007, Liu et al., 2004).
        Having specified most of the relationships, the CGE modeler manipulates a subset of parameters so that the
model will replicate detailed data for one `base' year--this is known as `calibration.' To calculate the effects of a
policy change the model is solved once without the change and once with it in an otherwise identical universe and the
difference in outcome calculated. Often the `without' scenario is just the data base year, although in our case we
project key features of the global economic policy environment economy forward to 2005 in order to facilitate our
analysis of the Doha Round.
        In summary, CGE modeling is a very powerful tool, allowing economists to explore numerically a huge
range of issues on which econometric estimation would be impossible; in particular to forecast the effects of future
policy changes. The models have their limitations, however. First, CGE simulations are not unconditional predictions
but rather `thought experiments' about what the world would be like if the policy change had been operative in the
assumed circumstances and year. The real world will doubtless have changed by the time we get there. Second, while
CGE models are quantitative, they are not empirical in the sense of econometric modeling: they are basically
theoretical, with limited possibilities for rigorous testing against experience. Third, conclusions about trade policy are
very sensitive to the levels assumed for trade restrictions in the base data. One can readily do sensitivity analysis on
the parameter values assumed for economic behavior (as we have done in this paper) but less so on the data, because
altering one element of the base data requires compensating changes elsewhere in order to keep the national accounts
and social accounting matrix in balance. Of course, many of these criticisms apply to other types of economic
modeling, and therefore, while imperfect, CGE models remain the preferred tool for analysis of global trade policy
issues.



                                                              5

       These data distinguish farm households' places in the wealth distribution,

commodity sources of farm income, and detailed information on off-farm income so that

changes in total income and welfare can be calculated in the wake of agricultural reforms.

Keeney (2005) uses these data to analyze the distributional consequences of stylized

WTO scenarios, representing the only previous analysis of US farm household impacts of

a Doha agreement. The ARMS data have served as the source for other disaggregate

analyses (most notably Hanson and Somwaru's (2003) work on the WTO acceptability of

counter-cyclical payments) but in these cases the distributional character has been

focused on farm structure rather than the welfare focus of Keeney (2005), and global

reforms have not been considered.

       Distributional Impacts for Poor Country Households: Winters (2002) and

Winters et al. (2004) provide an analytical framework and evidence on tracing the effects

of trade policy through to individual households and poverty. Hertel and Reimer (2005)

develop this framework in the context of CGE modeling. We believe that the impact of

trade reform on individual households will vary widely depending on their sector of

primary employment, their endowments, and their consumption patterns. Therefore for

each of our 15 focus developing countries we utilize household survey data to divide

households into seven classes (strata) according to their principal income source and

estimate factor-specific poverty elasticities for each country and stratum combination.

These elasticities are incorporated directly into our global CGE model and embody

information about the shape of income distribution and income sources in the

neighborhood of the poverty line. When combined with estimates of consumption

behavior at the poverty line, those estimates allow for accurate assessments of how



                                               6

poverty headcounts will likely change in the wake of WTO trade reforms. Drawing on

the results for the 15 developing countries in our sample, we seek to arrive at some

general conclusions about the poverty impacts of trade policy reforms in rich and poor

countries.

        The remainder of the paper is organized as follows. We proceed with discussion

of the unique analytical framework created for this study's analysis of the distributional

impacts of WTO reforms in both rich and poor countries. Following that, we outline the

policy scenarios to be applied in this framework. The results section begins with

discussion of changes in macro indicators for trade, prices, and national welfare as well

as changes in US farm household welfare and change in developing country poverty

focusing on the impacts of agricultural reforms undertaken in rich countries. We extend

this analysis to global reforms and non-agricultural sectors, separately identifying the

contributions of these reforms to the poverty headcount results. The concluding section

summarizes our findings and offers policy recommendations.


Analytical Framework

        Figure 1 provides an overview of the analytical framework used in this paper. The

boxed items in the top two rows represent inputs to the framework, and the boxes at the

bottom of the Figure represent outputs of particular relevance to this study. The other

entries represent intermediate steps in the analysis. As can be seen, we begin with three

fundamental sources of data: household survey data from the US, household survey data

from the 15 focus countries, and the GTAP data base. Agricultural earnings data in the

latter two sources are reconciled, as the GTAP data are notoriously weak when it comes




                                             7

to the estimation of returns to self-employed labor in the farm sector (see Annex III for

details). The reconciled survey data are used to compute the poverty elasticities discussed

in Box 2, while the revised GTAP data are used to specify agricultural technology in the

global CGE model. Other inputs to the global modeling exercise include: farm income

sources by farm type for the US, the poverty headcounts, by region, for $1/day and

$2/day, the estimated parameters for our consumer demand system, estimates of farm

factor supply and demand elasticities from the OECD, as well as the trade reform

scenarios (see Table 6). These inputs are combined with a modified version of the GTAP

CGE model of the global economy.


        With this overview in mind, a bit more needs to be said about some key aspects of

the analytical framework. Our starting point for this analysis is the GTAP version 6.1

data base (Dimaranan, 2006). Virtually all contemporary analyses of the Doha

Development Agenda start at this same point. Data availability is easily the most limiting

resource for global analysis and GTAP version 6.1 represents the only data base covering

global economic activities with bilateral trade and protection data that reflects tariff

preferences. This also permits us to draw on the carefully constructed Doha reform

scenarios developed and utilized in the recent books by Anderson and Martin (2006), and

Hertel and Winters (2006).3 These scenarios also involve a pre-experiment in which key

trade policies are updated to 2005, and it is from that benchmark that the trade

liberalization experiments proceed.



3These tariff cutting scenarios are now available on the GTAP web site to those wishing to replicate this
work. For purposes of this paper, we have used scenarios S0 (pre-simulation with China's WTO accession,
EU enlargement, etc.) and S8: the central Doha scenario used in the Hertel-Winters volume.



                                                   8

        Our modifications to the standard GTAP model focus on features that enhance

analysis of agricultural reforms and simulation of distributional impacts. We retain the

simplistic yet empirically robust assumptions of constant returns to scale and perfect

competition typically featured in agricultural trade studies.4 Our modifications are aimed

at permitting us to shed new light on the distributional consequences of WTO reforms ­

focusing particularly on the seemingly intractable problem of agriculture liberalization in

the industrial countries. We turn now to these modifications.


        Factor Markets: Since the work of T.W. Schultz (1945), economists have

recognized the importance of off-farm factor mobility in determining farm incomes.

Significant wage differentials between farm and non-farm employment persist in the

United States and other high income economies (Gardner, 1992; Kilkenny, 1993). The

limitations of agricultural labor markets have also been prominently featured in the

development economics literature, as an explanation for the very low level of agricultural

supply response (de Janvry et al., 1991). The common CGE assumption of perfect

mobility of labor and capital from agriculture to non-agriculture forcing wages to

equalize at each point in time for farm and non-farm workers with comparable skills, is at

odds with historical observation.


        Effectively modeling the complex processes leading to limited farm/non-farm,

rural/urban mobility for the full range of countries in our model would be a lifetime

project. Instead, we specify a constant elasticity of transformation function which


4Francois et al. (2004) introduce monopolistic competition in the manufacturing sector into their analysis
of WTO reforms. The resulting variety and scale effects generally boost the gains to rich countries and




                                                  9

"transforms" farm-labor into non-farm labor and vice-versa. This transformation function

permits wages to diverge between the farm and non-farm sectors, a key driver for our

distributional analysis. With segmented labor markets, the impact of reduced subsidies to

agriculture in the rich economies will not be shared equally between the farm and non-

farm labor forces. Similarly, the benefits from higher farm prices in developing countries

following rich country reforms will not be shared as widely with non-farm households in

the presence of factor market segmentation.

         Much of the reasoning behind differing agricultural and non-agricultural labor

rewards similarly applies to returns to agricultural investment. Therefore, we also

introduce a constant elasticity of transformation function governing capital movements

between agriculture and non-agriculture, with full capital mobility (a unique rental rate

on capital) only applying across uses within these two broad sectors.

         The extent of burden shifting between farm and non-farm labor and capital will

depend on the size of the associated factor supply elasticities. In order to calibrate these

key parameters, we draw on the OECD's (2001) parameterization of agricultural factor

markets which derive from comprehensive econometric reviews for the EU (Salhofer,

2001) and for North America (Abler, 2001) as well as an modeling panel's assumptions

for the Japanese economy.

         We assume a constant aggregate level of land, labor, and capital employment

reflecting the belief that the aggregate supply of factors is unaffected by trade policy.

This is not the `full employment' assumption sometimes ridiculed by advocates of



dampens the gains to poor countries from rich country reforms. However, this makes their model less
stable, and, given our focus on agricultural reforms, this feature seems less critical.



                                                      10

structuralist models of development, rather it assumes that aggregate employment is

determined by factors such as labor market norms and regulation that are largely

independent of trade policy in the long run. Absent sufficient detail on these employment

drivers, we look to wage changes to clear farm and non-farm the labor markets in each

country.5

        Rich Country Farm Household Impacts: The potential for adverse impacts on

rich country farm household incomes has received far less attention than the

distributional impacts in poor countries, yet it represents a key component of the political

economy of WTO trade reform. A primary factor in determining the impact of

agricultural reforms on farm household welfare in rich countries is the share of their

income that currently comes from the farm sector. If farm income is only 10% of total

household income, then a 10% drop in farm income translates into just a 1% drop in

overall household income (for constant non-farm income). Recent OECD (2003)

statistics report the on and off-farm income split for farm households in numerous

member countries ­ see Annex Table A.6.1. Farm income provides only 8% of the total

income of US farm households and 10% and 12% in Canada and Japan respectively. In

Europe the share is larger, in 60%-70% range.

        In the global CGE model, we model a representative farm household for each

region and explicitly track the allocation of its labor and capital between the farm and

non-farm sectors and the allocation of its land across agricultural uses. As returns in

agriculture fall when subsidies are removed, farm households reallocate some farm-



5 This market clearing assumption means that our model does not generate the large changes in
competitiveness that Polaski finds when real wages become misaligned.



                                                   11

owned resources to the non-farm sector as well adjusting the output composition to

changes in relative land returns. Total farm household income in the model is then

determined as the sum of returns on their endowments employed in agriculture, plus the

returns on those employed in non-agriculture.

        While the average farm household's welfare change is an important component in

assessing WTO outcomes for any given country, greater detail on the distribution around

this average is required to develop insight into the political economy of agricultural

reform. This requires more disaggregate data. We have obtained these data for the US,

and we use a "micro-simulation" technique that uses the general equilibrium changes in

product and factor prices to evaluate the welfare impact on different groups of farm

households in this country. These different groups are defined first by their product

specialization and then by their place in the wealth distribution of similarly specialized

producers. The households and their initial income sourcing are benchmarked using the

ARMS annual survey data of the United States farm household population for 2004. The

ARMS survey data has no longitudinal component, and Hill (1996) argues that in such

cases wealth provides a suitable substitute for multi-period averages necessary to

accurately gauge the income position of farm households.

        Table 1 identifies the disaggregate US farm households of our study. They

represent income specialized households in four highly protected sub-sectors: dairy,

cotton, rice and sugar, and a residual category of non-specialized farm households. The

specialization criterion is that at least 1/3 of farm revenue be derived from rice, cotton, or

dairy (to be specialized in those products), and 1/5 of farm revenue from sugar (to be

specialized in sugar). The second line of delineation among households distinguishes



                                                12

eleven intervals in the wealth distribution of each specialization group. The farm income

share for the specialized groups ranges from 0.22 to 0.92 with larger dependencies for

wealthier farms. The residual category "Other", is by far the largest in the population and

mirrors the aggregate distribution of US farm households. Its low farm income shares

contrast sharply with those of the specialized farms.

        The choice of dairy, sugar, rice, and cotton as focus households is driven by the

level of support and protection these products enjoy in the US: about 50% of total

producer revenue for US milk, sugar and rice is attributable to farm programs (OECD

2002) while government programs provide about 35% of revenue for cotton producers

(Sumner, 2005). Other products like maize and oilseeds receive less support in the US

(25%) as do livestock products (less than 5 %). In addition, maize, oilseed, and livestock

producers in the US tend to be much more product-diversified in farm revenue. Thus, the

focus of our analysis is squarely on those households specialized in highly protected

products. In particular, we believe that high levels of support foster income specialization

and specialization enhances interest group formation and lobbying around a specific

agricultural product. Our results will provide insight into this dynamic that disfavors

policy reforms in the most needed areas.

        Poverty Assessment: There are many dimensions through which rich country

reforms affect developing countries. Here we focus on the poverty headcount ­ that is,

the proportion of the population that falls below the poverty line. This is the most widely

cited figure in the literature, and, by considering two different poverty criteria ($1/day

and $2/day), we explore the sensitivity of our findings to the choice of poverty line. We

do this for 15 focus countries for which we have been able to assemble comparable



                                              13

household survey data. These countries are listed in Table 2 and together they span the

continents of Africa, Asia and Latin America. In the aggregate, they account for nearly 1

billion people, and more than 400 million poor (measured at the $2/day poverty line; 150

million poor when evaluated at the $1/day poverty line). While they are not a random

sample, they do span a wide range of per capita income levels as well as differing degrees

of industrialization. Therefore, as we will see, the location and earnings patterns of the

poor in these 15 countries vary greatly.


        There are many alternative approaches to estimating the poverty impacts of trade

reforms (Annex II). The analytical approach used here builds on that of Hertel et al.

(2004), which employs a sequential, macro-micro modeling strategy in which results

from the global model are passed on to a series of micro-simulation models. In this paper

we summarize the key characteristics of these micro-simulation models using highly

disaggregated poverty elasticities ­ describing the impact of a change in various

components of earnings on poverty within a given population group, or stratum. This

permits us to present and analyze our results for all 15 focus countries in a compact and

easy to understand manner while maintaining the diversity of poverty outcomes under

global trade reform.


        A key finding in the work of Hertel et al. (2004) is the importance of stratifying

households by their primary source of income. Unlike some of the rich countries (and

particularly the US, as discussed above), farm households in developing countries often

rely on the farm enterprise for virtually all of their income and are likely to be highly

diversified in the products grown on the farm. Furthermore, the share of national poverty



                                              14

concentrated in agriculture-specialized households is quite high in the poorest countries

in our sample ­ between one-quarter and one-half of the $1/day headcount in Chile,

Colombia, Indonesia, Malawi, Mozambique and Zambia. On the other hand, this share is

relatively small in Mexico, Peru and Thailand, where a much smaller proportion of the

households are engaged in farming, as well as Vietnam, where rural households are more

likely to have substantial off-farm income.


         Not only are farm households in the poorest countries more likely to be

specialized in farming, these specialized farm households also tend to be poorer, on

average, than the rest of the population. This point is evident from Figure 2 which plots

the poverty headcount in the entire population (horizontal axis) against the poverty rate in

the agriculture-specialized group (vertical axis). With the exception of Peru, Mexico and

Venezuela, which lie slightly below the 45 degree-line, it is clear that agriculture

specialized households have a higher poverty rate ­ indeed, in the case of Brazil, this is

about six times the national poverty rate. The implication of this pattern of farm income

specialization is that the poorest households in the poorest countries are more

concentrated on agriculture and therefore more likely to benefit from producer price

increases engendered by multilateral trade reforms.


         We follow Hertel et al. (2004) in identifying five household groups that rely

almost exclusively (95% or more) on one source of income: agricultural self

employment, non-agricultural self-employment, rural wage labor, urban wage labor, or

transfer payments. The remaining households are grouped into rural and urban diversified

strata, leading to seven total strata. Table 2 reports the share of the total national poverty



                                               15

headcount ($1/day) arising in each stratum, for each of our 15 focus countries.

Agriculture specialized households and rural diversified households tend to dominate the

poverty headcount, although exceptions are Colombia, Venezuela and Peru, where self-

employed, non-agriculture households contain a large share of the poor.

        The change in the national poverty rate is calculated from the changes in the

poverty headcount in each stratum, The latter depend on the density of the income

distribution in the neighborhood of the poverty line. This can be usefully captured by the

stratum-specific poverty elasticities which have been computed numerically based on the

cumulative income distribution taken from the household survey data for each of the

focus countries (Box 2). These are reported in Table 3, and they answer the question: If

incomes in a given stratum rise by one percent, what percentage reduction in the poverty

headcount will be achieved? They range from a low of 0.0006 in the self-employed

agriculture stratum in Zambia, where nearly all of the population is well below the

poverty line, to a high of 3.63 in the urban diversified stratum of Brazil, where the

population density at the poverty line is quite high.

        However, all income sources are not equally important for households in poverty.

In most cases these households own few assets, and have few skills, so their primary

endowment is unskilled labor. Increased returns to capital in the wake of trade reforms

will do little to reduce poverty. However, a rise in the unskilled wage will make a great

deal of difference. This fact is captured in our work by disaggregating the poverty

elasticities by income source, as shown in Table 4 for the case of Peru. These elasticities

measure the percentage change in stratum poverty headcount, in response to a one

percent increase in returns to different types of household endowments.



                                              16

Box 2. Estimating Poverty Impacts in the Focus Countries
         The unifying theme of our results is that different households are affected differently by trade reforms. Thus
how we derive and treat differences among households is central to the analysis. The most consistent approach embeds
household behavior fully within the national CGE model, but this is computationally burdensome (Rutherford, Tarr and
Shepotylo, 2006) and would add significant complexity to an already complex global analysis. A popular simplification
involves solving a national CGE model and combining the resulting changes in commodity prices, factor prices and
possibly quantities and employments with household data on earnings and expenditures to estimate a (first-order)
approximation of the welfare effects on households. Chen and Ravallion (2004) apply this to 80,000 households to
estimate the poverty effects of Chinese accession to the WTO. Hertel and Winters (2006) is conceptually similar in its
estimates of the poverty implications of the Doha Round, but with up to three levels of modeling: a global multi-country
CGE model to calculate the effects of the Round on each country's prices of imports and export demand; more detailed
national CGE models for twelve country case studies to estimate the effects of these on local prices etc, and, in the cases
where the national models do not embed households directly, household modules to calculate the first order welfare
approximations by household.
         A further simplification is again to solve a CGE model with a single representative consumer, but now to
consider the effects of a shock only on a few summary statistics such as average incomes, unskilled wages and food
prices. Then, applying `poverty elasticities' to these statistics allows one to estimate the implied change in poverty. (The
poverty elasticity relates the proportionate change in poverty to the proportionate change in per capita GDP ­ see, for
example, Ravallion (1997)). This is the approach in Cline (2004), and Anderson et al. (2006) among others. These
studies differ inter alia in the base poverty levels to which they apply the elasticities.
         For purposes of this paper, we adopt a hybrid of the alternatives. For a global model of the size we have used to
explore the DDA, it is not computationally feasible to embed households or even many representative household groups
into the CGE model. And neither do we have the requisite data on factor earnings by household for the majority of
developing countries. However, we believe that the impact of trade reform on individual households will vary widely
depending on their primary sector of employment, their endowments, as well as their consumption patterns. Therefore
we reject the single poverty elasticity approach. Instead we utilize the factor earnings and income distribution data for
our 15 target developing countries, where this has been obtained and processed in a uniform manner, and we estimate
country-stratum-factor price-poverty line specific poverty elasticities. These elasticities embody information about the
shape of the income distribution as well as the composition of household earnings in the neighborhood of the poverty
line for key subgroups (strata) of the population.
         The specifics of our approach are as follows: For each of our 15 countries we have household surveys that
identify sources of income. As described in the Annex, we first adjust the earnings data in order to ensure that the
composition of factor incomes in agriculture match those reported in the National Accounts. (A comprehensive
reconciliation of these two, mutually inconsistent, sources of income would be a monumental, so we focus on the sector
most central to our analysis, which also happens to be the easiest to reconcile.) We then divide households into seven
strata, or groups of households, according to their predominant source of income and location: for five the criterion is
that 95% or more of income comes from the named source--agricultural self-employment, non-agricultural self-
employment, urban wages, rural wages, and transfers; for the remaining two we distinguish urban diversified and rural
diversified. Each stratum is then ordered by total income and divided into twenty vigntiles to create a total of 140
classes of households in each of the 15 focus countries.
         For each stratum we calculate poverty elasticities with respect to each source of income, where the shares of
income come from the vigntiles in which the poverty line falls, and the density of households around the poverty line
from the overall distribution. The density determines the change in headcount poverty resulting from a given change in
stratum income, while the earnings shares determine how a change in (e.g.) unskilled wages change stratum income in
the neighborhood of the poverty line. Once the CGE model has been solved for a new set of factor prices, we can use
these poverty elasticities to determine the change in stratum poverty ­ and, given the relative importance of stratum
poverty in national poverty ­ the change in the latter may also be obtained.
         Of course, a rise in factor earnings is only meaningful from a poverty perspective if commodity prices do not
also rise by the same amount. Therefore we must deflate the factor price changes by the change in the real cost of living
at the poverty line. This is obtained by solving the AIDADS demand system for the expenditure necessary to achieve
the poverty level of utility at the post-reform prices.



                                                             17

        So, for example, from the first entry in row 2 of Table 4, we see that a one percent

increase in unskilled wages in Peruvian agriculture reduces the $1/day poverty headcount

in the agriculture stratum by 1.41%. It also contributes to poverty reductions in the

diversified households. Indeed, the elasticity is slightly higher for urban diversified

households than for rural diversified ones, indicating that these households earn a non-

negligible share of their income from agriculture self-employment, despite their urban

status in the survey. Labor income is also dominant in the other strata, although in the

case of non-agriculture, it is non-agricultural labor, and so on. Note also that the non-

agriculture and wage-labor specialized households receive income from both skilled and

unskilled labor.

        Returning to the agriculture stratum poverty elasticities in the first column of

Table 4, we see that, in addition to unskilled labor, there are also small elasticities for

land, agriculture capital and transfers. If returns to all of these income sources were to

rise by one percent, then stratum income would rise by one percent for all households,

including the households at the poverty line. Therefore, the elasticities in Table 4 sum to

the same figure displayed in Table 3 for this particular stratum (compare row total with

Peru row in Table 3).

        As noted in Table 2, in addition to the agriculture stratum, the rural diversified

stratum is a very important repository for the poor in most of our focus countries. For this

reason, it is interesting to examine the poverty elasticities for this particular stratum

across the full range of focus countries. These are reported in Table 5. To facilitate

comparison across countries, we have normalized these elasticities, by dividing by their

total (e.g., 1.05 for the rural diversified households in Peru, Table 4). So the elements in



                                               18

each row represent the contribution of each endowment to the total poverty elasticity for

the rural diversified stratum in a given country. Clearly the composition of the aggregate

poverty elasticity for the rural diversified stratum varies considerably across countries ­

further evidence of the great variety of developing countries included in our sample.

        As expected, unskilled earnings are generally dominant in the rural diversified

households' earnings profile, with the type of earnings depending on the sector in which

the labor is employed. Land rents are generally unimportant for the poor, excepting in the

case of the Philippines, and, to a lesser degree, Uganda. Skilled labor also plays a small

role in earnings at the poverty line in these countries, and hence contributes little to the

poverty elasticities. Agriculture and non-agriculture capital plays a more important role in

some countries ­ most notably non-agriculture capital in Vietnam, where it accounts for

more than half of the poverty elasticity for the rural diversified households. Transfer

payments are quite significant at the poverty line in the wealthier countries ­ most

notably Brazil, Chile and Thailand, where they account for more than a third of the total

poverty elasticity for the rural diversified households.

        The income sources in Table 5 must be mapped to factor earnings in the general

equilibrium model. For example, agricultural labor and capital receive the corresponding

farm factor returns from the general equilibrium model, as do non-agricultural labor and

capital. Wage labor reported in the survey presents a problem, since we don't know how

much of this is employed in agriculture vs. non-agriculture activities. For this reason, we

simply assign to it the economy-wide average wage ­ a blend of the farm and non-farm

wages. Finally, transfer payments are indexed by the growth rate in net national income

(Annex V offers elaboration on this choice).



                                               19

       Of course our evaluation of household welfare depends not only on earnings, but

also on what happens to consumer prices. With food prices likely to rise in the wake of

rich country agricultural reforms, and with the poorest households potentially spending

the bulk of their income on food, this could have adverse consequences for poverty.

Therefore, we turn next to our treatment of consumer preferences.

       Household Preferences and Welfare: Given the emphasis in this paper on

household welfare ­ in both rich and poor countries ­ it is important that we pay close

attention to the specification of household preferences and the resulting pattern of

demands across the income spectrum. The approach used here follows closely that of

Hertel et al. (2004) insofar as we begin with an econometrically estimated, international,

cross-section demand system, which is then systematically adjusted to reproduce national

per capita demands. These national preferences are then used to predict demands across

the income spectrum within each country; in particular they are used to assess the impact

of consumer price changes on households at the poverty line in our 15 focus countries. In

the US, the national demand system is used to evaluate welfare for each of the farm

household groups discussed above.

       The demand system chosen for this task must be flexible enough to explain the

broad pattern of consumption in Malawi, on the one hand, and the United States on the

other. Accordingly, we follow Hertel et al.(2004) in using a demand system ­ nick-named

AIDADS -- which features highly non-linear Engel curves and has been shown to

perform very well in out-of-sample predictions of per capita international demand

behavior (Cranfield et al., 2003; see Annex IV for a detailed discussion). For our

purposes, the key feature is that the chosen demand system allocates two-thirds of its



                                             20

parameters to predicting behavior at extremely low income levels, which is what we need

to predict the consumption impacts on the poor. Estimation of this demand system is

undertaken using the 80-country, per capita consumption data set offered by GTAP,

version 6.1 and the resulting parameters are reported in Annex IV.

        The best way to understand the implications of this approach to estimation of

demands across the income spectrum is to view the results for a particular country. Figure

3 plots the predicted household budget shares for Peru, across the income spectrum.

These show how the pattern of consumer expenditures are predicted to vary from the

subsistence level (origin of horizontal axis), where expenditures on food and clothing are

dominant (budget share of nearly 60%), to the national per capita expenditure level where

the household budget is more diversified (the horizontal axis reports the natural logarithm

of consumption expenditure, per capita). Vertical lines denote the $1/day, $2/day and

national per capita expenditure levels. Note that at $1/day poverty line, 49% of the budget

is devoted to food, with the bulk of this spent on crops. The initial levels of utility at the

two poverty lines are each fixed, and the estimated demand system is used to determine

the change in the cost of attaining this exogenous poverty level of utility when prices and

demands change due to trade liberalization.

Policy Scenarios

        Our attention in this paper is on the distributional impacts of WTO reforms in

agriculture. Since such reforms are most contentious in the rich countries, we focus

initially on impacts from liberalizing agricultural policies in only the rich countries. The

OECD produces annual estimates of the producer support estimate for its member

countries. Rice is far and away the most protected commodity by this measure, with on



                                              21

average OECD rice producers receiving eighty-plus percent of their revenue as a result of

some policy intervention. Both sugar and milk producers in the OECD generate over

forty percent of their revenue from some combination of market intervention and direct

government support, while other grains and oilseeds lie below that level.

       Across countries, the producer support for OECD member countries varies widely

ranging from a low value of 1 percent of producer revenue in New Zealand to a high

value of 69 percent of producer revenue in Switzerland (Annex Table A.6.2). For the

OECD in aggregate, transfers to producers account for 31 percent of revenues. Producer

support in the EU is near the OECD-wide average. In Western Europe and East Asia

producer support is considerably above the OECD average, while that in North America

and Central and Eastern Europe is somewhat below. Australia and New Zealand provide

minimal support to producers through agricultural policies.

       The OECD producer support estimate is a combined measure of all support to

producers capturing the transfer of treasury monies paid to farmers as well as the

transfers from commodity transactions to producers resulting from prices supported

above world price levels. Thus this subsidy measure can be broadly decomposed into

market price support (i.e. policy measures applied at a country's border) and farm policy

transfers including output and input subsidies, area and headage based payments, and the

various payments tied to land use, farm income, and historical payments. The relative

importance of these differs across countries but in most instances the division between

market price and other support is roughly equal. The primary exception is in East Asia

(Japan and Korea) where producer support is nearly all provided as market price support.




                                            22

        The WTO separates support policies into three groups, with separate negotiating

modalities for each of them. Translating from the OECD producer support measure to the

WTO's aggregate measure of support framework is not straightforward. The market price

support component captures both the market access and export subsidy pillars of the

WTO agricultural negotiations. The remaining portion of the OECD measure poses a

significant challenge for quantifications in the context of the WTO domestic support

negotiations, as these are differentiated according to "traffic light" designations (amber,

blue, and green boxes) that intend to characterize the level of distortion created by a

particular policy implementation. This complexity of moving from the OECD's

comprehensive domestic support data base to the WTO domestic support framework is

the reason we draw on the published study by Jensen and Zobbe (2006) for our Doha

agricultural scenarios. These authors consider in detail not only the WTO designations of

support, but also the associated binding overhang versus actual support levels that we can

not evaluate by looking at the OECD producer support estimates in isolation.

        The Doha scenario considered in this paper derives from the so-called July 2004

Framework Agreement (WTO, 2004) as embodied in the core scenario from the Hertel

and Winters volume (2006) and is summarized, along with the other policy scenarios

considered in this paper, in Table 6. The first column of this table highlights the

implications for cuts in support in the rich countries' agricultural sectors ­ the main focus

of this paper. This Doha scenario assumes that industrial countries with domestic support

in excess of 20 percent of production cut their bound commitments by 75 percent, while

others cut by 60 percent. However, even with these ambitious reductions, the gap

between bindings and applied policies, as well as the inclusion of market price support



                                              23

concepts mean that effectively only five WTO members would be required to reduce

actual support, based on 2001 notifications: Australia, EU, Iceland, Norway, and US

(Jensen and Zobbe, 2006). Export subsidies are the one area where bold cuts (full

elimination) are on the table, and we assume this outcome in our Doha scenario. When it

comes to developing countries (see column three) domestic subsidy bindings are cut by

40 percent. In this case, Jensen and Zobbe (2006) estimate that only Thailand's subsidies

would be affected.

       Agricultural tariffs in the rich countries are reduced using a tiered formula, with

marginal cuts changing at 15 and 90 percent bound tariff rates. The marginal cuts are 45

percent on the first 15 percentage points of the tariff, 70 percent for the range between 15

and 90 percent, and 75 percent on the remainder.6 For developing countries, the inflection

points are placed at 20, 60 and 120 percent bound tariff levels in agriculture, with

marginal cuts of 35, 40, 50 and 60 percent, respectively.

       Of course, cross-sector trade-offs are at the heart of the WTO negotiations, so we

also consider the impact of non-agricultural elements of a prospective Doha Development

Agenda on both rich and poor countries. Given the importance of non-agricultural

income to farm households in many of the rich countries, this also could have a direct

bearing on farm household welfare. In the case of poverty impacts in developing

countries, improved access to rich country manufactures markets, as well as access to the

markets of other developing countries can have an important impact on the demand for

unskilled labor, and hence poverty rates.



6For example, a tariff of, say, 100% is cut by 66.95%: = [15%*0.45 + (90-15)%*0.70 + (100-90)%*0.75].
By applying the cuts at the margin we avoid the discontinuities implied by the July Framework.



                                                   24

         Following Hertel and Winters (2006), we focus the attention of our non-

agricultural shocks on market access (see column 3 of Table 6), since barriers to services

trade and investment remain difficult to quantify and those WTO negotiations appear

unlikely to yield significant changes in the near term. Specifically, non-agriculture tariffs

are subjected to proportional cuts of 50 percent for developed and 33 percent for

developing countries. The Least Developed Countries are not required to cut tariffs under

this central scenario (see Anderson and Martin, 2006). As a consequence of these

relatively ambitious tariff cuts in both farm and non-farm trade, average world-wide

tariffs for all merchandise trade drop from 4.7% in the baseline to 3.2%.

         In order to establish a benchmark set of liberalization results from which to make

comparisons, we begin by examining the distributional consequences of the complete

elimination of rich country support for agriculture. We then consider the portion of this

impact that would be delivered under the particular Doha scenario discussed above. After

this, we add, in turn, non-agricultural reforms in the rich countries, and liberalization in

the developing countries (agricultural and non-agricultural).

         Finally, we consider the likely scenario that governments in rich regions will opt

to compensate adversely affected farm households through WTO green-box means.

These green-box payments are tied to land use, not output, and are designed to be neutral

across farm products (i.e. the subsidy is not contingent of a specific use of the land). As

such they generate minimal distortions in world markets and so are in line with WTO

guidelines as their primary effect is simply the transfer of income from taxpayers

(including farmers) to farmers.




                                             25

        Throughout our analysis, we employ a macroeconomic closure which fixes the

ratios of government spending, tax revenue, net national savings, and the trade balance,

all relative to net national income. This (relatively standard) closure facilitates linking the

aggregate and disaggregate welfare impacts of trade reform (see the Annex for an

extended discussion of our closure assumptions and their implications).

Results

        Agriculture Liberalization by the Rich Economies: Before discussing the farm

household impacts, we consider briefly the macro-economic impacts of these policies.

Complete liberalization of rich country farm policies generates some very large trade

volume increases for rice, sugar and beef products where border protection is dominant,

whereas world trade in coarse grains and cotton actually falls, as domestic subsidies are

eliminated and rich country exports are reduced. Under the Doha scenario, which

emphasizes trade volume-reducing export subsidy elimination, as opposed to trade-

increasing tariff reductions, the global trade volumes for wheat and dairy products also

fall. Details are available in Annex Table A.6.3.

        We turn next to the national, macro-economic impacts of the reforms. Table 7

reports two key national indicators for the rich countries and our 15 focus developing

countries (results for other regions are reported in the annex). These variables are: the

percentage change in the regional terms of trade (an index of export prices, relative to

import prices), and the percentage change in real aggregate consumption (national

welfare derived from the consumption of private goods and services). We see from the

first two pairs of columns (Rich-Agr-Full and Rich-Agr-Doha), that agricultural

liberalization is good for the rich countries (welfare rises). Furthermore, these changes



                                              26

are statistically significant, once we take into account the parametric uncertainty

associated with our CGE model. (See the footnote to Table 7 as well as Annex I.) The

fact that reform of this highly distorted sector will benefit the rich countries should come

as little surprise, and it is well-established in the literature (Anderson, Martin and van der

Mensbrugghe, 2006; Francois et al., 2005; Dimaranan, Hertel and Keeney, 2004). The

roadblock to agricultural reform has to do with the concentration of losses among key

interest groups ­ a point to which we will turn shortly. Note also that the Doha reforms

capture a significant share of total available gains to Europe under full agricultural

reform, and a little under half in other rich countries.

         A somewhat more controversial point has to do with the impact of rich country

agricultural reforms on the developing countries. Here, the key mechanism for

transmission for economic welfare is through the terms of trade (ToT). If a country is a

net importer of food products and the world price of food products rises, then the ToT

might be expected to deteriorate. This is the case of Bangladesh, for example, which,

according to Table 7, experiences a 0.58% ToT deterioration under Rich-Agr-Full

liberalization, and a 0.21% ToT decline under the Rich-Agr-Doha scenario. This is

primarily due to higher prices for cotton, wheat and oilseeds. With a deteriorating ToT,

Bangladesh can afford fewer imports for a given amount of exports, and real

consumption is expected to decline. On the other hand, Brazil, with a 5.48% ToT

appreciation, can now consume more imports, or export less and consume more domestic

production, so its welfare rises.

         Of course, the story is a bit more complex for two reasons. First of all, in a world

of differentiated products, there is no single "world price" for a good. Even a commodity



                                                27

like rice is differentiated and many different prices can co-exist in the world market at

one point in time. So it can matter whether you source your rice from a country whose

price is rising, for example due to the elimination of an export subsidy. This is the case

with dairy imports into Venezuela from the EU and US. Venezuela also suffers from

higher import prices for manufactures from Brazil, since the latter country experiences a

real appreciation. In short, Venezuela is an example of a country that experiences ToT

and consumption losses due to its specific pattern of imports. (A full decomposition of

the ToT results is available in Annex Table A.6.5.) Overall, we find that the ToT

deteriorate in 7 of the 15 focus countries in the case of full agricultural reform in the rich

countries, with the number being somewhat larger (9 of 15) in the case of the Doha

reforms. The latter result follows from the greater emphasis of Doha on export subsidies

as opposed to market access.

        The second complication to the simple "ToT drive welfare" story described above

arises due to the presence of domestic tax and subsidy distortions. Note in particular, that

in the case of the Philippines and Tanzania (Rich-Agr-Full) and Peru (Rich-Agr-Doha),

the ToT improve, but welfare falls. This stems from fact that all three countries have

domestic tax policies that favor agriculture, relative to industry. Therefore an expansion

of agriculture at the expense of industry has an adverse effect on economic efficiency and

overall welfare.

        Now let us turn to the distributional results of rich country agricultural reforms.

Table 8 reports the percentage change in real on-farm income and off-farm household

income, as well as the implied change in real household income for the aggregate farm

household in each of the rich economies. From the on-farm income results, it is clear why



                                              28

there is so much opposition to these reforms. The average decline in Japan is 16 % under

the Doha scenario and 28 % under the Full Liberalization scenario and 6% and 13%

respectively in the EU. On-farm income losses in the US are much smaller ­ indeed they

are negligible under the Doha scenario, while Canadian and Australia/New Zealand

producers see gains in real on-farm income.

        However, as noted above, farm households in many of these countries are quite

diversified in their earnings. If we factor in the change in real, off-farm income, which

tends to rise (albeit modestly, since there are no reforms outside of agriculture), the total

impact on real farm household income is considerably moderated. Indeed, in Japan, the

losses drop by a full order of magnitude ­ from -16 % to just -1.4 % under the Doha

scenario. In the US, the losses become negligible, even under full liberalization. The

dampening factor is less prevalent in Europe, where the role of off-farm income is much

more modest than in Japan and US.

        Given the very modest aggregate farm household losses in the US, the question

arises: Why is the opposition to reform so strong? This becomes quite clear when we

delve more deeply into the US impacts, using the farm household survey data and

associated methodology outlined above. Table 9 reports the welfare impacts on

representative households in each of the 11 wealth classes across the five US producer

groups. It is clear that under the Rich-Agr Full Liberalization scenario, the losses to the

richest, and likely most influential, producer groups are very large­ nearly 20% of

income in the case of the wealthiest rice producers. The wealthiest sugar producers are

also hard-hit, as are cotton producers across the board. The surprising thing about the

results in Table 9 is the impact on rice producers under the Doha scenario. Here, they



                                               29

switch from being the biggest losers to the biggest gainers (based on this 5-way producer

grouping)!

        To further investigate this result we have performed a decomposition (using the

methodology of Harrison, Horridge and Pearson, 1999) that separately identifies the

partial impact of US rice reforms, US non-rice reforms, Japanese rice reforms, and the

residual category of all other agricultural reforms on US Farm Household welfare.

Results (available in annex Table A.6.6) show that the US agricultural reforms contribute

negatively to rice producer welfare. The initial level of support for rice production is very

high and even the modest reduction of the Doha scenario would generate an average real

income loss of -4.5% for rice producers if applied in isolation. Other US agricultural

reforms have lower but significant impact (-2% average income change) since rice

households lose support on any other crops they might produce and non-rice reforms

lower returns to labor and capital in agriculture. Therefore, the positive Doha welfare

impact derives from non-US policy reforms.

        The US rice producer gains under Rich-Agr reforms are dominated by the gains

owing to increased access to the lucrative Japanese market. Japanese rice protection cuts

increases average rice producer welfare by 8 percent, with the average contribution of

other countries liberalizing adding an additional one percent. So US producers gain under

Rich country reforms, following the Doha Agenda, since their cuts in domestic support

are modest (28%), while the improvement in market access to Japan is substantial. Of

course, Japanese negotiators will strive to have rice treated as a sensitive product, thereby




                                              30

limiting the increase in market access, and this will obviously limit the final gains under

any agreement.7

        Having considered the impact of rich country agricultural reforms on farm

households in the rich economies, we now turn to the impact of these reforms on the

poorest farm households in the some of the poorest countries in the world. As noted

previously, we do this via a set of disaggregated poverty elasticities -- each of which

relates to one of the income sources for the poor. We focus our analysis on the Rich-Agr-

Full liberalization results, subsequently comparing these to the Doha impacts.

        Table 10 reports the change in cost-of-living deflated factor returns, by country

under the $1/day poverty line assumption. With the exception of Uganda, which is the

only focus country to experience a real depreciation in the face of rich country

agricultural liberalization,8 these returns rise for all agricultural factors in all regions ­ a

simple consequence of the higher world prices for farm products. The biggest increases

are in land prices (the least mobile factor of production) ­ with very substantial increases

(from 15 ­ 39%) in Brazil, Mexico, Peru and Thailand. This is followed by unskilled

agricultural labor and capital. Note that the poverty-deflated earnings fall for

nonagricultural labor and capital in most countries. This will translate into higher poverty

rates for the self-employed, non-agriculture households. However, the economy-wide

average wage for unskilled labor rises in Brazil, Chile, Malawi, Peru, Philippines,


7 Anderson and Martin (2005) provide a systematic analysis of the case in which sensitive and special
commodities are exempted from steep tariff cuts, facing instead a modest 15% cut in bound rates (the Doha
scenario considered in this paper). In the case where just 2% of industrial country tariff lines and 4% of
developing country tariff lines in agriculture are exempted, the overall average tariff cuts are greatly
reduced. Furthermore, Anderson, Martin and van der Mensbrugghe (2005) find that such exemptions erase
any potential for poverty reduction under our Doha scenario.




                                                   31

Thailand and Vietnam, so that modest poverty reductions in the wage-labor households

are expected. The final column of Table 10 shows that transfers, which are assumed to be

indexed by net national income, generally do not rise fast enough to offset the higher cost

of living at the poverty line. So we expect poverty in the transfer strata to rise.

         Table 11 reports the consequent changes in $1/day poverty, by stratum. As

expected, poverty rates in the agriculture stratum fall in all countries, excepting Uganda.

Due to its relatively higher poverty elasticities, the largest percentage reductions in

poverty are in Thailand. However, there are also double-digit percentage reductions in

poverty among the self-employed agricultural households in Brazil, Chile, and Peru.

Clearly the same policies that bolster incomes of the richest farm households in US, and

other developed countries, have an adverse impact on the poorest farm households in

some of the poorest countries in the world. The removal of these policies serves to reduce

poverty incidence among poor country farm households. The diversified household strata

(both urban and rural) also show substantial poverty reductions in a number of cases ­

particularly Brazil, Chile and Thailand. On the other hand, higher food prices consistently

push more of the non-agriculture, self-employed and the transfer dependent households

into poverty.

         The net effect of Rich-Agr Liberalization on the national poverty headcount is

reported in the first set of columns in Table 12. National poverty at the $1/day level falls

in 10 of the 15 countries, with small percentage increases in Mozambique (unskilled

wages fall), Uganda (factor prices fall), Venezuela (high share of poor in the non-




8In the case of Uganda, the impact of preference erosion in the EU market is particularly severe.




                                                   32

agriculture stratum), Vietnam (large poverty elasticity for non-agricultural capital) and

Zambia (negligible poverty elasticity in agriculture stratum). The next column of Table

12 converts these percentage changes into thousands of people. Here, the reductions in

Brazil, Indonesia, Philippines and Thailand are clearly dominant. When we move to the

$2/day of Rich-Agr Lib (next two columns of the table), the national poverty picture is

reversed in two cases: Bangladesh (small decrease becomes a small increase) and

Vietnam (small insignificant increase becomes a small decrease), so once again poverty

falls in two-thirds of the 15 countries. On balance, the largest changes involve poverty

reductions, with Brazil, Indonesia, Philippines and Thailand standing out.

         We can contrast these outcomes to those that would be achieved under the

prospective DDA (Rich-Agr reforms only), and this is done in the final four columns of

Table 12. More modest rises in agriculture earnings and lesser increases in the unskilled

wage rate (adjusted for the cost of living at the poverty line) means that now poverty rises

(albeit slightly) in more than half the countries (8 of 15) in the case of $1/day poverty.

Clearly even the ambitious Doha Development Agenda under examination here is less

poverty friendly than would be a proportionately scaled back version of full liberalization

in rich country agriculture. The latter would presumably show poverty reduction in all the

same countries ­ just to a lesser degree. Yet the Doha scenario results in fewer countries

showing poverty reductions than under the full liberalization of Rich Agriculture.

         Global Liberalization Scenarios: We now turn to a set of liberalization scenarios

that involve tariff cuts in both agriculture and non-agriculture sectors and in both the rich

and the poor countries. Developing country agricultural tariffs are quite high, so

abolishing them increases world agricultural trade volumes relative to rich-only



                                              33

liberalization. Reforming them on Doha terms, however, makes little difference because

the large binding overhangs and modest cuts in developing country bound tariffs (no cuts

for LDCs) translate into little additional market access. Adding tariff cuts in

manufactures on the other hand leads to significant increases in manufacturing trade

under both full and Doha scenarios and for both developed and developing countries (see

Annex table A.6.3 for detailed results).

        Returning to Table 7, we consider now the second group of columns reporting the

aggregate welfare and terms of trade impacts of the global reforms. Comparing Rich

Agriculture with Global reforms, the most striking change in the rich countries is the

improvement in the terms of trade for Japan, which benefits from manufacturing tariff

cuts. On the other hand, the Canadian terms of trade deteriorate more as a result of

preference erosion in the US manufactures market. However, despite this terms of trade

loss, Canadian welfare rises by more under global full liberalization than under Rich-Agr

liberalization alone.

        Turning to the focus countries, we see very different terms of trade and welfare

impacts than those stemming from Rich Agriculture reforms only. The terms of trade for

these developing countries fall in more cases (9 of 15 cases), due to the expansion of poor

country exports in the wake of own and other developing country tariff cuts and the

erosion of preferences in manufacturing. However, welfare only falls for six of these

countries, with efficiency gains dominating the ToT losses in the other three cases

(Philippines, Vietnam and Zambia). In contrast, under Global Doha, there are fewer ToT

losses, but also fewer (and smaller) welfare gains. These mixed aggregate welfare effects

for developing countries from global trade reforms are quite comparable to those reported



                                             34

in other studies of the aggregate impacts of global trade reforms on developing countries

(Francois et al., Anderson and Martin, Hertel and Winters).

        The changes in real farm income under global reforms (Annex Table A.6.7) ­

both full liberalization and Doha reforms ­ are dominated by the Rich-Agr reforms

previously discussed. Liberalizing rich country non-agricultural merchandise trade is

slightly beneficial to the farm households ­ by lowering the price of non-agricultural

goods, but tariffs on most of these products are already quite low and so the impact is

minimal. On the other hand, trade reforms in the poor countries as a group tend to be

slightly adverse for the welfare of rich country farm households. This is due to a complex

set of factors, including the tendency for tariff cuts to encourage labor and capital to shift

back to the food and agriculture sector, as well as the impact of increased demand on the

general price level in rich countries. But these effects are very small, relative to the

primary impact of the Rich-Agr policies themselves.

        Given these results for the average farm household in the rich countries, it is

hardly surprising that the impacts of global reforms on individual US farm households

are quite similar to that reported previously in the Rich-Agr reform scenario (Table

A.6.8). Welfare for the wealthiest farm households is driven first and foremost by their

own national policies, with the largest international interactions occurring among the

world's richest (and largest) markets ­ as in the case of US-Japan rice trade.

        However, when it comes to the poverty impacts of global trade reform,

agricultural policies in the rich countries are only part of the story ­ trade policies in the

developing countries themselves assume much greater prominence. Table 13 reports the

impacts of global trade reform on national poverty in our 15 focus countries. We



                                              35

decompose these effects by the broad set of policies being reformed using the technique

of Harrison, Horridge, and Pearson (1999). As noted previously, Rich-Agr reforms

contribute to poverty reduction in the majority of countries. However the most striking

thing about the results in Table 13 is the uniformly favorable impact of food and

agriculture tariff cuts in the poor countries on poverty (column Poor-Agr). The impact of

non-agriculture reforms in both rich and poor countries is generally of lesser magnitude,

and mixed in sign.

        Table 14 provides a comprehensive picture of the poverty impacts of global

reforms, including those implied by the Doha scenario. Here, we see that the latter is

indeed less poverty friendly than the global liberalization scenario, with poverty rising in

8 of the 15 countries, as opposed to just 6 in the full liberalization case. Hertel and Ivanic

(2006) emphasize that this is due to the heavy weight given to export subsidy elimination

(which raises import prices for food), while the developing countries make only mild cuts

to their applied tariffs under the Doha scenario and the least developed countries are not

required to cut tariffs at all.

        Compensation for Rich Country Farmers: The farm household welfare impacts in

rich countries are dominated by liberalization of the agricultural pillars. Inclusion of

agricultural and developing country reforms do little to make-up the lost income when we

broaden the scope of reform. With this in mind we consider a final scenario that ensures

that aggregate farm income is left unchanged under the global full liberalization

experiment. This requires solving for an endogenous green-box subsidy to land in the

following rich regions where aggregate farm income declines: Japan (-28.4%), Europe

(-11.5%), and the US (-3.7%). The choice of aggregate farm income as a compensation



                                              36

target reflects the expectation that the policy process will continue to focus on this readily

available measure to gauge the well-being of the farm population. In an alternative

compensation simulation we investigate the cost savings generated in these three

countries by compensating on the basis of aggregate farm household income (inclusive of

off-farm income).

        Farm income compensation at the level of a representative farm household in each

of these countries leads to sizable increases in WTO green-box outlays in each country.

In Japan, agricultural land is subject to net taxation in the initial situation, and

compensation requires a change to net subsidization at the level of $9.1 billion in land-

based payments. Both the EU and US have significant land-based payments initially and

the compensation scheme here indicates that the EU would need a 63 percent increase

over that initial level at a cost of $11.8 billion. For the US, the percentage increase is

smaller at 27.4%, coming at a cost of $3.3 billion.

        As discussed previously, the use of farm income as a welfare indicator for the

population of farm households in wealthy countries is incomplete and in this case would

lead to considerable over-compensation in welfare terms. Using the full farm household

welfare criterion as opposed to solely farm income, we find that Japanese and US policy-

makers need only compensate these farmers with $6.3 billion and $2.4 billion dollars,

nearly a one-third reduction. The reduction in the European Union is much smaller (only

$300 million less than when compensating based on losses in farm income alone). This

follows directly from the relatively large share of income from farming in the European

Union (60 percent) as well as the less favorable developments in off-farm income.

Conclusions



                                                37

       This paper has sought to identify the impacts of WTO reforms on farm

households in rich and poor countries. It has done so via innovative use of newly

available household survey data that identify the income sources and degree of earnings

specialization of households. This proves to be a critical factor in assessing the household

welfare impacts of trade reforms. In the rich countries, we focus our attention on the

United States, where survey data permit us to assess the impacts of trade reforms by

wealth decile and commodity specialization. In the poor countries, we focus our attention

on the poverty headcount ­ among both farm and non-farm households.

       Our findings highlight the fact that wealthy farmers are the main beneficiaries of

current trade policies aimed at protecting agriculture in the rich countries. Furthermore,

these benefits tend to be concentrated in a few products that receive very high levels of

support presently. In the United States, rice stands out ­ followed by cotton, sugar and

dairy. When we look at aggregate farm household welfare in the United States, it is little

affected by agricultural trade policy reforms. This is because many of the farm products

receive little or no support and improved market access in other countries benefits export-

oriented producers. Indeed, this is why the average farm household in Australia, Canada

and New Zealand is expected to gain from rich country agricultural trade reforms. A

second reason why the average farm household in the US is not more severely affected

by trade policy reform stems from the degree of earnings diversification in that country.

On average, only 8% of farm household income in the US is derived from farming. This

income diversification is also critical in Japan where just 12% of farm income is obtained

from on-farm earnings. As a consequence, while Doha trade reforms cause on-farm

incomes to drop by 16% in Japan, the average farm household impact is just 1.4%.



                                              38

       This finding of generally modest impacts on the average farm household stands in

sharp contrast to the strong opposition from agricultural lobbies in the rich countries. This

opposition can be better understood when we use our household survey data for the US to

show that the degree of earnings diversification diminishes for the wealthiest farms in the

highly protected commodities, and this provides them with strong incentives to prevent

the very substantial drop in household welfare that can be expected under trade reform.

Consequently, we can expect that some compensation mechanism is necessary to solve

the political impasse currently plaguing the Doha talks. We explore one such mechanism

by which payments are aimed at neutralizing the loss in average on-farm income for each

commodity group. This program would introduce around $25 billion of new agricultural

subsidies into global agriculture from the three countries where farm income declines

(Japan, EU, and US) and would likely make the Doha scenario much more palatable to

the farm lobbies.

       In the poorest countries, we find that, with one minor exception, rich country

agriculture reforms benefit low-income farm households. Regardless of the poverty line

considered, the poverty headcount in this part of the developing world falls. However, the

impact on non-farm population groups is mixed. In those countries where agriculture

makes up a large share of the unskilled labor force, rich country reforms tend to increase

the demand for labor sufficiently to benefit unskilled workers throughout the economy.

However, self-employed households in the non-agricultural economy, as well as those

dependent on transfer payments, systematically lose. Therefore the national poverty

outcome inevitably depends on the relative weights of these different groups in the

national poverty picture. Since a large share of the poor reside in agriculture, national



                                             39

poverty falls in two-thirds of the focus countries in the wake of rich country agricultural

liberalization.

        Finally, we conclude that, if policy makers are serious about poverty reduction,

they should push for more poor country farm and food tariff cuts, as these products loom

large in the household budgets of the poor. Giving them access to food at world market

prices (adjusted for marketing margins) is a sure way to reduce poverty. Yet this is

precisely the component that is mostly omitted under the current Doha proposals. Indeed,

global trade liberalization is the policy configuration with the most favorable poverty

outcomes in the 15 developing countries examined in this study.




                                              40

References

Abler, D.G. 2001. "Elasticities of Substitution and Factor Supply in Canadian, Mexican,
       and U.S. Agriculture," Appendix A.2 in Market Effects of Crop Support
       Measures, OECD, Paris.

Adelman, I., and S. Robinson. 1978. Income Distribution Policy in Developing
       Countries: A Case Study of Korea, Oxford University Press and The World Bank,
       Oxford and New York.

Anderson, K. 1995. "Lobbying Incentives and the Pattern of Protection in Rich and Poor
       Countries," Economic Development and Cultural Change, 43: 401-424.

Anderson, K., and W. Martin (eds). 2006. Agricultural Trade Reform and the Doha
       Development Agenda, Palgrave Macmillan and The World Bank, New York.

Anderson, K., W. Martin, and D. van der Mensbrugghe. 2006. "Market and Welfare
       Implications of Doha Reform Scenarios," Chapter 12 in Agricultural Trade
       Reform and the Doha Development Agenda, K. Anderson and W. Martin eds.,
       Palgrave Macmillan and The World Bank, New York.

Anderson, K., W. Martin, and D. van der Mesbrugghe. 2006. "Global Impacts of the
       Doha Scenarios on Poverty," Chapter 17 in Poverty and the WTO: Impacts of the
       Doha Development Agenda, T.W. Hertel and L.A. Winters (eds), Palgrave
       Macmillan and The World Bank, New York.

Bourguignon, F., and L.A. Pereira da Silva. 2003. The Impact of Economic Policies on
       Poverty and Income Distribution: Evaluation Techniques and Tools, World Bank
       and Oxford University Press, Oxford and New York.

Chen, S., and M. Ravallion. 2004."Welfare Impacts of WTO Accession to the World
       Trade Organization," World Bank Economic Review 18: 29-57.

Cline, W.R. 2004. Trade Policy and Poverty. Institute for International Economics,
       Washington, DC.

Cranfield, J.A.L., J.S. Eales, T.W. Hertel, and P.V. Preckel. 2003. "Model Selection
       when Estimating and Predicting Consumer Demand Using International, Cross-
       Section Data," Empirical Economics 28: 353-64.

Cranfield, J. A. L., P. V. Preckel, J. S. Eales, and T. W. Hertel. 2004. "Simultaneous
       Estimation of an Implicit Directly Additive Demand System and the Distribution
       of Expenditure--an Application of Maximum Entropy." Economic Modeling, 21:
       361­385.




                                              41

de Janvry, A., M. Fafchamps, and E. Sadoulet. 1991. "Peasant Household Behavior with
       Missing Markets: Some Paradoxes Explained," Economic Journal, 101: 1400-
       1417.

Emini, C.A., J. Cockburn, and B. Decaluwe. 2006. "The Poverty Impacts of the Doha
       Round in Cameroon: The Role of Tax Policy," Chapter 13 in Poverty and the
       WTO: Impacts of the Doha Development Agenda, T.W. Hertel and L.A. Winters
       (eds), Palgrave Macmillan and The World Bank, New York.

Finger, J. Michael. 1981. "Policy Research," Journal of Political Economy 89: 1270-71.

Francois, J., H. van Meijl, and F. van Tongeren. 2005. "Trade Liberalization in the Doha
       Development Round," Economic Policy, 20: 349-391.

Gardner, B.L. 1969. "Determinants of Farm Family Income Inequality," American
       Journal of Agricultural Economics, 51: 753-769.

Gardner, B.L. 1992. "Changing Economic Perspectives on the Farm Problem," Journal of
       Economic Literature, 30: 62-101.

Gardner, B.L. 2000. "Economic Growth and Low Incomes in Agriculture," American
       Journal of Agricultural Economics, 82: 1059-1074.

Golub, A., T.W. Hertel, H. Lee and N. Ramankutty. 2006. "Modeling Land Supply and
       Demand in the Long Run," paper presented at the Ninth Annual Conference on
       Global Economic Analysis, Addis Ababa, June 15-17, also available at:
       https://www.gtap.agecon.purdue.edu/resources/download/2628.pdf

Hanson, K. and A. Somwaru. 2003. Distributional Effects of U.S. Farm Commodity
        Programs: Accounting for Farm and Non-farm Households. Paper presented at
        the 6th Annual Conference on Global Economic Analysis, June, The Hague, The
        Netherlands.

Harrison, G., T. Rutherford, and D. Tarr. 2002. "Trade Policy Options for Chile: The
       Importance of Market Access," World Bank Economic Review 16: 49-79.

Harrison, G., T. Rutherford, D. Tarr and Gurgel, A. (2003) Regional, multilateral and
       unilateral trade policies of MERCOSUR for growth and poverty reduction in
       Brazil. Policy Research Working Paper 3051. World Bank, Washington, DC.

Hertel, T.W. and M. Ivanic. 2006. "Understanding the Poverty Implications of the Doha
       Development Agenda," mimeo, Center for Global Trade Analysis, Purdue
       University, W. Lafayette, IN.




                                           42

Hertel, T.W., M. Ivanic, P.V. Preckel, and J.A.L. Cranfield. 2004. "The Earnings Effects
        of Multilateral Trade Liberalization: Implications for Poverty," World Bank
        Economic Review 18: 205-36.

Hertel, T.W., and R. Keeney. 2006. "What is at Stake: The Relative Importance of Import
        Barriers, Export Subsidies, and Domestic Support," Chapter 2 in Agricultural
        Trade Reform and the Doha Development Agenda, K. Anderson and W. Martin
        eds., Palgrave Macmillan and The World Bank, New York.

Hertel, T.W., and J.J. Reimer. 2005. "Predicting the Poverty Impacts of Trade Reform,"
        Journal of International Trade and Economic Development, 14: 377-405.

Hertel, T.W., and L.A. Winters (eds). 2006. Poverty and the WTO, Palgrave Macmillan
        and The World Bank, New York.

Hill, B. 1996. Farm Incomes, Wealth, and Agricultural Policy, Ashgate Publishing Co.,
        Brookfield, Vermont.

Jensen, H.T., and F. Tarp. 2005. "Trade Liberalization and Spatial Inequality: A
        Methodological Innovation in a Vietnamese Perspective," Review of Development
        Economics 9:69-86.

Jensen, H.G., and H. Zobbe. 2006. "Consequences of Reducing Limits on Aggregate
        Measurements of Support," Chapter 9 in Agricultural Trade Reform and the Doha
        Development Agenda, K. Anderson and W. Martin eds., Palgrave Macmillan and
        The World Bank, New York.

Keeney, R. 2005. Decoupling and the WTO: Farm Sector and Household Impacts in the
        United States, Unpublished Ph.D. Dissertation, Purdue University.

Keeney, R., and T.W. Hertel. 2005. GTAP-AGR: A Framework for Assessing the
        Implications of Multilateral Changes in Agricultural Policies, GTAP Technical
        Paper No. 24, Center for Global Trade Analysis, Purdue University. Available at:
        https://www.gtap.agecon.purdue.edu/resources/tech_papers.asp

Kilkenny, M. 1993. "Rural/Urban Effects of Terminating Farm Subsidies," American
        Journal of Agricultural Economics, 75: 968-980.

Organization for Economic Cooperation and Development (OECD).2001. Market Effects
        of Crop Support Measures, OECD Publications, Paris.

Organization for Economic Cooperation and Development (OECD). 2003. Farm
        Household Income: Issues and Policy Responses, OECD Publications, Paris.




                                            43

Polaski, S. 2006. Winners and Losers: Impact of the Doha Development Round on
       Developing Countries, Carnegie Foundation, Washington, DC.

Ravallion, M. 1997. "Can High-Inequality Developing Countries Escape Absolute
       Poverty?" Economics Letters 56: 51-57.

Reimer, J.J., and T.W. Hertel. 2004. "Estimation of International Demand Behavior for
       Use with Input-Output Based Data," Economic Systems Research, 16: 347-366.

Rimmer, M., and A. Powell. 1996. "An Implicitly Additive Demand System," Applied
       Economics, 28: 1613-1622.

Rutherford, T., D. Tarr and O. Shepotylo. 2006. "The Impact on Russia of WTO
       Accession and the DDA: The Importance of Liberalization of Barriers against
       Foreign Direct Investment in Services for Growth and Poverty Reduction,"
       Chapter 16 in Poverty and the WTO: Impacts of the Doha Development Agenda,
       T.W. Hertel and L.A. Winters (eds), Palgrave Macmillan and The World Bank,
       New York.

Salhofer, K. 2001. "Elasticities of Substitution and Factor Supply Elasticities in
       European Agriculture: A Review of Past Studies," Annex A.2 in Market Effects of
       Crop Support Measures, OECD Publications, Paris.

Schattschneider, E.E. 1935. Politics, Pressure, and the Tariff, Prentice Hall, New York.

Schultz, T.W. 1945. Agriculture in an Unstable Economy. McGraw Hill Press, New
       York.

Surry, Y. 1990. "Econometric Modeling of the European Compound Feed Sector: An
       Application to France," Journal of Agricultural Economics, 41: 404-421.

       US Department of Agriculture-Economic Research Service. 2005. Agricultural
       Resource Management Survey Briefing Room. USDA-ERS Website.
       http://www.ers.usda.gov/Briefing/ARMS/

Valenzuela, E., T.W. Hertel, R. Keeney, and J.J. Reimer. 2005. Assessing Global CGE
       Model Validity Using Agricultural Price Volatility, GTAP Working Paper No. 33,
       Center for Global Trade Analysis, Purdue University, Available at:
       https://www.gtap.agecon.purdue.edu/resources/working_papers.asp

Van der Mensbrugghe, D. 2006. "Estimating the Benefits: Why Numbers Change,"
       Chapter 4 in Trade, Doha, and Development: Window into the Issues, R.
       Newfarmer (ed), The World Bank, Washington DC.




                                            44

Winters, L.A. 1987. "The Political Economy of the Agricultural Policy of Industrial
      Countries, " European Review of Agricultural Economics, 14: 285-304.

Winters, L.A. 2002. "Trade Liberalisation and Poverty: What are the Links?" World
      Economy, 25: 143-171.

Winters, L.A., N. McCulloch, and A. McKay. 2004. "Trade Liberalization and Poverty:
      The Evidence So Far," Journal of Economic Literature, 27: 481-506.

The World Bank. 2001. Global Economic Prospects, The World Bank, Washington DC.

The World Bank. 2003. Global Economic Prospects, The World Bank, Washington DC.

World Trade Organization (WTO). 2004. Decision Adopted by the General Council on 1
      August 2004, WT/L/579, WTO Secretariat, Geneva.




                                          45

Table 1. US Farm Income Shares by Household Type and Wealth Group
Wealth Group      Rice       Sugar       Cotton    Dairy    Other
10-%ile              0.39        0.22        0.67     0.56     -0.01
20-%ile              0.39        0.22        0.67     0.47     0.03
30-%ile              0.58        0.22        0.75     0.72     0.01
40-%ile              0.58        0.78        0.64     0.48     -0.01
50-%ile              0.84        0.78        0.82     0.59     0.07
60-%ile              0.55        0.78        0.59     0.57     0.07
70-%ile              0.76        0.31        0.64     0.71     0.11
80-%ile              0.80        0.31        0.63     0.61     0.12
90-%ile              0.75        0.66        0.83     0.81     0.20
95-%ile              0.74        0.91        0.68     0.83     0.21
100-%ile             0.74        0.91        0.91     0.92     0.41
Source: 2004 USDA-ERS ARMS.



Table 2. Stratum Contributions to the $1/day Poverty Population in each Country
                                                         Strata

Country
                 Agriculture        Non-       Urban   Rural             Urban   Rural
                                 Agriculture   Labor   Labor    Transfer Diverse Diverse Total

Bangladesh               0.15            0.13    0.04    0.22       0.03    0.07    0.37  1.00
Brazil                   0.14            0.09    0.24    0.15       0.32    0.04    0.03  1.00
Chile                    0.26            0.01    0.09    0.09       0.28    0.15    0.12  1.00
Colombia                 0.28            0.43    0.03    0.04       0.12    0.05    0.04  1.00
Indonesia                0.42            0.12    0.02    0.07       0.04    0.06    0.28  1.00
Malawi                   0.54            0.11    0.00    0.03       0.07    0.01    0.25  1.00
Mexico                   0.05            0.06    0.05    0.12       0.28    0.14    0.29  1.00
Mozambique               0.41            0.13    0.01    0.05       0.14    0.06    0.19  1.00
Peru                     0.07            0.35    0.01    0.02       0.22    0.11    0.23  1.00
Philippines              0.12            0.06    0.03    0.05       0.03    0.23    0.49  1.00
Thailand                 0.06            0.02    0.00    0.06       0.11    0.07    0.68  1.00
Uganda                   0.10            0.04    0.00    0.03       0.02    0.07    0.75  1.00
Venezuela                0.08            0.24    0.17    0.10       0.28    0.08    0.05  1.00
Vietnam                  0.04            0.11    0.00    0.00       0.05    0.10    0.70  1.00
Zambia                   0.34            0.23    0.10    0.07       0.07    0.09    0.11  1.00
Source: Household surveys for each country.




                                                   46

Table 3. Elasticity of Poverty Headcount ($1/day) with Respect to Total Income
                                                            Strata
                                      Non-        Urban       Rural          Urban    Rural
Country           Agriculture      Agriculture     Labor      Labor Transfer Diverse Diverse
Bangladesh                1.64            2.02       1.58      0.63    0.56     1.74    1.09
Brazil                    0.75            1.28       1.94      2.19    0.34     3.63    2.69
Chile                     1.90            2.24       2.06      1.55    2.45     2.29    2.60
Colombia                  0.79            0.60       1.73      1.72    0.93     1.14    1.00
Indonesia                 2.35            2.14       2.38      2.89    1.17     2.58    2.87
Malawi                    0.49            0.30       2.26      1.97    0.43     1.04    0.76
Mexico                    1.73            1.90       3.33      2.08    2.28     1.63    1.80
Mozambique                0.28            0.94       0.97      0.76    0.48     1.58    0.99
Peru                      1.50            1.32       2.37      1.73    0.44     1.09    1.05
Philippines               2.25            1.96       2.98      2.44    1.69     2.42    1.98
Thailand                  2.30            2.42       2.98      2.45    2.78     2.42    2.59
Uganda                    0.28            0.40       1.71      0.34    0.01     0.36    0.21
Venezuela                 0.69            1.16       2.57      2.17    0.01     1.72    1.53
Vietnam                   0.48            1.12       2.81      8.98    0.84     0.86    1.01
Zambia                    0.00            0.64       2.28      0.91    0.45     1.29    0.37
Source: Authors' calculations, based on household survey data.




                                                        47

Table 4. Poverty Elasticities, by Stratum and Income Source, $1/day: Peru
Factor        Agriculture         Non-          Urban                              Urban    Rural
                               Agriculture      Labor       Rural Labor Transfer  Diverse  Diverse

Land                  0.04            0.00           0.00          0.00      0.00     0.01     0.03

Ag.
Unskilled             1.41            0.00           0.00          0.00      0.00     0.25     0.21
Labor
Ag. Skilled
Labor                 0.00            0.00           0.00          0.00      0.00     0.01     0.00

Non-Ag.
Unskilled             0.00            1.08           0.00          0.01      0.00     0.31     0.32
Labor
Non-Ag
Skilled               0.00            0.14           0.00          0.00      0.00     0.05     0.07
Labor
Wage
Labor                 0.00            0.00           2.19          1.58      0.00     0.21     0.13
Unskilled
Wage
Labor                 0.00            0.00           0.16          0.12      0.00     0.01     0.00
Skilled
Agricultural
Capital               0.05            0.00           0.00          0.00      0.00     0.01     0.03

Non-
agricultural          0.00            0.09           0.00          0.00      0.00     0.05     0.12
Capital

Transfers             0.01            0.01           0.02          0.02      0.44     0.18     0.15


Total                 1.50            1.32           2.37          1.73      0.44     1.09     1.05

Source: Authors' calculations, based on household survey data.




                                                        48

Table 5. Poverty Elasticities for Rural Diversified Stratum, $1/day

                               Ag.        Ag.      Non-Ag.     Non-    Wage     Wage                   Non-
Country          Land      Unskilled    Skilled    Unskilled   Ag      Labor    Labor   Agricultural agricultural Transfers Total
                             Labor       Labor       Labor    Skilled                    Capital
                                                               Labor  Unskilled Skilled               Capital

Bangladesh        0.01           0.18      0.00         0.20    0.00      0.43    0.04         0.01         0.03      0.10   1.00
Brazil            0.00           0.10      0.04         0.12    0.00      0.32    0.01         0.01         0.00      0.41   1.00
Chile             0.05           0.16      0.00         0.02    0.00      0.35    0.00         0.07         0.00      0.35  1.00
Colombia          0.00           0.22      0.00         0.30    0.00      0.22    0.02         0.00         0.02      0.21  1.00
Indonesia         0.06           0.32      0.00         0.20    0.00      0.26    0.00         0.04         0.08      0.04   1.00
Malawi            0.03           0.38      0.00         0.07    0.00      0.08    0.00         0.06         0.11      0.27  1.00
Mexico            0.01           0.14      0.00         0.06    0.00      0.48    0.00         0.01         0.01      0.30   1.00
Mozambique        0.01           0.43      0.00         0.07    0.00      0.07    0.00         0.02         0.20      0.20   1.00
Peru              0.02           0.20      0.00         0.30    0.07      0.13    0.00         0.03         0.11      0.14   1.00
Phillippines      0.22           0.00      0.02         0.14    0.01      0.30    0.01         0.12         0.08      0.11  1.00
Thailand          0.04           0.21      0.03         0.03    0.01      0.24    0.07         0.02         0.02      0.35   1.00
Uganda            0.14           0.15      0.00         0.06    0.00      0.09    0.06         0.26         0.14      0.10   1.00
Venezuela         0.00           0.10      0.00         0.32    0.01      0.28    0.04         0.00         0.00      0.26   1.00
Vietnam           0.01           0.09      0.00         0.14    0.00      0.00    0.00         0.00         0.55      0.21   1.00
Zambia            0.01           0.03      0.00         0.20    0.00      0.43    0.05         0.03         0.13      0.12   1.00
Source: Authors' calculations, based on household survey data.




                                                                      49

    Table 6. Overview of Scenarios

                                                    Rich Agriculture                                       Global
                                                                                          (All countries and merchandise)
    Instrument                                      Doha                    Full                      Doha                  Full

    Agr. Tariffs Rich                    -45 %, -70 %, -75 %c             -100 %           -45 %, -70%, -75 %              -100 %

    Agr. Tariffs
    Poor (Non-LDCa)                                  n.a.                   n.a.        -35%, -40%, -50%, -60%d            -100 %


    Agr. Export Subsidies                          -100 %                 -100 %                    -100 %                 -100 %

    Amber Box                                -75 % Group 1                                    -75 % Group 1
    Subsidiesb                               -60 % Group 2                -100 %              -60 % Group 2                -100 %


    Non-Agr. Tariffs Rich                            n.a.                   n.a.                      -50%                 -100 %

    Non-Agr. Tariffs Poor
    (Non-LDCa)                                       n.a.                   n.a.                     -33 %                 -100 %


    Green Box Subsidies                              n.a.                   n.a.                       n.a.                 n.a.

    aLeast developed countries are not required to make any tariff reductions under Doha scenarios.
    bGroup 1 countries have amber box subsidies accounting for more than 20% of producer revenue. Group 2 countries
    have support less than 20 percent of producer revenue. A third grouping exists for developing countries where 40
    percent reductions are required, but adequate data on amber box subsidies is available to model this.
    c
    These three percentage cuts are applied in a tiered formula whereby higher portions of the tariff are more deeply cut.
    Tiers are defined over the tariff rate and the reductions increase at 15% and then 90%.
    dThese four percentage cuts are applied in a tiered formula whereby higher potions of the tariff are more deeply cut.
    Tiers are defined over the tariff rate and the reductions increase at 20%, 60%, and 120%.
Table 7. Welfare and Terms of Trade Results

                                Rich Agriculture                                          Global
                                                                           (All countries and merchandise)
                           Full                      Dohaa                      Fulla                    Dohaa
                     ToT Welfare                 ToT Welfare                ToT Welfare              ToT Welfare
Rich Countries
Aust. and
New Zlnd.             2.79           0.72         1.09          0.25        2.30         0.76        1.09         0.28

Japan                -1.32            0.96       -0.38          0.44         0.28        1.29         0.04        0.54
Canada               -0.07           0.27        -0.02          0.10        -0.64        0.39        -0.23        0.06
US                    0.36           0.02         0.13          0.01         0.29        0.06        0.01         0.01
Europe FTA           -0.30           0.21        -0.06          0.18        -0.12        0.42        -0.12        0.20
Focus Countries
Bangladesh           -0.58          -0.27        -0.21         -0.10        -5.66       -0.65        -0.04       -0.04
Brazil                5.48            0.75        1.91          0.26         3.72        0.67         2.03        0.31
Chile                 0.74           0.13         0.15          0.01        0.58         0.32        0.18         0.01
Colombia              1.27          0.00#         0.60          0.01        -1.52       -0.54        0.33        -0.07




                                                                 50

Indonesia             -0.23         -0.26        -0.10          -0.09       1.11   0.51   0.23   0.07
Malawi                  2.67        1.92          0.18           0.22      3.56    3.83  0.34    0.32
Mexico               -0.14#         -0.25        -0.11          -0.08      -2.02  -0.20  -0.43  -0.12
Mozambique            -0.39         -0.54        -0.15          -0.14       0.00   1.19  -0.13  -0.08
Peru                    3.47         0.58         0.16          -0.01       0.66   0.60  0.11   -0.02
Philippines           0.01#         -0.21        -0.08          -0.11      -0.33   0.49  0.12    0.12
Tanzania              0.04#         -0.08        -0.21          -0.08      -2.05  -0.66  -0.29  -0.08
Thailand                1.21         0.77         0.24           0.15       1.50   2.08   0.54   0.51
Uganda                -0.50         -0.21        -0.63          -0.18      -0.99  -0.32  -0.64  -0.18
Venezuela             -0.41         -0.11        -0.23          -0.06      -2.19  -0.26  -0.68  -0.03
Vietnam                 0.34        0.12#        -0.12          -0.18      -1.25   5.73  -0.85  -1.17
Zambia                -0.17         -0.20        -0.01          -0.05      -0.46   0.28   0.04  -0.03
Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.
a
Sensitivity analysis is not conducted for these liberalization simulations.




                                                                 51

Table 8. Percent Change in Farm Income for Rich Regions by Source (On/Off-Farm)
                      Rich Region Doha Ag. Reforms                        Rich Region Ag. Full Reform
Region                On-farm         Off-farm           Total           On-farm       Off-farm        Total
Australia
and New                     7.3             -0.0           4.4                17.3         -0.0#       10.5
Zealand

Japan                     -15.5              0.6          -1.4              -28.2           1.2         -2.5

Canada                      3.5              0.0           0.4                 6.3          0.1          0.7

US                         -0.3              0.0          -0.0                -4.4          0.1         -0.3

EU and
Other                      -5.8              0.3          -3.5              -12.7           0.5         -7.7
Europe
Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.



Table 9. Disaggregate U.S. Farm Household Income Impacts of Ag. Reforms
Income             Rice Hhld.            Sugar Hhld.         Cotton Hhld.        Dairy Hhld.    Other Hhld.
Group         Doha              Full     Doha      Full     Doha        Full     Doha    Full   Doha    Full
 10%ile         1.36           -5.08     -0.12    -0.97      -2.09    -8.31      -0.30  -2.44    0.02  0.08
 20%ile         1.37           -5.11     -0.12    -0.97      -2.09    -8.31      -0.30  -2.00   -0.03  -0.12
 30%ile         1.89           -6.55     -0.34    -2.64      -1.63    -7.03      -0.43  -2.94    0.00  0.00#
 40%ile         1.89           -6.57     -0.87    -4.80      -2.13    -8.14      -0.32  -2.11    0.01  0.08
 50%ile         6.32         -16.63      -0.87    -4.80      -1.60    -7.64      -0.41  -2.56   -0.08  -0.35
 60%ile         1.63           -7.68     -0.87    -4.80      -1.18    -5.00      -0.44  -2.74   -0.08  -0.34
 70%ile         4.64         -14.92      -0.37    -1.98      -1.47    -6.66      -0.66  -3.74   -0.17  -0.70
 80%ile         5.53         -17.08      -0.37    -1.98      -1.15    -5.13      -0.47  -2.92   -0.18  -0.78
 90%ile         5.60         -17.79      -0.65    -3.73      -1.81    -8.94      -0.71  -4.26   -0.31  -1.31
 95%ile         5.33         -18.91      -1.33    -6.49      -1.61    -6.77      -0.46  -3.56   -0.30  -1.31
 100%ile        5.31         -18.83      -1.33    -6.49      -3.53   -12.68      -0.50  -4.04   -0.56  -2.39
Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.




                                                             52

Table 10. Percentage Change in Cost of Living Adjusted Factor Returns: $1/day Poverty

Country          Land    AgUnskl    AgSkl     NagUnskl    NagSkl      WgUnskl      WgSkl     AgCap  NagCap Transfer

Bangladesh      1.64        0.77     0.57        -0.38      -0.46       -0.06#      -0.46      0.53 -0.55   -0.33

Brazil         39.28       16.06    14.73        -1.68      -1.99         0.46      -1.82    14.63  -2.31   -0.69

Chile          12.55        6.13     5.44        -0.96      -1.19         0.12      -1.18      5.42 -1.29   -0.72

Colombia        9.75        4.41     3.74        -1.70      -1.84        -0.58      -1.84      3.68 -2.19   -1.13

Indonesia       2.56        1.22     0.81        -0.75      -0.95        -0.12      -0.94      0.82 -0.94   -0.57

Malawi          1.78        1.37     1.13         0.78      0.53          1.02      0.54       1.22  0.61    1.35

Mexico         16.73#       4.82     4.13        -1.20      -1.50       -0.18#      -1.50      4.13 -1.60   -1.11

Mozambique      1.38        0.51     0.31        -0.61      -0.65        -0.26      -0.64     0.32# -0.72   -0.46

Peru           14.61        7.90     6.34        -1.53      -1.88         1.30      -1.61      6.19 -2.05   -0.71

Philippines     2.20        1.07     0.65        -0.79      -0.90       -0.03#      -0.86      0.54 -1.16   -0.57

Thailand       22.67       10.93     8.28        -1.66      -2.70         2.41      -2.42      7.83 -3.27   -1.27

Uganda         -0.14#       -0.16   -0.19        -0.15      -0.22        -0.16      -0.22     -0.18 -0.21   -0.19

Venezuela       2.02        0.90     0.78        -0.39      -0.43        -0.20      -0.43      0.79 -0.45   -0.33

Vietnam         4.23        2.03     1.61        -0.66      -0.86       -0.04#      -0.86      1.68 -0.76   -0.32

Zambia          1.56        0.75     0.60        -0.32      -0.38        -0.07      -0.38      0.59 -0.54   -0.21

Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.
Note: All earnings have been deflated by the country-specific cost of living at the poverty line.




                                                               53

Table 11. Percent Change in the Poverty Headcount ($1/day) across Developing Country
Stratums, when Rich Countries undertake Full Agricultural Reform
                                   Non-         Urban                                  Urban     Rural
Country         Agriculture    Agriculture      Labor        Rural Labor     Transfer  Diverse  Diverse

Bangladesh             -1.27          0.78           0.09#         0.05          0.18    -0.05#    0.02#

Brazil               -10.45           2.21         -0.57#          -0.79         0.23     -7.10    -4.81

Chile                -12.53           2.24         -0.22#         -0.14#         1.78     -4.65    -4.25

Colombia               -3.37          1.04            1.04         1.03          1.06     0.13#    0.01#

Indonesia              -2.86          1.62            0.45         0.46          0.66     -0.56    -0.80

Malawi                 -0.67         -0.22           -2.10         -1.88        -0.57     -1.20    -0.92

Mexico                 -7.83          2.35            0.77         0.37#         2.57    -0.29#   -0.52#

Mozambique             -0.15          0.62            0.25          0.20         0.22      0.21    0.06#

Peru                 -10.83           2.14           -2.53         -1.82         0.32     -1.61    -1.32

Philippines            -3.52          1.68           0.22#         0.15#         0.97     -0.55    -0.54

Thailand             -22.04           4.35           -6.68         -5.49         3.43     -7.69    -7.90

Uganda                  0.05          0.07            0.34          0.06         0.00      0.06     0.04

Venezuela              -0.61          0.45            0.54          0.46         0.00      0.37     0.31

Vietnam                -0.99          0.78           0.13#         1.87          0.26     -0.64    0.36

Zambia                  0.00          0.26            0.40          0.12         0.09      0.19     0.06

Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.




                                                             54

Table 12. National Poverty Impacts due to Rich Country Liberalization of Agriculture: Full
versus Dohaa Reform
                         Rich Agriculture Full Reform                      Rich Agriculture Doha Reform
                        $1/day                     $2/day                    $1/day                 $2/day
Country               %        1000s              %        1000s            %     1000s            %     1000s
Bangladesh        -0.06#        -27#           0.06            62         0.00         0         0.02      21
Brazil            -1.88         -431          -2.61         -958         -0.73     -167        -0.96      -352
Chile             -3.99          -12          -2.48           -35        -0.99        -3       -0.57        -8
Colombia          -0.29          -12          -0.67           -59        -0.17        -7       -0.46       -40
Indonesia         -1.18         -177          -0.20         -210         -0.13       -20         0.00        0
Malawi            -0.72          -31          -0.32           -25         0.41       17          0.15       12
Mexico              0.34          32          -0.10           -25         0.15       14          0.03        7
Mozambique          0.09            5          0.06             8         0.05         3         0.02        3
Peru              -0.43          -19          -1.71         -157          0.04         2       -0.18       -17
Philippines       -0.66          -75          -0.41         -143          0.03         3         0.00        0
Thailand           -7.10         -84          -4.15         -806         -1.43       -17       -0.83      -161
Uganda              0.04            7          1.12          220          0.04         7         1.58      310
Venezuela           0.24            8          0.18            13         0.11         4         0.09        6
Vietnam             0.25            4         -0.24           -62         0.14         2         0.12      31
Zambia              0.13            8          0.03             2         0.03         2         0.01        1
Source: Authors' simulations.
#Result can not be distinguished from zero change at the 95% confidence level.
aNo sensitivity analysis is conducted for Doha liberalization simulation.




                                                                55

Table 13. Percentage change in the $1/day Head Count
                                                    Rich                 Rich            Poor              Poor
                               Total            Agricultural           Non-agr.       Agricultural      Non-agr.
    Country                                       Reforms              Reforms         Reforms          Reforms
    Bangladesh                        0.29                 -0.11             -0.05            -0.18             0.62
    Brazil                           -1.42                 -1.79             -0.07            -0.15             0.60
    Chile                           -4.99                  -3.89             0.00             -1.41            0.31
    Colombia                          0.10                 -0.29             -0.04            -0.28            0.71
    Indonesia                        -1.45                 -1.24             -0.46            -0.82            1.07
    Malawi                           -1.84                 -0.74              0.02            -0.96            -0.16
    Mexico                            1.35                  0.31              0.31             0.61             0.12
    Mozambique                       -0.69                  0.07              0.04            -1.08             0.28
    Peru                             -0.80                 -0.40              0.03            -0.23            -0.19
    Philippines                      -0.76                 -0.76             -0.70            -0.56            1.27
    Thailand                         -8.87                 -6.63             -0.53            -4.55             2.84
    Uganda                            0.07                  0.04              0.00            -0.06             0.08
    Venezuela                         0.85                  0.26              0.05            -0.15             0.70
    Vietnam                          -5.85                  0.22             -1.84            -1.70            -2.53
    Zambia                            0.09                  0.14              0.05            -0.29             0.19
Source: Authors' simulations.




Table 14. Change in National Poverty due to Global Liberalization: Doha versus Fulla
                                   Global Liberalization                           Global Doha Liberalization
                               $1/day                      $2/day                 $1/day               $2/day
    Country                %          1000s            %         1000s        %        1000s       %         1000s
    Bangladesh              0.29           130          0.13          133      -0.05       -22      0.01         10
    Brazil                 -1.42          -325         -1.73         -635      -0.79      -181     -1.09       -400
    Chile                  -4.99           -15         -3.26           -46     -1.28        -4     -0.76        -11
    Colombia                0.10             4          1.20          105     -0.09         -4     -0.13        -11
    Indonesia              -1.45          -218         -0.59         -619      -0.20       -30     -0.09        -94
    Malawi                 -1.84           -78         -0.77           -60     0.35        15       0.12           9
    Mexico                  1.35           125          0.67          166      0.13        12       0.03           7
    Mozambique             -0.69           -42         -0.30           -38      0.02         1      0.01           1
    Peru                   -0.80           -35         -1.87         -172       0.06         3     -0.15        -14
    Philippines            -0.76           -86         -0.19           -66     -0.25       -28     -0.12        -42
    Thailand               -8.87          -105         -4.49         -872      -1.97       -23     -1.02       -198
    Uganda                  0.07            12          2.19          430      0.04          7      1.58        310
    Venezuela               0.85            28          0.78            54     0.21          7      0.18         13
    Vietnam                -5.85           -90         -4.84       -1242       0.89        14       0.70        180
    Zambia                  0.09             5          0.03             2      0.03         2      0.01           1
Source: Authors' simulations.
a
No sensitivity analysis is conducted for global liberalization simulations.




                                                                 56

Figure 1. Overview of the Analytical Framework




                                            57

Figure 2. Total poverty Rate versus Poverty Rate among Agricultural Specialized households
(line denotes locus of points with equal poverty rates)




Source: Authors calculations.




                                               58

Figure 3. Estimated Budget Shares across the Income Spectrum in Peru


                                                   GTAP proxy                                     Consumption
                                              for $1/day and $2/day                                 Per capita
 0.6
                                                                                                                 Grains, other crops


                                                                                                                 Meat, dairy, fish
 0.5

                                                                                                                 Processed food,
                                                                                                                 beverages,
                                                                                                                 tobacco
 0.4                                                                                                             Textiles, apparel,
                                                                                                                 footwear

                                                                                                                 Utilities, other
                                                                                                                 housing services
 0.3
                                                                                                                 Wholesale/retail
                                                                                                                 trade

                                                                                                                 Manufactures/
 0.2                                                                                                             Electronics

                                                                                                                 Transport,
                                                                                                                 communication


 0.1                                                                                                             Financial and
                                                                                                                 business services

                                                                                                                 Housing,
                                                                                                                 education, health,

  0                                                                                                              public services


   3.51 3.71 3.91 4.11 4.31 4.51 4.71 4.91 5.11   5.31 5.51  5.71 5.90 6.10  6.30 6.50 6.70 6.90 7.10 7.30

                              Consumption per capita (natural logarithm, in 2001 $US)




Source: Authors' calculations




                                                                 59