View in EDS HTML Full Text PDF Full Text

Comparing Modeling Approaches for Assessing Priorities in International Agricultural Research

Saved in:
Bibliographic Details
Title: Comparing Modeling Approaches for Assessing Priorities in International Agricultural Research
Language: English
Authors: Petsakos, Athanasios, Hareau, Guy, Kleinwechter, Ulrich, Wiebe, Keith, Sulser, Timothy B.
Source: Research Evaluation. Apr 2018 27(2):145-156.
Availability: Oxford University Press. Great Clarendon Street, Oxford, OX2 6DP, UK. Tel: +44-1865-353907; Fax: +44-1865-353485; e-mail: jnls.cust.serv@oxfordjournals.org; Web site: http://applij.oxfordjournals.org/
Peer Reviewed: Y
Page Count: 12
Publication Date: 2018
Document Type: Journal Articles
Reports - Research
Descriptors: Agriculture, Research, Comparative Analysis, Models, Needs Assessment, Research Needs, Global Approach
DOI: 10.1093/reseval/rvx044
ISSN: 0958-2029
Abstract: This article examines how the estimated impacts of crop technologies vary with alternate methods and assumptions, and also discusses the implications of these differences for the design of studies to inform research prioritization. Drawing on international potato research, we show how foresight scenarios, realized by a multi-period global multi-commodity equilibrium model, can affect the estimated magnitudes of welfare impacts and the ranking of different potato research options, as opposed to the static, single-commodity, and country assumptions of the economic surplus model which is commonly used in priority setting studies. Our results suggest that the ranking of technologies is driven by the data used for their specification and is not affected by the foresight scenario examined. However, net benefits vary significantly in each scenario and are greatly overestimated when impacts on non-target countries are ignored. We also argue that the validity of the single-commodity assumption underpinning the economic surplus model is case-specific and depends on the interventions examined and on the objectives and criteria included in a priority setting study.
Abstractor: As Provided
Entry Date: 2018
Accession Number: EJ1176296
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHvaGNMccohoHoHeTgV_gWkAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDLQWkm7L1XmLoILazgIBEICBm3mTb300iKZ7eo-sx5jkOYneZxkcsYnnQryV2d_TOf-jjfUbnU5z2TY-SI696b3FuSMmWVKJuKrNR9bE0gtRCP5AAli2wAMOxRRLMhmVzqDiuFItiXWtbEoKwvsx6Tut59gwwU8MQ2Gj9VGWn14fnEVSRiHuEJsuWgOLFkdqi05CmkXdYz84PbbF0LDgbeC4n6clpjmTNiaQUANa
Text:
  Availability: 1
  Value: <anid>AN0128889675;p9101apr.18;2018Oct18.11:51;v2.2.500</anid> <title id="AN0128889675-1">Comparing modeling approaches for assessing priorities in international agricultural research. </title> <p>This article examines how the estimated impacts of crop technologies vary with alternate methods and assumptions, and also discusses the implications of these differences for the design of studies to inform research prioritization. Drawing on international potato research, we show how foresight scenarios, realized by a multi-period global multi-commodity equilibrium model, can affect the estimated magnitudes of welfare impacts and the ranking of different potato research options, as opposed to the static, single-commodity, and country assumptions of the economic surplus model which is commonly used in priority setting studies. Our results suggest that the ranking of technologies is driven by the data used for their specification and is not affected by the foresight scenario examined. However, net benefits vary significantly in each scenario and are greatly overestimated when impacts on non-target countries are ignored. We also argue that the validity of the single-commodity assumption underpinning the economic surplus model is case-specific and depends on the interventions examined and on the objectives and criteria included in a priority setting study.</p> <p>Keywords: priority setting; foresight analysis; economic surplus model; international agricultural research</p> <hd id="AN0128889675-2">1. Introduction</hd> <p>There is a growing demand from public and private donors and other decision makers for more efficient spending of resources used in agricultural research. At the same time, tightened budgets have led agricultural research organizations to formalize priority setting approaches for the identification of research activities with the highest possible impact in terms of economic efficiency, poverty alleviation, and other institutional, social, and environmental objectives, to inform decisions on the optimal allocation of research funds. Priority setting increases the credibility and objectivity of the decisions taken at the various institutional levels (institutes, programs, and projects) and offers a systematic way of planning and managing research which is consistent with informed scientific opinion and stakeholder needs ([<reflink idref="bib2" id="ref1">2</reflink>]).</p> <p>Various qualitative and quantitative approaches have been used for priority setting, including scoring methods, congruence rules, benefit–cost analysis, economic surplus models, and multi-objective programming, each with distinct advantages and drawbacks compared to the others ([<reflink idref="bib34" id="ref2">34</reflink>]; [<reflink idref="bib4" id="ref3">4</reflink>]). Priority setting can be limited to the subjective ranking of the various research options using expert judgement, or it may involve formal <emph>ex ante</emph> impact assessment to quantify the potential benefits of alternative technologies. In the latter case, an economic surplus model combined with discounted benefit–cost measures is probably the most common method ([<reflink idref="bib1" id="ref4">1</reflink>]). In its simplest form, the economic surplus model is a comparative static representation of the supply and demand equilibrium for a single market (commodity). It can estimate welfare changes for producers and consumers brought about by policies or technological innovations. Extensions to the basic model include price and technology spillovers among countries or regions connected via trade ([<reflink idref="bib7" id="ref5">7</reflink>]; [<reflink idref="bib13" id="ref6">13</reflink>]), while Geographical Information Systems have also been proposed for identifying similar agroecologies in neighboring regions to quantify the direction and intensity of these spillover effects ([<reflink idref="bib40" id="ref7">40</reflink>]).</p> <p>The economic surplus model has been widely applied in <emph>ex ante</emph> impact assessment studies to inform decisions for defining priorities, both in national research programs ([<reflink idref="bib8" id="ref8">8</reflink>]; [<reflink idref="bib23" id="ref9">23</reflink>]), and in International Agricultural Research Centers ([<reflink idref="bib5" id="ref10">5</reflink>]; [<reflink idref="bib12" id="ref11">12</reflink>]). One important limitation of the approach is that it offers a static representation of the modeled commodity market and thus ignores the socioeconomic and biophysical dynamics which can affect the expected impact of a given intervention. Although some of its implementations allow for an explicit representation of dynamics in production and consumption ([<reflink idref="bib13" id="ref12">13</reflink>], the economic surplus model is not able to analyze well-structured foresight scenarios, like those proposed by the Intergovernmental Panel on Climate Change (IPCC) which describe plausible future pathways in population, income, climate, and other drivers of change. Furthermore, with some few exceptions ([<reflink idref="bib7" id="ref13">7</reflink>]), the analysis of agricultural research priorities at the regional or international level with economic surplus models either ignores welfare impacts on horizontally related markets (countries not targeted for the release of new technologies) or does so aggregately, probably because it is difficult to estimate the individual demand and supply curves in each country ([<reflink idref="bib1" id="ref14">1</reflink>]). Finally, the economic surplus model cannot directly account for the effects of technological innovations on multiple commodity markets. Instead, cross-price effects are implicitly represented in the assumed supply and demand elasticities of the commodity examined ([<reflink idref="bib27" id="ref15">27</reflink>]). In this regard, several studies have used partial or general equilibrium models to estimate the direct and indirect welfare impacts of new technologies ([<reflink idref="bib32" id="ref16">32</reflink>]; [<reflink idref="bib18" id="ref17">18</reflink>]), yet not explicitly for ranking specific investment options. One possible reason is that multi-commodity equilibrium analyses seem to be beyond the scope of priority setting exercises ([<reflink idref="bib6" id="ref18">6</reflink>]).</p> <p>Given the aforementioned limitations, this article seeks to provide a quantitative assessment of whether the relaxation of the related assumptions in the economic surplus approach adds value to a priority setting exercise and can therefore justify a shift from the standard quantitative paradigm in studies aiming to inform research prioritization. More precisely, we examine how the simulated dynamics of a quantitative foresight scenario analysis can influence the returns to investment and the ranking of different crop research options, and to what extent the consideration of other countries can affect the welfare results, as opposed to focusing on target countries only. We also look at multiple commodity markets to assess the importance of cross-price effects which are ignored under the simple economic surplus model approach.</p> <p>Our study builds on the analysis of research priorities for potato which was carried out by the CGIAR Research Program (CRP) on Roots, Tubers, and Bananas (RTB) ([<reflink idref="bib12" id="ref19">12</reflink>]) and constitutes an analytical example of using economic surplus models for informing priority setting decisions in international agricultural research. In this article, we replicate the RTB exercise and then estimate the expected global welfare impacts of four of the same potato research options under different foresight scenarios with the IMPACT model (International Model for Policy Analysis of Agricultural Commodities and Trade), which is a recursive dynamic multi-commodity partial equilibrium representation of the global agricultural sector ([<reflink idref="bib30" id="ref20">30</reflink>]), developed at the International Food Policy Research Institute (IFPRI). For examining the role of dynamics, we compare the welfare results from foresight scenarios simulated with IMPACT against a counterfactual which is created by keeping constant all model parameters that are intrinsically dynamic, like population, income, and climate impact on yields. This counterfactual therefore represents a static scenario of technology adoption and serves as a proxy for the economic surplus model, since it only considers a research-induced increase in crop yields which finally leads to an endogenously calculated outward shift of the supply function, ceteris paribus. By also looking at different result aggregation levels (target countries and global) and the magnitude of cross-price effects, we examine how the underlying assumptions of the economic surplus model can influence the results of a priority setting study.</p> <p>In the next section, we introduce the IMPACT model and present the four potato research options compared in this article. In Section 3 we explain the approach for simulating a productivity shock with IMPACT and detail the assumptions behind the different foresight scenarios considered. Results are presented in Section 4, followed by a discussion in Section 5 on implications of our findings for studies aiming to inform research prioritization.</p> <hd id="AN0128889675-3">2. Materials and methods</hd> <p></p> <hd id="AN0128889675-4">2.1 The IMPACT partial equilibrium agricultural sector model</hd> <p>IMPACT is an integrated modeling framework which combines economic, crop, livestock, and water models designed for the analysis of future developments of global agricultural production, demand, trade, and prices up to 2050. The model has been used extensively in analyses related to projections of global, regional, and national food supply and demand ([<reflink idref="bib29" id="ref21">29</reflink>]; [<reflink idref="bib15" id="ref22">15</reflink>]; [<reflink idref="bib28" id="ref23">28</reflink>]), commodity-specific analyses ([<reflink idref="bib35" id="ref24">35</reflink>]), or analyses of issues related to the agricultural sector, such water scarcity ([<reflink idref="bib31" id="ref25">31</reflink>]) and climate change ([<reflink idref="bib24" id="ref26">24</reflink>]; [<reflink idref="bib18" id="ref27">18</reflink>]). IMPACT uses FAOSTAT and other data, and its projections are based on regional and global scenarios that draw on the fifth assessment report of the IPCC. A detailed description of the IMPACT modeling framework and underlying equations can be found in [<reflink idref="bib30" id="ref28">30</reflink>].</p> <p>The food production module in IMPACT comprises 62 agricultural commodities and distinguishes 159 geopolitical regions and 154 water basins globally, which combine to 320 geographic 'food production units' (FPUs). Crop production takes place at the FPU level and in modeling terms is defined as the product of response functions for crop yields and harvested areas. Cropland is divided between irrigated and rainfed areas, and the crop share allocations are determined by a market which simulates the equilibrium between total land supply and the area demanded by each crop. Yields depend on exogenous trend parameters, input and crop prices, and possible water stress or climate-induced shocks. On the demand side, a set of separate functions is used to represent different demand components for each country and commodity, namely, food, feed, biofuels, crush demand for oilseeds, and other uses, which add up to total demand for any single crop. Whereas all demand types are a function of the commodity's own price, demand for food additionally depends on the prices of other competing commodities (through cross-price elasticities), but also takes into account assumed changes in population and per capita income.</p> <p>The individual regions for which supply and demand is calculated are connected to each other via trade. Net trade adds to domestic production to finally equilibrate domestic supply and demand. For each year simulated by the model, global demand and supply for every commodity are brought into equilibrium and determine an endogenous world market price which leads to global market clearing. The domestic producer and consumer prices for all commodities in all regions are derived from this world equilibrium price. When an exogenous shock is applied to the model, like increases in crop productivity as done in this study, the world market price adjusts to those changes to establish new market equilibrium and produce a new set of country-level prices, demand, supply, and trade ([<reflink idref="bib35" id="ref29">35</reflink>]).</p> <p>IMPACT also includes a post-simulation module to estimate the welfare impacts of a given intervention scenario in relation to a reference (baseline) scenario. The estimations comprise producer surplus, consumer surplus, and net welfare effects to the agricultural sector. Further, the costs of a particular intervention, if available, can be taken into account to calculate the internal rate of return (IRR) as a measure of returns on investment. Consumer surplus is estimated by using the stored equilibrium point and the model's own price demand elasticities to calculate a slope and price intercept parameter to create a linear demand function for each commodity. In contrast, estimation of producer surplus is not straightforward because supply in IMPACT is modeled as the product of yield and area response functions and not through a traditional supply curve which reflects a producer's marginal cost curve. For this reason, a synthetic supply curve is created which expresses yields and harvested areas as a function of crop prices. Such supply curves for each commodity are created for every land type (irrigated, rainfed) in each FPU and are then aggregated at the national level.[<reflink idref="bib1" id="ref30">1</reflink>]</p> <hd id="AN0128889675-5">2.2 Proposed potato technologies</hd> <p>The research options (technologies) analyzed are (<reflink idref="bib1" id="ref31">1</reflink>) improved potato seed systems, (<reflink idref="bib2" id="ref32">2</reflink>) potato varieties resistant to bacterial wilt, (<reflink idref="bib3" id="ref33">3</reflink>) virus-resistant potato varieties, and (<reflink idref="bib4" id="ref34">4</reflink>) varieties resistant to late blight.[<reflink idref="bib2" id="ref35">2</reflink>] These technologies were previously examined in the RTB analysis of research priorities and were identified by RTB via a multistage participatory process ([<reflink idref="bib22" id="ref36">22</reflink>]). For each research option, specific target countries in Africa, Asia, and Latin America were also identified. Their selection was based on the current distribution of the constraints addressed by the different research options and on considerations about current and future target geographical regions for international potato research for development, as carried out by RTB, the International Potato Center (CIP), and partners.</p> <p>The quantification of every research option in each target country was based on expert judgment and consists of assumptions regarding the probability of research success, the expected adoption rates, and the expected changes in yield and input costs.[<reflink idref="bib3" id="ref37">3</reflink>] An important note is that the maximum adoption rates and the probability of success should ideally be defined as empirical distributions ([<reflink idref="bib1" id="ref38">1</reflink>]). However, the RTB priority assessment study only provides country-specific point estimates for both parameters, elicited from expert workshops, and considers a 'lower' and a 'higher' adoption scenario to partially account for the uncertainty on the adoption process. To simplify the analysis, in this article we do not assume any variability for either the maximum adoption rates or the probability of success. All adoption data come from the 'higher' adoption scenario, and we define the probability of success as the mean of the odds for achieving the research objectives.</p> <p>Further assumptions were made on the costs of research, development, and dissemination. Annual costs for research and development are an estimation of costs incurred by CIP and by the national agricultural research systems. For the dissemination cost, a fixed figure per hectare is assumed for the marginal area of adoption. This cost depends on the type of technology analyzed. Varietal technologies are assumed to require an investment of US$50 per hectare of finally adopted area, while more knowledge-intensive technologies, e.g. seed systems interventions analyzed herein, are assumed to require US$80 per hectare. Table 1 provides a summary of the technologies, while Sections 2.1.1–2.1.4 give a more detailed description for each research option and its parameterization as an adoption scenario. Detailed lists with the target countries and parameter values for each research option are provided as Supplementary Material.</p> <p>Summary of technology scenarios</p> <p> <ephtml> <table><tr><th>Technology parameters </th><th>Improved seed systems </th><th>Bacterial wilt-resistant varieties </th><th>Virus-resistant varieties </th><th>Late blight- resistant varieties </th></tr><tr><td>Research lag (years) </td><td>3 </td><td>10 </td><td>2 </td><td>2 </td></tr><tr><td>Adoption lag (years) </td><td>5 </td><td>10 </td><td>10 </td><td>10 </td></tr><tr><td>Countries targeted</td><td>27 </td><td>22 </td><td>27 </td><td>32 </td></tr><tr><td>Adoption ceiling (%) </td><td>3–20 </td><td>10–60 </td><td>15–40 </td><td>10–60 </td></tr><tr><td>Total annual R&D costs (million US$) </td><td>8 </td><td>4 </td><td>8 </td><td>16 </td></tr><tr><td>Dissemination costs (US$/ha) </td><td>80 </td><td>50 </td><td>50 </td><td>50 </td></tr><tr><td>Maximum expected yield change (%) </td><td>20 </td><td>10–30 </td><td>40 </td><td>12–32 </td></tr><tr><td>Production cost change (%) </td><td>20 </td><td>0 </td><td>−5 </td><td>−2 to −5 </td></tr><tr><td>Probability of success (%) </td><td>60–80 </td><td>50 </td><td>70 </td><td>80 </td></tr></table> </ephtml> </p> <p>1 aSee Supplementary Data for country-level specifications.</p> <p>2 Source: [<reflink idref="bib12" id="ref39">12</reflink>].</p> <hd id="AN0128889675-6">2.2.1 Seed systems: Improving seed production and distribution</hd> <p>Poor seed quality is a major constraint to potato yields in many regions ([<reflink idref="bib10" id="ref40">10</reflink>]; [<reflink idref="bib14" id="ref41">14</reflink>]). The improvement of potato seed production and distribution through the development and implementation of improved seed systems reduces yields losses from seed-borne diseases and improves farmers' access to seed potato tubers of higher physical and physiological quality. The quantified impacts of the improved seed systems technology modeled here across 27 target countries are assumed to include an increase in potato yield of 20% and a rise in production costs by the same percentage as a result of the higher cost for quality seed and the increased use of other inputs. The research lag was set to 3 years, reflecting the investment that CIP has undertaken in previous years to develop methods and approaches for providing quality seed materials to farmers. Improved potato seed systems therefore represent a relatively well-developed technology which requires only some adjustment to local conditions in the target countries. Maximum adoption rates vary across the target countries from 3 to 20%, while the probability of research success is between 60 and 80%, based on the experience of CIP in seed systems work in the respective countries.</p> <hd id="AN0128889675-7">2.2.2 Potato varieties resistant to bacterial wilt</hd> <p>Potato varieties resistant to bacterial wilt aim at reducing yield losses caused by the pathogenic bacterium <emph>Ralstonia solanacearum</emph>, which is a major biotic constraint in many developing countries, particularly in Africa ([<reflink idref="bib10" id="ref42">10</reflink>]) and in the Andean region ([<reflink idref="bib33" id="ref43">33</reflink>]). The improved varieties are expected to increase yields in target countries by 10–30%, based on expert opinions about the incidence of the disease in each country. Since no chemical treatment is available for bacterial wilt, adoption is not expected to bring about input use changes, and thus production costs are assumed to remain constant. The technology has a research lag of 10 years, which is the longest among all four research options, since breeding efforts targeted at this trait need to be re-initiated. The probability of research success is assumed to be 50% because the development of bacterial wilt-resistant varieties requires more upstream research, and success is therefore more uncertain than for the other technologies considered.</p> <hd id="AN0128889675-8">2.2.3 Virus-resistant potato varieties</hd> <p>Viruses are considered the most important biotic constraint for seed potato. Virus-resistant potato varieties aim at reducing yield losses caused by potato virus diseases, namely, potato virus Y, potato leaf roll virus, potato virus S, and potato virus X. Although some virus resistance has already been achieved, this technology aims at introducing even higher levels of virus resistance in future potato varieties. The impact pathway of virus-resistant potato varieties partially overlaps with the seed system technology in the development of high-quality virus-free potato seed. The technology further assumes high levels of virus resistance during the crop's entire life cycle, and therefore it is expected to lead to the highest yield increases of all four options assessed. In addition, a small reduction in production costs of 5% across all countries is also assumed. This reflects potentially lower costs for seed replacement, since virus-resistant varieties are less prone to seed degeneration, and also lower labor costs due to reduced need for roguing infected plants. The development of virus-resistant varieties is rather complex, and therefore the estimated probability of success is assumed to be 70%. Since varieties with lower degrees of virus resistance are already available at CIP, the diffusion of the technology can start after a comparatively short research lag of only 2 years.</p> <hd id="AN0128889675-9">2.2.4 Late blight-resistant potato varieties</hd> <p>Yield losses from late blight (<emph>Phytophtora infestans</emph>) are still considered to be the most important biotic constraint on potato production worldwide ([<reflink idref="bib3" id="ref44">3</reflink>]). Breeding for late blight-resistant varieties at CIP began in 1975, and today it is the primary trait of one of the two major breeding populations of the Center (the other being virus resistance). Although late blight-resistant varieties have already been developed and released by CIP, continuous breeding efforts are necessary since resistance typically breaks down a few years after the varieties are introduced into the field ([<reflink idref="bib36" id="ref45">36</reflink>]).</p> <p>Maximum rates of adoption for late blight-resistant varieties in the 32 target countries range from 10 to 60% and the expected yield gains range from 12 to 32%. This technology also entails cost reductions as a result of reduced use of fungicides. However, anecdotal evidence in some regions where late blight-resistant varieties are already cultivated reveals that farmers tend to keep spraying as if the varieties were susceptible. Therefore, the 5% cost reductions assumed reflect only a small proportion of the full potential that can be achieved in experimental trials. Since late blight-resistant varieties have been an important part of CIP's research programs for a long time, a probability of success of 80% is given to this research option.</p> <hd id="AN0128889675-10">3. Model simulations</hd> <p></p> <hd id="AN0128889675-11">3.1 Simulating the adoption and the productivity impacts of new technologies</hd> <p>To simulate the effects of the selected crop productivity improvement technologies in the IMPACT model, different technologies can be defined at the FPU level. Each technology can occupy a specific share of the total production area dedicated every year to a particular crop in each FPU, and its adoption is modeled through the implementation of a logistic or linear diffusion curve which can be specified and parameterized prior to running the model. To simulate the yield advantage of a new technology over an old one, we appropriately modify the crop yield response functions included in the model. More precisely, in every FPU <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>n</mi></math> </ephtml></p> <p>, the yield equation for crop <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>j</mi></math> </ephtml></p> <p>and land type <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>k</mi></math> </ephtml></p> <p>(irrigated or rain-fed) for each year is expressed as:</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>Y</mi></mrow><mrow><mi>j</mi><mo>,</mo><mi>n</mi><mo>,</mo><mi>k</mi></mrow></msub><mtext>=</mtext><msub><mrow><mi>Y</mi><mi>I</mi><mi>n</mi><mi>t</mi></mrow><mrow><mi>j</mi><mo>,</mo><mi>n</mi><mo>,</mo><mi>k</mi></mrow></msub><mtext>×</mtext><msub><mrow><mi>Y</mi><mi>I</mi><mi>n</mi><mi>t</mi><mn>2</mn></mrow><mrow><mi>j</mi><mo>,</mo><mi>n</mi><mo>,</mo><mi>k</mi></mrow></msub><mtext>×</mtext><msub><mrow><mi>Y</mi><mi>W</mi><mi>a</mi><mi>t</mi></mrow><mrow><mi>j</mi><mo>,</mo><mi>n</mi><mo>,</mo><mi>k</mi></mrow></msub><mtext>×</mtext><msub><mrow><mi>Y</mi><mi>C</mi><mi>l</mi><mi>i</mi><mi>m</mi></mrow><mrow><mi>j</mi><mo>,</mo><mi>n</mi><mo>,</mo><mi>k</mi></mrow></msub><mtext>×</mtext><msup><mrow><mo>(</mo><msub><mrow><mi>P</mi><mi>S</mi></mrow><mrow><mi>j</mi><mo>,</mo><mi>c</mi></mrow></msub><mo>)</mo></mrow><mrow><mi>ɛ</mi></mrow></msup><mtext>×</mtext><msup><mrow><mi>P</mi><mi>F</mi></mrow><mrow><mi>η</mi></mrow></msup><mo>,</mo></math> </ephtml> </p> <p>(<reflink idref="bib1" id="ref46">1</reflink>) where <emph>YInt</emph> is base year crop yield reported in FAOSTAT (in tons per ha), <emph>YInt2</emph> is an exogenous crop yield growth parameter, <emph>YWat</emph> is the exogenous yield shock from water stress (water availability), <emph>YClim</emph> is the exogenous yield shock from climate change (increased temperature), <emph>PS</emph> is the crop net price mapped to country <emph>c</emph>,[<reflink idref="bib4" id="ref47">4</reflink>]<emph>ɛ</emph> denotes the price response with respect to net price, <emph>PF</emph> represents prices of inputs, and <emph>η</emph> is the yield supply elasticity with respect to input prices. The impact from water and climate shocks on yields is estimated with the SUBSTOR potato growth model ([<reflink idref="bib11" id="ref48">11</reflink>]) in the DSSAT (Decision Support System for Agrotechnology Transfer) farming system simulator ([<reflink idref="bib20" id="ref49">20</reflink>]), which is linked to the IMPACT modeling suite. All parameters in Equation (<reflink idref="bib1" id="ref50">1</reflink>) are specified as percentage changes over <emph>YInt</emph>. For example, when climate and water shocks are not considered, <emph>YWat</emph> and <emph>YClim</emph> take a value of one. More details about the yield equation and the how the different types of shock are defined in IMPACT can be found in [<reflink idref="bib30" id="ref51">30</reflink>].</p> <p>The four potato technology options modeled here not only lead to changes in yield but also include changes in input costs and are contingent on the probability of success of the research efforts undertaken. In traditional economic surplus models, such as those proposed by [<reflink idref="bib1" id="ref52">1</reflink>], the proportional shift in the supply curve ( <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>k</mi></mrow><mrow><mi>t</mi></mrow></msub></math> </ephtml></p> <p>) caused by a new technology in each target country at any year <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml></p> <p>is calculated as:</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>k</mi></mrow><mrow><mi>t</mi></mrow></msub><mtext>=</mtext><mfrac><mrow><mtext>Δ</mtext><mi>Y</mi></mrow><mrow><msub><mrow><mi>ɛ</mi></mrow><mrow><mi>s</mi></mrow></msub></mrow></mfrac><mo>-</mo><mfenced separators="|"><mrow><mfrac><mrow><mtext>Δ</mtext><mi>C</mi></mrow><mrow><mn>1</mn><mtext>+Δ</mtext><mi>Y</mi></mrow></mfrac></mrow></mfenced><mtext>×</mtext><mi>p</mi><mtext>×</mtext><msub><mrow><mi>A</mi></mrow><mrow><mi>t</mi></mrow></msub></math> </ephtml> </p> <p>(<reflink idref="bib2" id="ref53">2</reflink>) where <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>Δ</mi><mi>Y</mi></math> </ephtml></p> <p>is the percentage maximum yield change, <emph>ɛ</emph><subs>s</subs> is the price elasticity of supply, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>Δ</mi><mi>C</mi></math> </ephtml></p> <p>is the percentage input cost change, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>p</mi></math> </ephtml></p> <p>is the probability of success in research, and <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>A</mi></mrow><mrow><mi>t</mi></mrow></msub></math> </ephtml></p> <p>is the adoption rate at year <emph>t</emph>, which is given by the logistic function:</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>A</mi></mrow><mrow><mi>t</mi></mrow></msub><mi /><mo>=</mo><mi /><mfrac><mrow><msub><mi>A</mi><mi>max</mi></msub></mrow><mrow><mn>1</mn><mo>+</mo><msup><mrow><mi>e</mi></mrow><mrow><mo>-</mo><mo>(</mo><mi>α</mi><mo>+</mo><mtext>β</mtext><mi>t</mi><mo>)</mo></mrow></msup></mrow></mfrac><mi /><mtext>for</mtext><mi /><mi>t</mi><mtext>=</mtext><mn>1</mn><mo>,</mo><mn>2</mn><mo>,</mo><mtext>...</mtext><mo>,</mo><mi>T</mi><mo>.</mo></math> </ephtml> </p> <p>(<reflink idref="bib3" id="ref54">3</reflink>)</p> <p>In the above equation, <emph>A</emph><subs>max</subs> is the maximum adoption rate, whereas <emph>α</emph> and <emph>β</emph> are parameters estimated using two points in the adoption curve, namely, <emph>A<subs>0</subs></emph> and <emph>A</emph><subs>mid</subs> which correspond to the initial and the 50% adoption rate occurring at years <emph>t<subs>0</subs></emph> (year of initial adoption) and <emph>t</emph><subs>mid</subs>, respectively.</p> <p>Although <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>k</mi></mrow><mrow><mi>t</mi></mrow></msub></math> </ephtml></p> <p>adequately describes the supply shift in an economic surplus model, it cannot be applied in IMPACT because the supply shift itself is endogenously calculated when determining the market equilibrium and also incorporates feedback effects from other related commodity markets (cross-price effects). The same applies to the adoption rate and is specified prior to running the model. For these reasons, instead of a supply shift, we focus on modifying parameter <emph>YInt2</emph> in Equation (<reflink idref="bib1" id="ref55">1</reflink>) which represents a non-price-induced exogenous yield trend for a 'baseline' or 'business as usual' scenario of productivity growth. Any factor that modifies—additively or multiplicatively—the initial <emph>YInt2</emph> values corresponds to a scenario of accelerated growth in yields, such as that brought about by technological innovations from agricultural research. Given that the research-induced yield growth is conditional on the probability of success and needs to take into account possible cost changes, we calculated the percentage change of <emph>YInt2</emph> (denoted by Δ<emph>YInt2</emph>) with the following formula, which derives directly from the definition of <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi>k</mi></mrow><mrow><mi>t</mi></mrow></msub></math> </ephtml></p> <p>in Equation (<reflink idref="bib2" id="ref56">2</reflink>):</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>Δ</mi><mi>Y</mi><mi>I</mi><mi>n</mi><mi>t</mi><mn>2</mn><mtext>=Δ</mtext><mi>Y</mi><mo>-</mo><mfenced separators="|"><mrow><mfrac><mrow><mtext>Δ</mtext><mi>C</mi></mrow><mrow><mn>1</mn><mtext>+Δ</mtext><mi>Y</mi></mrow></mfrac></mrow></mfenced><mtext>×</mtext><mi>p</mi><mo>.</mo></math> </ephtml> </p> <p>(<reflink idref="bib4" id="ref57">4</reflink>)</p> <p>By introducing an increase in productivity at the release year of the new technology, IMPACT simulates a new supply and demand equilibrium for the entire adoption period. In modeling terms, this one-time increase in productivity is sustained over time and translates to an upward shift of the projected yield growth over and above the growth which is already assumed by the model (expressed by <emph>YInt2</emph>). The difference between Equations (<reflink idref="bib2" id="ref58">2</reflink>) and (<reflink idref="bib4" id="ref59">4</reflink>) is that the former represents an outward shift in supply in economic surplus models, whereas Equation (<reflink idref="bib4" id="ref60">4</reflink>) represents an upward shift in the projected yield growth trend in IMPACT which finally leads to an endogenously calculated supply increase when running the model.</p> <hd id="AN0128889675-12">3.2 Foresight scenarios</hd> <p>To compare alternative approaches, one static and three foresight scenarios were specified for each of the four potato research options. The initial year of the simulations for all scenarios was 2015, and the length of the simulation period was 25 years, i.e. until 2040, which included the research and dissemination phases for each technology, as defined in Section 2.</p> <p>The foresight scenarios derive from combinations of Representative Concentration Pathways (RCPs) and Shared Socioeconomic Pathways (SSPs) adopted by the IPCC. RCPs represent alternative trajectories on emissions and concentrations of greenhouse gases, and they are defined according to the radiative forcing projected by the end the 21st century ([<reflink idref="bib37" id="ref61">37</reflink>]). The climate projections from each RCP are performed with Earth System Models (ESMs) at various spatial and temporal resolutions and are used as data inputs for other applications, such as in crop modeling. SSPs are characterized by different challenges for mitigation and adaptation to climate change, and their quantification is based on assumed changes in income, population, and urbanization (among other variables) for most countries in the world ([<reflink idref="bib25" id="ref62">25</reflink>]).</p> <p>The RCP-SSP combinations selected for this article draw on the compatibility matrix approach which has been proposed for developing scenarios in climate change analyses ([<reflink idref="bib38" id="ref63">38</reflink>]). More specifically, they are based on SSPs 1, 2, and 3 which represent pathways of low, medium, and high overall socioeconomic challenges, respectively. SSP1 is a 'sustainability pathway' with low population growth rates, high economic growth, and high levels of R&D investments that result in more environment-friendly technologies. SPP2 is a 'middle of the road' pathway that generally follows historic trends, while SSP3 is a more pessimistic scenario with higher population growth rates and more uneven economic growth ([<reflink idref="bib25" id="ref64">25</reflink>]). In this exercise, we consider SSP2 to be the benchmark pathway for which the initially assumed maximum adoption rates apply. To be more consistent with the assumptions of the different SSPs, <emph>A</emph><subs>max</subs> has been increased by 10% under SSP1 for all countries and technologies to account for higher R&D investments, whereas for SSP3 it was reduced by a similar percentage.</p> <p>Each of the above SSPs can be combined with a different RCP to produce a consistent foresight scenario and to represent different levels of mitigation and adaptation costs. For this article, we selected RCP-SSP combinations that correspond to similar (medium) mitigation costs: SSP1-RCP4.5; SSP2-RCP6.0; and SSP3-RCP8.5.[<reflink idref="bib5" id="ref65">5</reflink>] Among the three RCPs examined, RCP8.5 is the most extreme climate pathway, as it posits high green house gas emissions without mitigation policies, whereas RCP4.5 is a more optimistic pathway which assumes mitigation policies that lead to lower overall emissions. The same SSP-RCP combinations have also been used by [<reflink idref="bib39" id="ref66">39</reflink>] in their analysis of the global climate change impacts on agriculture by 2050. Specific climate data for deriving the crop yield shock parameter <emph>YClim</emph> in Equation (<reflink idref="bib1" id="ref67">1</reflink>) with DSSAT were retrieved from the ESM model HadGEM2 (Hadley Centre Global Environment Model version 2) ([<reflink idref="bib19" id="ref68">19</reflink>]).</p> <p>For the static scenario, we set a zero-growth rate to all parameters that change over time and are either linked to RCP and SSP assumptions (population, income, and climate-induced yield shocks) or are intrinsic to the model structure (e.g. exogenous yield growth). Thus, the static scenario approximated an economic surplus analysis with cross-price effects included in a multi‐market structure.</p> <p>When modeling multiple commodity markets, the literature suggests that welfare effects can be captured by examining the market in which the distortions appear, since the equilibrium will also reflect changes in all other vertically or horizontally related markets ([<reflink idref="bib21" id="ref69">21</reflink>]). Therefore, to avoid double-counting ([<reflink idref="bib1" id="ref70">1</reflink>]), we estimated the expected welfare changes by focusing only on the potato market. Welfare changes from every technology in each scenario were estimated by comparing the outcomes against a baseline that involved an initial model run under the same scenario but without any productivity shifts. The baseline for the static scenario produced the same result for every year of the adoption period, since all dynamic elements of the model were held constant. The results were then aggregated (<reflink idref="bib1" id="ref71">1</reflink>) for all countries and (<reflink idref="bib2" id="ref72">2</reflink>) for target countries only, and we estimated the net present value (NPV) of the net welfare benefits (sum of consumer and producer surplus minus research and dissemination costs), the IRR, and the modified internal rate of return (MIRR) of each research option with both aggregation methods.[<reflink idref="bib6" id="ref73">6</reflink>]</p> <hd id="AN0128889675-13">4. Results</hd> <p>Table 2 reveals that the ordering of the different technologies in terms of net benefits is the same across country groups for all four scenarios, and it is also identical to what is reported in the original RTB study ([<reflink idref="bib12" id="ref74">12</reflink>]); late blight-resistant varieties produce the highest welfare benefits, followed by virus resistance, bacterial wilt resistance, and improved seed systems.[<reflink idref="bib7" id="ref75">7</reflink>] The ordering of the technologies according to their MIRR is slightly different, yet still consistent across all country groups and scenarios, with the virus-resistant varieties yielding the highest rate of return. In contrast, the ordering of the technologies according to their IRR is not the same between country groups, since improved seed systems produce a higher IRR than bacterial wilt-resistant varieties when only target countries are considered. The previous results suggest that the ordering of the different technologies, according to the three measures of investment performance used in this article, is independent of the foresight scenario examined, but it may be affected by the aggregation method employed.[<reflink idref="bib8" id="ref76">8</reflink>]</p> <p>Global economic impacts and benefit–cost results of potato research options under different scenarios</p> <p> <ephtml> <table><tr><th rowspan="2">Benefit-cost indicators </th><th colspan="4">All countries</th><th colspan="4">Target countries</th></tr><tr><th>Static </th><th>SSP1-RCP4.5 </th><th>SSP2-RCP6.0 </th><th>SSP3-RCP8.5 </th><th>Static </th><th>SSP1-RCP4.5 </th><th>SSP2-RCP6.0 </th><th>SSP3-RCP8.5 </th></tr><tr><td colspan="9">Improved seed systems </td></tr><tr><td> Producer surplus </td><td>−264 </td><td>−561 </td><td>−504 </td><td>−493 </td><td>158 </td><td>279 </td><td>258 </td><td>230 </td></tr><tr><td> Consumer surplus </td><td>300 </td><td>790 </td><td>696 </td><td>687 </td><td>127 </td><td>369 </td><td>328 </td><td>324 </td></tr><tr><td> Net welfare benefits </td><td>−24 </td><td>156 </td><td>125 </td><td>131 </td><td>226 </td><td>575 </td><td>517 </td><td>490 </td></tr><tr><td> IRR </td><td>0.04 </td><td>0.27 </td><td>0.25 </td><td>0.25 </td><td>0.42 </td><td>0.60 </td><td>0.58 </td><td>0.57 </td></tr><tr><td> MIRR </td><td>0.07 </td><td>0.16 </td><td>0.15 </td><td>0.17 </td><td>0.20 </td><td>0.26 </td><td>0.25 </td><td>0.25 </td></tr><tr><td colspan="9">Bacterial wilt-resistant varieties </td></tr><tr><td> Producer surplus </td><td>−378 </td><td>−916 </td><td>−785 </td><td>−782 </td><td>198 </td><td>352 </td><td>317 </td><td>262 </td></tr><tr><td> Consumer surplus </td><td>529 </td><td>1,450 </td><td>1,240 </td><td>1,253 </td><td>202 </td><td>686 </td><td>586 </td><td>592 </td></tr><tr><td> Net welfare benefits </td><td>111 </td><td>488 </td><td>411 </td><td>427 </td><td>361 </td><td>992 </td><td>859 </td><td>811 </td></tr><tr><td> IRR </td><td>0.22 </td><td>0.33 </td><td>0.32 </td><td>0.32 </td><td>0.32 </td><td>0.40 </td><td>0.39 </td><td>0.39 </td></tr><tr><td> MIRR </td><td>0.17 </td><td>0.23 </td><td>0.23 </td><td>0.23 </td><td>0.22 </td><td>0.27 </td><td>0.26 </td><td>0.26 </td></tr><tr><td colspan="9">Virus-resistant varieties </td></tr><tr><td> Producer surplus </td><td>−1,923 </td><td>−3757 </td><td>−3,212 </td><td>−3,121 </td><td>926 </td><td>1,367 </td><td>1,222 </td><td>1,015 </td></tr><tr><td> Consumer surplus </td><td>2,847 </td><td>6,110 </td><td>5,204 </td><td>5,110 </td><td>1,138 </td><td>2,923 </td><td>2,496 </td><td>2,461 </td></tr><tr><td> Net welfare benefits </td><td>892 </td><td>2,316 </td><td>1,957 </td><td>1,957 </td><td>2,032 </td><td>4,253 </td><td>3,684 </td><td>3,444 </td></tr><tr><td> IRR </td><td>0.70 </td><td>0.92 </td><td>0.88 </td><td>0.88 </td><td>0.98 </td><td>1.18 </td><td>1.14 </td><td>1.12 </td></tr><tr><td> MIRR </td><td>0.29 </td><td>0.33 </td><td>0.33 </td><td>0.33 </td><td>0.34 </td><td>0.38 </td><td>0.37 </td><td>0.37 </td></tr><tr><td colspan="9">Late blight-resistant varieties </td></tr><tr><td> Producer surplus </td><td>−2,868 </td><td>−6,149 </td><td>−5,366 </td><td>−5353 </td><td>1,202 </td><td>1,813 </td><td>1,672 </td><td>1,373 </td></tr><tr><td> Consumer surplus </td><td>4,187 </td><td>9,584 </td><td>8,366 </td><td>8,420 </td><td>1,640 </td><td>4,610 </td><td>4,020 </td><td>4,061 </td></tr><tr><td> Net welfare benefits </td><td>1,254 </td><td>3,358 </td><td>2,928 </td><td>2,999 </td><td>2,776 </td><td>6,346 </td><td>5,620 </td><td>5,366 </td></tr><tr><td> IRR </td><td>0.62 </td><td>0.84 </td><td>0.81 </td><td>0.81 </td><td>0.87 </td><td>1.07 </td><td>1.04 </td><td>1.03 </td></tr><tr><td> MIRR </td><td>0.27 </td><td>0.32 </td><td>0.31 </td><td>0.32 </td><td>0.32 </td><td>0.36 </td><td>0.36 </td><td>0.35 </td></tr></table> </ephtml> </p> <p>3 aResults are calculated as changes over the baseline (IMPACT simulations for each SSP-RCP combination without the enhanced technologies); all welfare measures are expressed in NPVs, calculated with a discount rate of 10% and measured in million US$ at 2005 constant prices.</p> <p>Although changes in consumer surplus are unambiguously positive for all technologies everywhere, potato producers in non-target countries suffer significant surplus losses because they face lower world prices without the benefit of higher yields. Lower prices are a disincentive for potato production and lead to decreases in yields and in land allocated to potato cultivation (Table 3). For target countries, the technology-induced supply increase is greater in effect than the decrease in world prices, leading to a positive producer surplus change. The difference in net benefits among country groups is particularly pronounced for the improved seed systems technology, and it is the principal reason for the inconsistent ordering of the different research options according to their IRR. Furthermore, improved seed systems are assumed to have short research and adoption lags, increased production costs, and the highest dissemination costs among all four technologies. These assumptions lead to the concentration of negative cash flows early in the beginning of the simulated 25-year period, thus yielding lower net benefits and IRR values compared to technologies with more uniformly distributed costs.</p> <p>Impacts of improved technologies on world potato production in 2040 (% change from baseline)</p> <p> <ephtml> <table><tr><th>Production indicators </th><th>Improved seed systems </th><th>Bacterial wilt- resistant varieties </th><th>Virus-resistant varieties </th><th>Late blight- resistant varieties </th></tr><tr><td colspan="5">Static scenario </td></tr><tr><td colspan="5"> Yield </td></tr><tr><td>Target countries</td><td>0.20 </td><td>0.78 </td><td>1.52 </td><td>2.03 </td></tr><tr><td>Non-target countries</td><td>−0.02 </td><td>−0.11 </td><td>−0.20 </td><td>−0.29 </td></tr><tr><td>Global</td><td>0.05 </td><td>0.16 </td><td>0.41 </td><td>0.52 </td></tr><tr><td> Land </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.04 </td><td>0.23 </td><td>0.25 </td><td>0.49 </td></tr><tr><td>Non-target countries</td><td>−0.05 </td><td>−0.21 </td><td>−0.39 </td><td>−0.56 </td></tr><tr><td>Global</td><td>−0.01 </td><td>−0.02 </td><td>−0.10 </td><td>−0.07 </td></tr><tr><td> Supply </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.24 </td><td>1.01 </td><td>1.77 </td><td>2.53 </td></tr><tr><td>Non-target countries</td><td>−0.07 </td><td>−0.31 </td><td>−0.59 </td><td>−0.84 </td></tr><tr><td>Global</td><td>0.04 </td><td>0.18 </td><td>0.31 </td><td>0.44 </td></tr><tr><td colspan="5">SSP1-RCP4.5 </td></tr><tr><td colspan="5"> Yield </td></tr><tr><td>Target countries</td><td>0.24 </td><td>0.86 </td><td>1.43 </td><td>2.31 </td></tr><tr><td>Non-target countries</td><td>−0.04 </td><td>−0.15 </td><td>−0.24 </td><td>−0.41 </td></tr><tr><td>Global</td><td>0.07 </td><td>0.25 </td><td>0.48 </td><td>0.74 </td></tr><tr><td> Land </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.03 </td><td>0.23 </td><td>0.18 </td><td>0.40 </td></tr><tr><td>Non-target countries</td><td>−0.07 </td><td>−0.30 </td><td>−0.48 </td><td>−0.79 </td></tr><tr><td>Global</td><td>−0.02 </td><td>−0.03 </td><td>−0.12 </td><td>−0.14 </td></tr><tr><td> Supply </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.27 </td><td>1.09 </td><td>1.61 </td><td>2.72 </td></tr><tr><td>Non-target countries</td><td>−0.11 </td><td>−0.45 </td><td>−0.72 </td><td>−1.20 </td></tr><tr><td>Global</td><td>0.05 </td><td>0.22 </td><td>0.36 </td><td>0.61 </td></tr><tr><td colspan="5">SSP2-RCP6.0 </td></tr><tr><td colspan="5"> Yield </td></tr><tr><td>Target countries</td><td>0.24 </td><td>0.80 </td><td>1.23 </td><td>2.11 </td></tr><tr><td>Non-target countries</td><td>−0.04 </td><td>−0.14 </td><td>−0.21 </td><td>−0.37 </td></tr><tr><td>Global</td><td>0.07 </td><td>0.23 </td><td>0.39 </td><td>0.66 </td></tr><tr><td> Land </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.03 </td><td>0.21 </td><td>0.19 </td><td>0.38 </td></tr><tr><td>Non-target countries</td><td>−0.04 </td><td>−0.27 </td><td>−0.41 </td><td>−0.71 </td></tr><tr><td>Global</td><td>−0.02 </td><td>−0.03 </td><td>−0.08 </td><td>−0.11 </td></tr><tr><td> Supply </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.26 </td><td>1.01 </td><td>1.42 </td><td>2.50 </td></tr><tr><td>Non-target countries</td><td>−0.10 </td><td>−0.41 </td><td>−0.62 </td><td>−1.07 </td></tr><tr><td>Global</td><td>0.05 </td><td>0.20 </td><td>0.31 </td><td>0.54 </td></tr><tr><td colspan="5">SSP3-RCP8.5 </td></tr><tr><td colspan="5"> Yield </td></tr><tr><td>Target countries</td><td>0.20 </td><td>0.70 </td><td>1.05 </td><td>1.89 </td></tr><tr><td>Non-target countries</td><td>−0.03 </td><td>−0.13 </td><td>−0.20 </td><td>−0.36 </td></tr><tr><td>Global</td><td>0.07 </td><td>0.23 </td><td>0.38 </td><td>0.68 </td></tr><tr><td> Land </td><td /><td /><td /><td /></tr><tr><td>Target countries</td><td>0.02 </td><td>0.16 </td><td>0.14 </td><td>0.28 </td></tr><tr><td>Non-target countries</td><td>−0.06 </td><td>−0.26 </td><td>−0.39 </td><td>−0.69 </td></tr><tr><td>Global</td><td>−0.02 </td><td>−0.04 </td><td>−0.09 </td><td>−0.15 </td></tr><tr><td colspan="5"> Supply </td></tr><tr><td>Target countries</td><td>0.23 </td><td>0.87 </td><td>1.19 </td><td>2.18 </td></tr><tr><td>Non-target countries</td><td>−0.10 </td><td>−0.39 </td><td>−0.58 </td><td>−1.05 </td></tr><tr><td>Global</td><td>0.05 </td><td>0.20 </td><td>0.29 </td><td>0.53 </td></tr></table> </ephtml> </p> <p>The static scenario also produces negative net benefits for the improved seed systems technology when all countries are considered. Although from an investment viewpoint this result may suggest that improved potato seed systems, as specified in this scenario, are not a reasonable research option, it should be viewed with caution, as it contradicts common agronomic knowledge that low-quality seed constrains the expression of the crop's full yielding ability ([<reflink idref="bib14" id="ref77">14</reflink>]). In fact, negative net benefits are a direct consequence of the underestimation of the welfare impacts of all technologies under the static scenario. More precisely, Table 2 shows that the disparity in net benefits is mainly driven by consumer surplus, which in the static scenario is less than half of those seen in each of the foresight scenarios and can be explained by the static demand which ignores the expected growth in population and income. Similar results are also observed on the production side, namely, for potato yields, harvested land, and total supply in non-target countries, for which the static scenario generally underestimates the negative impacts of the new potato technologies (Table 3). As a consequence, it leads to smaller producer surplus losses compared to the foresight scenarios when all countries are considered.</p> <p>All technology alternatives analyzed in this article demonstrate higher areas of adoption under SSP3-RCP8.5 and consequently exhibit higher dissemination costs (Table 4). Since the adoption rates for each technology are exogenously defined as a percentage of the total potato harvested area, this result also implies that the area allocated to potato in target countries under SSP3-RCP8.5 is higher than in other scenarios. The same scenario also yields the highest net benefits for all technologies, except for virus-resistant varieties which produce higher returns under SSP1-RCP4.5. More specifically, RCP8.5 is the driest climate change scenario among the three examined and consequently leads to lower overall yields and smaller producer surplus changes in target countries compared to the other two scenarios. At the same time, SSP3 represents the highest population growth among all SSPs (but also the slowest income growth) and thus results in higher demand for staple agricultural commodities. These assumptions are reflected in the higher consumer surpluses observed under SSP3-RCP8.5. In contrast, SSP1-RCP4.5 represents a less rapid climate change scenario[<reflink idref="bib9" id="ref78">9</reflink>] and assumes the lowest population increase among all SSPs, thus leading to opposite impacts than SSP3-RCP8.5.</p> <p>Adoption of improved potato varieties</p> <p> <ephtml> <table><tr><th>Adoption indicators </th><th>Improved seed systems </th><th>Bacterial wilt- resistant varieties </th><th>Virus-resistant varieties </th><th>Late blight- resistant varieties </th></tr><tr><td colspan="5">Dissemination costs (million US$)</td></tr><tr><td> Static scenario </td><td>37.7 </td><td>12.8 </td><td>16.6 </td><td>35.1 </td></tr><tr><td> SSP1-RCP4.5 </td><td>50.9 </td><td>19.3 </td><td>21.5 </td><td>46.4 </td></tr><tr><td> SSP2-RCP6.0 </td><td>46.2 </td><td>17.4 </td><td>19.3 </td><td>41.9 </td></tr><tr><td> SSP3-RCP8.5 </td><td>42.2 </td><td>16.0 </td><td>17.6 </td><td>38.3 </td></tr><tr><td colspan="5">Total costs (million US$)</td></tr><tr><td> Static scenario </td><td>59.6 </td><td>39.9 </td><td>31.9 </td><td>65.6 </td></tr><tr><td> SSP1-RCP4.5 </td><td>72.8 </td><td>46.4 </td><td>36.7 </td><td>77.0 </td></tr><tr><td> SSP2-RCP6.0 </td><td>68.0 </td><td>44.4 </td><td>34.6 </td><td>72.5 </td></tr><tr><td> SSP3-RCP8.5 </td><td>64.1 </td><td>43.0 </td><td>32.9 </td><td>68.9 </td></tr><tr><td colspan="5">Total adoption area (thousand hectares) </td></tr><tr><td> Static scenario </td><td>774.4 </td><td>1,020.8 </td><td>618.0 </td><td>1,303.3 </td></tr><tr><td> SSP1-RCP4.5 </td><td>1,044.0 </td><td>1,614.4 </td><td>827.5 </td><td>1,798.8 </td></tr><tr><td> SSP2-RCP6.0 </td><td>963.8 </td><td>1,445.3 </td><td>742.3 </td><td>1,617.7 </td></tr><tr><td> SSP3-RCP8.5 </td><td>865.8 </td><td>1,330.4 </td><td>675.4 </td><td>1,483.4 </td></tr></table> </ephtml> </p> <p>4 aReported results are NPVs expressed at 2005 constant prices and calculated with a discount rate of 10%.</p> <p>Although our results reveal substantial differences in the net benefits among the various scenarios, the parameters used in the specification of the technologies lead to rather limited supply growth in target countries, which ranges from 0.23% (improved seed systems—SSP3-RCP8.5) to 2.72% for the late blight resistance technology in the SSP1-RCP4.5 scenario (Table 3). Except for China and India, target countries are responsible for only a small share of the global potato production, and therefore the resulting decrease in world prices is also small and estimated at around 0.1–1.8%, depending on the technology and the scenario examined (Table 5). The above changes in the potato sector are also expected to alter the allocation of land among different activities, shift total supply, and modify the clearing prices for other agricultural commodities. The new equilibrium is determined by changes in relative prices (substitution effects) and the increase in available income due to a fall of consumer expenditures on potato as a result of its lower price. Table 5 also presents the world price changes for rice and wheat, which are crops commonly related to potato either in consumption (rice in Asia) or in production (potato and winter cereals are grown at the same time of the year, although they do not necessarily compete for the same land). Our results show that the price feedback effects are small, which can be explained by the relatively smaller size of the potato sector compared to cereals that dominate agricultural production in target countries.[<reflink idref="bib10" id="ref79">10</reflink>] Therefore, the role of foresight on cross-price effects depends on the relative magnitude of the examined intervention.</p> <p>Impacts of potato technologies on world producer prices of potato and related commodities (% change from baseline)</p> <p> <ephtml> <table><tr><th>Commodities </th><th>Improved seed systems </th><th>Bacterial wilt- resistant varieties </th><th>Virus-resistant varieties </th><th>Late blight- resistant varieties </th></tr><tr><td colspan="5">Static scenario </td></tr><tr><td> Potato </td><td>−0.11 </td><td>−0.48 </td><td>−0.91 </td><td>−1.30 </td></tr><tr><td> Rice </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td> Wheat </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td colspan="5">SSP1-RCP4.5 </td></tr><tr><td> Potato </td><td>−0.16 </td><td>−0.66 </td><td>−1.07 </td><td>−1.78 </td></tr><tr><td> Rice </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td> Wheat </td><td /><td>−0.01 </td><td>−0.01 </td><td>−0.01 </td></tr><tr><td colspan="5">SSP2-RCP6.0 </td></tr><tr><td> Potato </td><td>−0.15 </td><td>−0.60 </td><td>−0.92 </td><td>−1.59 </td></tr><tr><td> Rice </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td> Wheat </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td colspan="5">SSP3-RCP8.5 </td></tr><tr><td> Potato </td><td>−0.14 </td><td>−0.58 </td><td>−0.86 </td><td>−1.55 </td></tr><tr><td> Rice </td><td /><td /><td>−0.01 </td><td>−0.01 </td></tr><tr><td> Wheat </td><td /><td>−0.01 </td><td>−0.01 </td><td>−0.01 </td></tr></table> </ephtml> </p> <p>5 aLower than −0.01%.</p> <hd id="AN0128889675-14">5. Discussion</hd> <p>Our analysis reveals that introducing dynamics in the form of foresight scenarios in this priority setting exercise does not affect the ranking of the examined technologies according to their expected net benefits or either measure of rate of return to investment. In fact, the ordering of the different potato research options is identical to the original RTB study which was based on a simple economic surplus model. Therefore, foresight considerations in assessing research options become irrelevant in this case, if the objective is simply to provide an ordinal comparison of the returns from investing in different research options. However, foresight scenarios greatly affect the estimated absolute level of welfare impacts. The discrepancy in the estimated net benefits between the static and the foresight scenarios also suggests that simple economic surplus models may not always be appropriate for setting priorities and allocating resources because such analyses typically rely on the NPV of net benefits of each candidate activity. Given that the welfare impacts of each examined technology vary significantly among the three foresight scenarios, there seems to be scope in including foresight considerations in priority setting exercises depending on the research objective, especially when the uncertainty about climate change needs to be explicitly accounted for.</p> <p>Another aspect of the analysis sheds light on the shortcomings of the use of single market economic surplus models in priority setting, in which welfare impacts are typically calculated for target countries only. According to our results, ignoring non-target countries leads to an overestimation of net benefits by an order of almost two in all simulation scenarios. This finding highlights the 'accuracy versus cost' trade-off that economists face when carrying out such exercises, which require the collection and consolidation of data and information from various sources and across different regions and countries. It also raises concerns because priority setting studies in international agricultural research do not always consider non-target countries in their analysis. This may also have important implications for resource allocation, not only among different research activities but also among different social purposes in different world regions. Notwithstanding the obvious challenges for modeling the global agricultural sector, our results clearly show that priority setting studies using economic surplus models should nonetheless account for the global welfare impacts of the examined interventions, even if this can only be done aggregately by assuming a single market for the rest of the world. The previous postulate applies to any objective used in a priority setting study and not just to net benefits. For example, nutritional security and poverty reduction have not been considered in the present analysis, but they usually rank high in policy agendas and are key determinants of resource allocation. The consequences of a new agricultural technology on these two objectives will not be limited to target countries only but will bring about global changes in food supply which must be also taken into account.</p> <p>The impact of the new potato technologies on other related commodities has also been found to be very small, which implies that the single-commodity focus of the simple economic surplus model is an acceptable assumption for the present analysis, given how the examined potato technologies were specified. However, the added value of a multi-commodity approach, compared to the simple economic surplus model, depends on the extent to which one expects cross-price effects. The added value will therefore be small for an innovation that affects a small or isolated market but can prove important for an innovation that affects a large market with close links across countries and commodities. The major drawback of the single-commodity model concerns the assessment of impacts related to objectives that span across multiple commodity markets and cannot be well represented by price changes alone. For instance, even if individual cross-price effects are small (thus leading to similarly small changes in supply), the assessment of non-economic impacts, like nutritional security, still requires accounting for supply changes in all agricultural commodities. Therefore, the objectives and criteria included in a priority setting study determine the validity of the single-commodity assumption, even in cases where the cross-price effects of the examined interventions are expected to be limited.</p> <p>All priority setting exercises are carried out in an <emph>ex ante</emph> framework and make use of a set of assumptions regarding adoption and productivity gains. As in the case of the original RTB study, values for these parameters may come from various sources, which are usually qualitative in nature, and complement quantitative approaches that can make more assumptions explicit and examine their validity and importance in ranking research priorities. However, results are also conditional on the validity of the assumptions regarding the realization of the foresight scenarios, the calculation of the different measures of investment performance, and the representation of the global multi-market equilibrium with the IMPACT model. For example, we have assumed different maximum adoption rates for the three SSPs to account for the different underlying socioeconomic conditions that may affect technology dissemination and adoption. However, modeling results not reported in this article reveal that the ordering of the technologies remains unaffected even if the maximum adoption rates for each technology are instead kept constant across all SSPs.</p> <p>We have implicitly assumed that there are no technological spillovers between countries and regions; yet such indirect effects are possible and have already attracted much attention in the relevant literature. Technological spillovers could mitigate partially—or fully—the negative impacts on potato production in non-target countries but could also reduce the positive impacts in target countries. In either case, the direction of changes in producer surplus would still depend on the relative magnitude of productivity gains and price decreases. Introducing spatial technology spillovers in IMPACT is straightforward and may be part of future work that also extends the present analysis to other commodities.</p> <p>Another set of assumptions concerns the numerical value of some key parameters in the IMPACT model which are subject to uncertainty. For example, we have introduced a productivity shift by modifying the exogenous yield growth rates in the model, which themselves are estimated from historical trends and adjusted according to expert judgment on regional potential for development of different commodities. We have used a single ESM to quantify the different RCPs in terms of yield losses, although other ESMs would probably produce different welfare results. Past experience, however, has shown that scenarios of alternative technologies usually give consistent results across both RCPs and SSPs in terms of the direction and relative magnitude of impacts. The size, complexity, and deterministic nature of IMPACT do not easily allow in-depth examination of model solution robustness across a broad set of parameters. For this reason, the definition of multiple consistent foresight scenarios can also serve as a sensitivity analysis exercise which can produce a band of plausible welfare outcomes for testing some key modeling assumptions as a means to improve our understanding of the role of selected dynamic parameters in a foresight study.</p> <p>Our analytical framework does not capture all possible welfare effects because it does not provide any information about possible impacts on other economic sectors outside of agriculture, for example the starch processing industry. Further work with general equilibrium approaches is required to examine the economy-wide effects of introducing new crop technologies and to test whether such elaborate models can better inform priority setting exercises. The shortcoming of computable general equilibrium (CGE) models is that they usually include an aggregate representation of the agricultural sector which makes them unsuitable for examining crop-specific technologies. In contrast, sector equilibrium models like IMPACT can distinguish between multiple crop and livestock activities and are thus able to capture spillover price effects across different commodity markets and regions, despite ignoring welfare consequences on other economic sectors. A solution to overcome the individual limitations of both types of models is to use them jointly in a sequential way. Hence, the results of agricultural sector models could be used as inputs in CGE models to estimate economy-wide impacts. An example is the effort to couple IMPACT with CGE models to tackle issues like poverty alleviation, which is difficult to answer when only focusing on the agriculture sector ([<reflink idref="bib30" id="ref80">30</reflink>]).</p> <hd id="AN0128889675-15">6. Conclusions</hd> <p>Resource allocation depends on the relative importance that decision makers and donors attach to their funding objectives but also on the available information regarding the expected impacts of research. In this article, we show how economic surplus analyses of the benefits of agricultural technologies can be incorporated in global dynamic multi-commodity models (specifically the IMPACT model). We argue that the assessment of research priorities which considers dynamics in the form of foresight scenarios may offer valuable insights on how possible biophysical and socioeconomic changes may affect the welfare impacts of a given intervention. We also show that limiting the analysis to target countries should be avoided, as it may lead to an overestimation of the expected welfare changes. On the other hand, the added value from the multi-market analysis seems to be low when the examined interventions have only small impacts on related commodity markets.</p> <p>These findings suggest that the assumptions underpinning an economic surplus model can seriously affect the estimation of the expected welfare impacts of the examined interventions. For international agricultural research, in particular, where new technologies target multiple countries and can have important direct and indirect welfare impacts, global dynamic multi-commodity models provide a more informative alternative to the economic surplus model approach. For this reason there is scope for including them in priority setting studies, although they come at additional cost; their complexity and their large set of assumptions for the representation of the global agricultural sector and the quantification of foresight scenarios may be a limiting factor for their wider use, especially when considering the simplicity of the single-commodity economic surplus model approach which provides a cost-effective way to implement priority setting in times of tightened budgets for international agricultural research. Improving the validity of these assumptions through better parameterization and stronger multi-disciplinary collaboration is therefore necessary for increasing the usefulness of such models in informing research prioritization and decision-making.</p> <hd id="AN0128889675-16">Acknowledgements</hd> <p>This work was undertaken as part of the Global Futures and Strategic Foresight project (GFSF) of the CGIAR Research Program on Policies, Institutions, and Markets (PIM). The authors would like to thank Sherman Robinson, Daniel Mason-D'Croz, and Shahnila Islam for their help in running the IMPACT model. The opinions expressed here belong to the authors and do not necessarily reflect those of PIM or CGIAR.</p> <hd id="AN0128889675-17">Funding</hd> <p>Funding support for this study was provided by the CGIAR Research Program on Policies, Institutions and Markets (PIM), the CGIAR Research Program on Roots, Tubers and Bananas (RTB), and by the Bill and Melinda Gates Foundation (BMGF) [grant number OPP1009468].</p> <ref id="AN0128889675-18"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref4" type="bt">1</bibl> <bibtext> More information about how the different surplus measures are calculated with IMPACT's welfare module can be found in Robinson et al. (2015).</bibtext> </blist> <blist> <bibl id="bib2" idref="ref1" type="bt">2</bibl> <bibtext> The RTB exercise identified two more research options, namely, the 'Development of potato value chains' and the 'Intensification of cereal-based potato systems: the <emph>agile potato</emph>'. These technologies are not considered in this article to simplify the analysis.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref33" type="bt">3</bibl> <bibtext> This section and the description of the potato research options draw on the original RTB priority assessment exercise (Hareau et al. 2014).</bibtext> </blist> <blist> <bibl id="bib4" idref="ref3" type="bt">4</bibl> <bibtext> The price parameters are country-specific and thus apply for all FPUs in the same country.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref10" type="bt">5</bibl> <bibtext> The number accompanying the selected RCPs (e.g. 8.5) represents the radiative forcing assumptions for 2100 (in Watts per m<sups>2</sups>), and can be interpreted as the intensity of the assumed change in climate.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref18" type="bt">6</bibl> <bibtext> Although the MIRR was not reported in the original RTB study, we calculate it in this article to address some of the criticism related to the use of IRR as an indicator of profitability. Specifically, the IRR assumes that the intermediate cash flows are reinvested at a discount rate equal to the IRR itself, which is considered unrealistic. In contrast, the calculation of the MIRR is based on the use of socially acceptable discount rates for borrowing and investing and leads to lower, albeit more plausible, rates of return compared to the IRR. A very comprehensive discussion about the mathematical properties, the underlying assumptions and the suitability of the IRR and MIRR for the evaluation of R&D investments can be found in Hurley et al. (2014, 2016) and Oehmke (2016).</bibtext> </blist> <blist> <bibl id="bib7" idref="ref5" type="bt">7</bibl> <bibtext> The original RTB study calculated the NPV of the expected welfare benefits with discount rates of 5 and 10%. To simplify the exposition of the results, in this article we only use a discount rate of 10%. However, simulation results not reported here due to space limitations reveal that the technology ordering does not change when a 5% discount rate is assumed instead.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref8" type="bt">8</bibl> <bibtext> Priority setting by comparing the rate of returns seldom makes sense as IRRs, and MIRRs do not consider the size effects that are shown with NPVs. The reason for presenting the IRR- and MIRR-based ranking is to identify differences between the modeling assumptions used (result aggregation between target countries versus all countries).</bibtext> </blist> <blist> <bibl id="bib9" idref="ref78" type="bt">9</bibl> <bibtext> Although RCP6.0 implies higher radiative forcing than RCP4.5 by 2100, during our simulation period (until 2040) RCP6.0 corresponds to a milder climate change pathway.</bibtext> </blist> <blist> <bibtext> On average for 2012–14, harvested potato areas in target countries amount at about 8.5% of rice areas and 13.5% of wheat areas (FAO 2017).</bibtext> </blist> </ref> <ref id="AN0128889675-19"> <title> References </title> <blist> <bibtext> Alston J., Norton G., Pardey P. (1998) Science Under Scarcity: Principles and Practice for Agricultural Research Evaluation and Priority Setting. Wallingford, Cambridge : CABI.</bibtext> </blist> <blist> <bibtext> Bantilan M. C. S., Keatinge J. D. H. (2007) 'Considerations for Determining Research Priorities: Learning Cycles and Impact Pathways', in Loebenstein G., Thottappilly G. (eds) Agricultural Research Management, pp. 37 – 64. Springer : The Netherlands.</bibtext> </blist> <blist> <bibtext> Birch P. R. J. et al. (2012) 'Crops That Feed the World 8: Potato: Are the Trends of Increased Global Production Sustainable?', Food Security, 4 / 4 : 477 – 508.</bibtext> </blist> <blist> <bibtext> Braunschweig T. (2000) Priority Setting in Agricultural Biotechnology Research Supporting Public Decisions in Developing Countries with the Analytic Hierarchy Process. Research Report No. 6. The Hague, The Netherlands: International Service for National Agricultural Research (ISNAR).</bibtext> </blist> <blist> <bibtext> Briones R. M. et al. (2008) 'Priority Setting for Research on Aquatic Resources: An Application of Modified Economic Surplus Analysis to Natural Resource Systems', Agricultural Economics, 39 / 2 : 231 – 43.</bibtext> </blist> <blist> <bibtext> Byerlee D. (2000) 'Targeting Poverty Alleviation in Priority Setting for Agricultural Research', Food Policy, 25 / 4 : 429 – 45.</bibtext> </blist> <blist> <bibtext> Davis J. S., Oram P. A., Ryan J. G. (1987) Assessment of Agricultural Research Priorities: An International Perspective. ACIAR Monograph No. 4. Canberra, Australia: Australian Centre for International Agricultural Research (ACIAR) and International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> Dey M. M., Norton G. W. (1992) 'Rice Research Priorities in Bangladesh: A Quantitative Analysis', Bangladesh Journal of Agricultural Economics, 15 / 2 : 1 – 17.</bibtext> </blist> <blist> <bibtext> FAO (2017) FAOSTAT Database, (Pubd online) < <ulink href="http://www.fao.org/faostat/en/#data/QC">http://www.fao.org/faostat/en/#data/QC</ulink> > accessed 24 Jan 2017.</bibtext> </blist> <blist> <bibtext> Gildemacher P. et al. (2009) 'Improving Potato Production in Kenya, Uganda and Ethiopia: A System Diagnosis', Potato Research, 52 : 173 – 205.</bibtext> </blist> <blist> <bibtext> Griffin T. S., Johnson B. S., Ritchie J. T. (1993) A Simulation Model for Potato Growth and Development: SUBSTOR-Potato Version 2.0. International Benchmark Sites Network for Agrotechnology Transfer (IBSNAT) Research Report Series 02. Honolulu: Department of Agronomy and Soil Science, College of Tropical Agriculture and Human Resources, University of Hawaii.</bibtext> </blist> <blist> <bibtext> Hareau G. et al. (2014) 'Strategic Assessment of Research Priorities for Potato', CGIAR Research Program on Roots, Tubers and Bananas (RTB) Working Paper No. 2014-8, (Pubd online) < <ulink href="http://www.rtb.cgiar.org">http://www.rtb.cgiar.org</ulink> > accessed 15 Jul 2016.</bibtext> </blist> <blist> <bibtext> HarvestChoice (1995) DREAM (Dynamic Research Evaluation for Management 3.1), (Pubd online) < <ulink href="http://harvestchoice.org/node/678">http://harvestchoice.org/node/678</ulink> > accessed 24 Jan 2017.</bibtext> </blist> <blist> <bibtext> Haverkort A. J., Struik P. C. (2015) 'Yield Levels of Potato Crops: Recent Achievements and Future Prospects', Field Crops Research, 182 : 76 – 85.</bibtext> </blist> <blist> <bibtext> Huang J., Rozelle S., Rosegrant M. (1999) 'China's Food Economy to the 21st Century', Economic Development and Cultural Change, 47 / 4 : 737 – 66.</bibtext> </blist> <blist> <bibtext> Hurley T. M., Rao X., Pardey P. G. (2014) 'Re-Examining the Reported Rates of Return to Food and Agricultural Research and Development', American Journal of Agricultural Economics, 96 / 5 : 1492 – 504.</bibtext> </blist> <blist> <bibtext> Hurley T. M., Rao X., Pardey P. G. (2016) 'Re-Examining the Reported Rates of Return to Food and Agricultural Research and Development: Reply', American Journal of Agricultural Economics, 99 / 3 : 827 – 36.</bibtext> </blist> <blist> <bibtext> Islam S. et al. (2016) 'Structural Approaches to Modeling the Impact of Climate Change and Adaptation Technologies on Crop Yields and Food Security', Global Food Security, 10 : 63 – 70.</bibtext> </blist> <blist> <bibtext> Jones C. D. et al. (2011) 'The HadGEM2-ES Implementation of CMIP5 Centennial Simulations', Geoscientific Model Development, 4 / 3 : 543 – 70.</bibtext> </blist> <blist> <bibtext> Jones J. W. et al. (2003) 'The DSSAT Cropping System Model', European Journal of Agronomy, 18 / 3 : 235 – 65.</bibtext> </blist> <blist> <bibtext> Just R. E., Hueth D. L., Schmitz A. (2004) The Welfare Economics of Public Policy: A Practical Approach to Project and Policy Evaluation. Northampton, MA : Edward Elgar Publishing.</bibtext> </blist> <blist> <bibtext> Kleinwechter U., Hareau G., Suarez V. (2014) 'Prioritization of Options for Potato Research for Development—Results from a Global Expert Survey', CGIAR Research Program on Roots, Tubers and Bananas (RTB) Working Paper No. 2014-7, (Pubd online) < <ulink href="http://www.rtb.cgiar.org">http://www.rtb.cgiar.org</ulink> > accessed 15 Jul 2016.</bibtext> </blist> <blist> <bibtext> Mutangadura G., Norton G. W. (1999) 'Agricultural Research Priority Setting under Multiple Objectives: An Example from Zimbabwe', Agricultural Economics, 20 / 3 : 277 – 86.</bibtext> </blist> <blist> <bibtext> Nelson G. C. et al. (2010) Food Security, Farming, and Climate Change to 2050. Washington, DC : International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> O'Neill B. C. et al. (2014) 'A New Scenario Framework for Climate Change Research: The Concept of Shared Socioeconomic Pathways', Climatic Change, 122 / 3 : 387 – 400.</bibtext> </blist> <blist> <bibtext> Oehmke J. F. (2016) 'Re-Examining the Reported Rates of Return to Food and Agricultural Research and Development: Comment', American Journal of Agricultural Economics, 99 / 3 : 818 – 26.</bibtext> </blist> <blist> <bibtext> Omamo S. W. et al. (2006) 'Strategic Priorities for Agricultural Development in Eastern and Central Africa', Research Report 150. Washington, DC : International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> Pandya-Lorch R., Rosegrant M. W. (2000) 'Prospects for Food Demand and Supply in Central Asia', Food Policy, 25 / 6 : 637 – 46.</bibtext> </blist> <blist> <bibtext> Pinstrup-Andersen P., Pandya-Lorch R., Rosegrant M. W. (1997) 'The World Food Situation: Recent Developments, Emerging Issues, and Long-Term Prospects', 2020 Vision Food Policy Report. Washington, DC : International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> Robinson S. et al. (2015). 'The International Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT). Model Description for Version 3', IFPRI Discussion Paper 01483. Washington, DC : International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> Rosegrant M. W., Cai X. (2001) 'Water Scarcity and Food Security: Alternative Futures for the 21st Century', Water Science and Technology, 43 / 4 : 61 – 70.</bibtext> </blist> <blist> <bibtext> Rosegrant M. W. et al. (2014) Food Security in a World of Natural Resource Scarcity—the Role of Agricultural Technologies. Washington, DC : International Food Policy Research Institute (IFPRI).</bibtext> </blist> <blist> <bibtext> Salazar L. F. (2006) 'Emerging and Re-Emerging Potato Diseases in the Andes', Potato Research, 49 : 43 – 7.</bibtext> </blist> <blist> <bibtext> Schumway C. R. (1977) 'Models and Methods Used to Allocate Resources in Agricultural Research: A Critical Review', in: Arndt T., Dalrymple D. G., Ruttan V. M., (eds) Resource Allocation and Productivity in National and International Agricultural Research, pp. 436 – 57. University of Minnesota Press : Minneapolis, Minnesota.</bibtext> </blist> <blist> <bibtext> Scott G., Rosegrant M., Ringler C. (2000) 'Global Projections for Root and Tuber Crops to the Year 2020', Food Policy, 25 / 5 : 561 – 97.</bibtext> </blist> <blist> <bibtext> Thiele G. et al. (2008) 'Varietal Change in Potatoes in Developing Countries and the Contribution of the International Potato Center: 1972-2007', Social Sciences Working Paper No. 2008-6. Lima, Peru: International Potato Center (CIP).</bibtext> </blist> <blist> <bibtext> Van Vuuren D. P. et al. (2011) 'The Representative Concentration Pathways: An Overview', Climatic Change, 109 / 1 : 5 – 31.</bibtext> </blist> <blist> <bibtext> Van Vuuren D. P. et al. (2014) 'A New Scenario Framework for Climate Change Research: Scenario Matrix Architecture', Climatic Change, 122 / 3 : 373 – 86.</bibtext> </blist> <blist> <bibtext> Wiebe K. et al. (2015) 'Climate Change Impacts on Agriculture in 2050 under a Range of Plausible Socioeconomic and Emissions Scenarios', Environmental Research Letters, 10 : 085010.</bibtext> </blist> <blist> <bibtext> You L., Johnson M. (2010) 'Exploring Strategic Priorities for Regional Agricultural R&D Investments in East and Central Africa', Agricultural Economics, 41 / 2 : 177 – 90.</bibtext> </blist> </ref> <aug> <p>By Athanasios Petsakos; Guy Hareau; Ulrich Kleinwechter; Keith Wiebe and Timothy B Sulser</p> </aug> <nolink nlid="nl1" bibid="bib34" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib13" firstref="ref6"></nolink> <nolink nlid="nl3" bibid="bib40" firstref="ref7"></nolink> <nolink nlid="nl4" bibid="bib23" firstref="ref9"></nolink> <nolink nlid="nl5" bibid="bib12" firstref="ref11"></nolink> <nolink nlid="nl6" bibid="bib27" firstref="ref15"></nolink> <nolink nlid="nl7" bibid="bib32" firstref="ref16"></nolink> <nolink nlid="nl8" bibid="bib18" firstref="ref17"></nolink> <nolink nlid="nl9" bibid="bib30" firstref="ref20"></nolink> <nolink nlid="nl10" bibid="bib29" firstref="ref21"></nolink> <nolink nlid="nl11" bibid="bib15" firstref="ref22"></nolink> <nolink nlid="nl12" bibid="bib28" firstref="ref23"></nolink> <nolink nlid="nl13" bibid="bib35" firstref="ref24"></nolink> <nolink nlid="nl14" bibid="bib31" firstref="ref25"></nolink> <nolink nlid="nl15" bibid="bib24" firstref="ref26"></nolink> <nolink nlid="nl16" bibid="bib22" firstref="ref36"></nolink> <nolink nlid="nl17" bibid="bib10" firstref="ref40"></nolink> <nolink nlid="nl18" bibid="bib14" firstref="ref41"></nolink> <nolink nlid="nl19" bibid="bib33" firstref="ref43"></nolink> <nolink nlid="nl20" bibid="bib36" firstref="ref45"></nolink> <nolink nlid="nl21" bibid="bib11" firstref="ref48"></nolink> <nolink nlid="nl22" bibid="bib20" firstref="ref49"></nolink> <nolink nlid="nl23" bibid="bib37" firstref="ref61"></nolink> <nolink nlid="nl24" bibid="bib25" firstref="ref62"></nolink> <nolink nlid="nl25" bibid="bib38" firstref="ref63"></nolink> <nolink nlid="nl26" bibid="bib39" firstref="ref66"></nolink> <nolink nlid="nl27" bibid="bib19" firstref="ref68"></nolink> <nolink nlid="nl28" bibid="bib21" firstref="ref69"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1176296
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Comparing Modeling Approaches for Assessing Priorities in International Agricultural Research
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Petsakos%2C+Athanasios%22">Petsakos, Athanasios</searchLink><br /><searchLink fieldCode="AR" term="%22Hareau%2C+Guy%22">Hareau, Guy</searchLink><br /><searchLink fieldCode="AR" term="%22Kleinwechter%2C+Ulrich%22">Kleinwechter, Ulrich</searchLink><br /><searchLink fieldCode="AR" term="%22Wiebe%2C+Keith%22">Wiebe, Keith</searchLink><br /><searchLink fieldCode="AR" term="%22Sulser%2C+Timothy+B%2E%22">Sulser, Timothy B.</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Research+Evaluation%22"><i>Research Evaluation</i></searchLink>. Apr 2018 27(2):145-156.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Oxford University Press. Great Clarendon Street, Oxford, OX2 6DP, UK. Tel: +44-1865-353907; Fax: +44-1865-353485; e-mail: jnls.cust.serv@oxfordjournals.org; Web site: http://applij.oxfordjournals.org/
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 12
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2018
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Agriculture%22">Agriculture</searchLink><br /><searchLink fieldCode="DE" term="%22Research%22">Research</searchLink><br /><searchLink fieldCode="DE" term="%22Comparative+Analysis%22">Comparative Analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Models%22">Models</searchLink><br /><searchLink fieldCode="DE" term="%22Needs+Assessment%22">Needs Assessment</searchLink><br /><searchLink fieldCode="DE" term="%22Research+Needs%22">Research Needs</searchLink><br /><searchLink fieldCode="DE" term="%22Global+Approach%22">Global Approach</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1093/reseval/rvx044
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0958-2029
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: This article examines how the estimated impacts of crop technologies vary with alternate methods and assumptions, and also discusses the implications of these differences for the design of studies to inform research prioritization. Drawing on international potato research, we show how foresight scenarios, realized by a multi-period global multi-commodity equilibrium model, can affect the estimated magnitudes of welfare impacts and the ranking of different potato research options, as opposed to the static, single-commodity, and country assumptions of the economic surplus model which is commonly used in priority setting studies. Our results suggest that the ranking of technologies is driven by the data used for their specification and is not affected by the foresight scenario examined. However, net benefits vary significantly in each scenario and are greatly overestimated when impacts on non-target countries are ignored. We also argue that the validity of the single-commodity assumption underpinning the economic surplus model is case-specific and depends on the interventions examined and on the objectives and criteria included in a priority setting study.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2018
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1176296
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1176296
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1093/reseval/rvx044
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 12
        StartPage: 145
    Subjects:
      – SubjectFull: Agriculture
        Type: general
      – SubjectFull: Research
        Type: general
      – SubjectFull: Comparative Analysis
        Type: general
      – SubjectFull: Models
        Type: general
      – SubjectFull: Needs Assessment
        Type: general
      – SubjectFull: Research Needs
        Type: general
      – SubjectFull: Global Approach
        Type: general
    Titles:
      – TitleFull: Comparing Modeling Approaches for Assessing Priorities in International Agricultural Research
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Petsakos, Athanasios
      – PersonEntity:
          Name:
            NameFull: Hareau, Guy
      – PersonEntity:
          Name:
            NameFull: Kleinwechter, Ulrich
      – PersonEntity:
          Name:
            NameFull: Wiebe, Keith
      – PersonEntity:
          Name:
            NameFull: Sulser, Timothy B.
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 04
              Type: published
              Y: 2018
          Identifiers:
            – Type: issn-print
              Value: 0958-2029
          Numbering:
            – Type: volume
              Value: 27
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Research Evaluation
              Type: main
ResultId 1