Evaluating the impact of improved technology adoption in traditional poultry farming on potential outcomes of farmers: Evidence from rural Togo

Soviadan, Mawussi Kossivi et al. Article — Published Version Evaluating the impact of improved technology adoption in traditional poultry farming on potential outcomes of farmers: Evidence from rural Togo Cogent Food & Agriculture Provided in Cooperation with: Leibniz Institute of Agricultural Development in Transition Economies (IAMO), Halle (Saale) Suggested Citation: Soviadan, Mawussi Kossivi et al. (2024) : Evaluating the impact of improved technology adoption in traditional poultry farming on potential outcomes of farmers: Evidence from rural Togo, Cogent Food & Agriculture, ISSN 2331-1932, Taylor & Francis, London, Vol. 10, Iss. 1, https://doi.org/10.1080/23311932.2024.2341091 , https://www.tandfonline.com/doi/full/10.1080/23311932.2024.2341091 This Version is available at: https://hdl.handle.net/10419/295210 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Food Science & Technology | ReSeaRch aRTicle Cogent Food & AgriCulture 2024, Vol. 10, no. 1, 2341091 Evaluating the impact of improved technology adoption in traditional poultry farming on potential outcomes of farmers: evidence from rural Togo Mawussi Kossivi Soviadana , osama ahmeda , Zaneta Kubikb , anselm anibueze enetec, chukwuemeka Uzoma okoyec and Thomas glaubena aleibniz institute of Agricultural development in transition economies (iAMo), department of Agricultural Markets, Marketing, and World Agricultural trade, Halle (Saale), germany; bChair of international economic Policy, university of göttingen, göttingen, germany; cdepartment of Agricultural economics, Faculty of Agriculture, university of nigeria, nsukka, nigeria ABSTRACT Social programs are designed to reach beneficiaries and achieve expected objectives. There is a need to understand whether development programs work and their level of impact on the beneficiaries involved. along these lines, the objective of this study is to evaluate the impact of improved Technology adoption in Traditional Poultry Farming (iTTPF) on farmers’ potential outcomes in Togo. Baseline and follow-up data were collected from 400 farmers and analyzed using difference-in-differences models. The study reveals that five years after the implementation of the program, the annual gross profit increased on average by US$ 1294 for each program participant. The results of the heterogeneous impacts assessment indicate that participating in the program is a necessary condition for iTTPF adoption, but not sufficient for profit optimization. overall, the program has a positive and significant impact on the potential outcomes of farmers in Togo. The government in its agricultural policy should mobilize more resources to enable considerably more farmers to adopt improved agricultural technologies. in addition, agricultural policymakers should implement the instruments of the chain planning, programming, budgeting, execution, monitoring and evaluation of all agricultural development programs and projects to make progressive adjustments for optimal results achievement and sustainable agricultural development. 1. Introduction The agricultural sector is at the heart of the economy of developing countries. it generates a large share of the gross domestic product (gdP) and employs a significant proportion of the active population. it is an important source of foreign exchange, produces the most basic foodstuffs and is the only source of livelihood and income for more than half of the population of developing countries (oecd/ Fao, 2023; Soviadan et al., 2022, 2023). The poultry sector is an agricultural sub-sector that continues to grow in many parts of the world. The increasing purchasing power of the population © 2024 the Author(s). Published by informa uK limited, trading as taylor & Francis group CONTACT Mawussi Kossivi Soviadan [email protected], [email protected] leibniz institute of Agricultural development in transition economies (iAMo), Agricultural Markets, Marketing, and World Agricultural trade, theodor-lieser-Straße 2, 06120 Halle (Saale), germany. https://doi.org/10.1080/23311932.2024.2341091 this is an open Access article distributed under the terms of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. the terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. ARTICLE HISTORY Received 11 december 2023 Revised 10 March 2024 accepted 4 april 2024 KEYWORDS Traditional poultry farming; improved technology adoption; impact evaluation; difference-in-differences models; potential outcomes; farmers REVIEWING EDITOR M. luisa escudero-gilete, University of Seville, Seville, Spain SUBJECTS agriculture & environmental Sciences; econometrics; african Studies; Rural development; Research Methods in development Studies; Microeconomics; development Studies; development Policy; economics and development; Sustainable development 2 M. K. SoViadan eTal. and urbanization have been powerful drivers of this growth. Therefore, poultry production has become highly specialized and increasingly productive and hence needs to be managed by specialists (Food and agriculture organization of the United nations [Fao], 2015). in addition, the development and transfer of technologies and farming techniques increased the efficiency of poultry production. This has led modern poultry farms to grow rapidly in size, concentrate near sources of inputs or output markets and opt for vertical integration (ao etal., 2021; chen etal., 2024; Mahanty et al., 2023). This structural reform is particularly reflected in the evolution of contract farming in the breeding of layers and broilers, which allows farmers with medium-sized units to access advanced technology with a relatively low initial investment (Fao, 2015). Traditional poultry farming is defined as small-scale, household-level poultry production using family labor and, to the greatest extent possible, locally available food supplies (Mcclaughlin et al., 2024; Wong et al., 2017). it is the most common type of breeding practiced by smallholder farmers in Sub-Saharan africa (SSa) due to low entry barriers. it plays a crucial role in rural areas of developing countries in sustaining livelihoods, supplying poultry products in rural, suburban and urban areas and representing important support (such as health care for family members, school fees for children, cultural events, rituals, etc.) for the most vulnerable groups (Fao, 2014). as long as poverty persists in developing countries, traditional poultry farming will continue to provide opportunities for high-quality income generation and nutrition for the human population (Fao, 2015). The modern largeand medium-sized poultry production system supplies integrated marketing chains while the free-range, low-income, family-scale poultry production system is supplying local or niche markets. Traditional poultry rearing is the most commonly used method of poultry production in Togo because it is less expensive than modern commercial poultry farming or other types of livestock production. it is well-established among smallholders in rural Togo due to low entry barriers and is considered an economic activity that could be easily accessed even by the most vulnerable social strata of the population (eg low-income, landless and female farmers) (Fao, 2014). however, this sector is characterized by low productivity because its production potential is inherently low combined with poor environmental and feeding conditions. losses are usually due to disappearances, theft and slaughter because of the extensive nature of this type of breeding. high mortality and slow growth of poultry are the major constraints to production. The most common mortality causes are diseases, predation, external parasites and accidents. The majority of constraints in traditional poultry farming are related to farm management techniques. in this context, and intending to make traditional poultry farming more productive and efficient to enhance food security, diversify income sources and thus strengthen the resilience of farmers for wealth creation and poverty alleviation, the introduction of iTTPF becomes crucial. The improved traditional poultry farming differs from free-range traditional poultry rearing in several respects. The improved poultry farming is semi-intensive as the poultry birds are raised in an enclosure with a well-built habitat called a semi-modern or improved traditional poultry housing. in addition, the bird species and feed quality are improved, the breeding equipment is semi-modern and the hygiene and health care are periodic (Fao, 2014, 2015; Soviadan et al., 2022, 2023; yadav et al., 2013). Since 2014, through the national Program of agricultural investment for Food and nutritional Security (PniaSan), and the agricultural Sector Support Project (PaSa), the government in Togo with the help from the Fao and the financial support from the World Bank has been assisting smallholder farmers in rural areas with the adoption of improved Technology in Traditional Poultry Farming (iTTPF) to create wealth, improve food security and alleviate poverty (Soviadan et al., 2022, 2023). The iTTPF is a commercial type of traditional poultry farming that differs from free-range traditional poultry rearing in terms of improvements in farm management, farm equipment, poultry housing, poultry feed and disease control. despite the measures taken by the government to make participation in the program more accessible to farmers, only 86 were able to be enrolled in it in 2014. Since not all farmers participated in the program, there is a need to categorize them into two groups, that is, the treated group (program beneficiaries) and the untreated group (program non-beneficiaries). Several previous studies referred to any intervention program as ‘treatment’ (alem & Ruhinduka, 2020; carter etal., 2019; cole & Fernando, 2021; gao etal., 2020; Rubin, 1974). The term ‘treatment’ refers to early work in the medical field that focused on determining the efficacy of treatments. although it is not cogenT Food & agRicUlTURe 3 the most appropriate term, it is used in econometrics to describe public intervention, subsidy policy, social assistance program or implementation of a development program that is being evaluated. Rubin’s (1974) counterfactual framework, which has been adopted by many researchers in statistics and econometrics, including Rosenbaum and Rubin (1983), heckman (1992, 1997), imbens and angrist (1994), angrist et al. (1996), heckman et al. (1997), allows for the definition of various treatment effects that may be of interest. once the treatment effect is defined, one can study ways to consistently estimate this effect (Wooldridge, 2010). The literature on treatment effects begins with a counterfactual, where each individual in the population has an outcome with and without treatment. To ensure an unbiased treatment effect evaluation, it is essential to identify the type of treatment based on the stochastic and behavioral attributes of the population. Therefore, to apply the most appropriate impact evaluation method to an intervention program, there is a need to compare the socioeconomic characteristics of beneficiaries and non-beneficiaries of this program before its implementation. The balanced socioeconomic characteristics for both treated and untreated groups should, therefore, be used as control variables for the impact evaluation analysis of the program. Furthermore, a development program is only relevant if the potential outcomes of beneficiaries differ significantly from those of non-beneficiaries after the program’s implementation. Before conducting the impact evaluation analysis, it is critical to confirm the existence of a significant difference in the potential outcomes of the two groups (treated and untreated) in the population. note that, since the introduction and implementation of iTTPF, there were no impact assessment studies to verify if the iTTPF is delivering on its intended promise of boosting livelihoods and reducing poverty. To the best of our knowledge, this is the first attempt to fill knowledge gaps in evaluating the impact of iTTPF adoption on farmers’ potential outcomes in Togo. The rest of this article is organized as follows. Section 2 highlights the implementation of iTTPF in Togo. Section 3 covers materials and methods. Research findings are presented and discussed in Section 4. The conclusion and policy recommendations are outlined in Section 5. 2. Poultry value chain in Togo Since the Maputo commitments1 in 2003 (Benin & yu, 2012), the comprehensive africa agriculture development Program (caadP) has been at the heart of many african governments’ efforts to accelerate growth and reduce poverty and hunger in african countries through the african Union (aU) and the new Partnership for africa’s development (nePad). The economic community of West african States’ Regional agricultural Policy (ecoWaS/ ecoWaP) was developed as a result of caadP implementation in 2005 (Kolavalli etal., 2012). Since 2010, Togo has been implementing the PniaSan with assistance from the Fao and the World Bank, as part of its agricultural policy for sustainable development (Soviadan et al., 2022, 2023). The objective of PniaSan was to improve food and nutritional security, increase farmers’ income and contribute to improving trade balance and rural people’s living conditions through sustainable development, with special attention paid to the poorest and most vulnerable groups (RoPPa, 2013; World-Bank, 2017). PniaSan comprised five projects, including the PaSa. The main goal of PaSa was to increase the productivity and/or competitiveness of strategic food crops, export crops and livestock farming and to promote an environment conducive to sustainable agricultural development. in this regard, a sub-component of PaSa was aimed at reviving the livestock sub-sector, the specific objective of which was to provide short-term emergency assistance to rehabilitate poultry and small ruminant production, assist small livestock farmers to develop and improve livestock farming in rural areas for wealth creation, food security and poverty reduction (Soviadan etal., 2022, 2023). The government, through this sub-component of PaSa, has made available to all farmers in rural areas a technical package to facilitate the adoption of iTTPF. This technical package includes the construction of semi-modern poultry houses (improved poultry farms), the provision of technical poultry rearing equipment, training on the composition of balanced and quality feed at lower cost, prophylaxis, vaccination of poultry, cleaning and hygiene of poultry farms, health care, etc. The cost (per farmer) of the technical package is about US$ 6364. Through PaSa, the government, with help from Fao and financial support from the World Bank, subsidized the cost of the acquisition of the technical package at the level of 90%, while farmers were required to contribute the remaining 10%, which is about US$ 636 per farmer. This counterpart or individual contribution from farmers interested in PaSa could be paid in cash or in kind. Most farmers opt for in-kind contributions, through 4 M. K. SoViadan eTal. land used as a site for the implementation of semi-intensive or improved poultry farms. Farmers who were aware of the benefits of PaSa in terms of food security, wealth creation and poverty alleviation in rural areas, but who lacked both financial capacity and land to cover their 10% counterpart, have taken out loans from financial institutions to participate in PaSa for the adoption of iTTPF (Soviadan et al., 2022, 2023). 3. Materials and methods 3.1. Agricultural technological change Since strict geographical limits exist on horizontal expansion (known as extensive margin growth), agricultural production growth will need to come mainly from vertical expansion (known as intensive margin growth) (Besedeš & Prusa, 2011; Black, 1929; chaney, 2008; Mamo et al., 2019). agricultural growth concerns have focused on technological and institutional innovations that can increase total factor productivity (TFP) (Bachewe etal., 2018; dias avila & evenson, 2010; el ghak et al., 2020; Fao, 2017; Fuglie et al., 2020; gong, 2020). For agricultural technology to be used and have an impact on agricultural production growth, it must meet three requirements: generation, adoption and diffusion (de Janvry, 1973; Foster & Rosenzweig, 1995, 2010; gallardo & Sauer, 2018; gollin et al., 2005; Magruder, 2018). The technology must be generated, socially profitable and locally available. it should be adopted by individual farmers for whom it is appropriate. it has to diffuse among farmers, generating partial and general equilibrium effects on prices and on the welfare of both producers and consumers (Feder et al., 1985; Feder & Umali, 1993; lee, 2005). The agricultural technology may consist of an overall change in the agricultural production function, as seen in the Solow model, where a gain in TFP will result in an increase in output for a certain combination of inputs, or an equivalent reduction of input costs to achieve a given output level (Bachewe et al., 2018; evenson & Fuglie, 2010). More particularly, though, the technology is intended to achieve specific gains that will be of interest to the adopter with various advantages or costs to others as a consequence of adoption (Balakrishnan et al., 1996; Kuan & chau, 2001; Robertson & gatignon, 1986). once the technology is available, it has to be adopted by farmers to have an impact on agricultural production and productivity. 3.2. Econometric methods of impact evaluation The real hurdle of an impact assessment study is determining what might have occurred to beneficiaries without the implementation of the program (Rubin, 1974). in the absence of any program or project, a beneficiary’s potential outcome would be its counterfactual. The counterfactual outcome framework can be applied in situations where each farmer in the population comes up with two possible potential outcomes, such as adopting or not adopting iTTPF. a program intervention, such as the implementation of iTTPF, aims to strengthen food security, increases income levels and improves the well-being of the targeted beneficiaries. note that it is not possible to draw any conclusions about the impact of the program on beneficiaries’ potential outcomes based on a single observation after the treatment. The problem with assessment is that, while the impact evaluation of the program can only be truly assessed by comparing actual and counterfactual potential outcomes, the counterfactual cannot be observed (heckman & Vytlacil, 2001; Khandker et al., 2009). if there is no information on the counterfactual, then the most appropriate option is to compare the potential results of treated participants or individuals to those of comparison or an untreated group that has not received treatment (Rubin, 1974). By doing this, one tries to select a subpopulation or comparison group that is highly comparable to the treated group, in such a way that those being treated would have had potential results similar to those in the untreated group if treatment had not been administered. Finding a good comparison group is critical to a successful impact evaluation. equation ( )1 illustrates the fundamental evaluation problem of comparing potential outcomes Y across treated and untreated individuals i (heckman & Vytlacil, 2001, 2005; Khandker et al., 2009) YXT ii i = ++ α βε i (1) T is a dichotomous variable which is 1 for participants in the program or 0 otherwise; X represents a set of observable socioeconomic characteristics specific to individual i ; and ε is a stochastic error term representing non-observed factors affecting Y . equation ( )1 captures a standard approach in impact assessment estimating the program’s direct effect T () on potential outcomes Y () . The issue with estimating equation ( )1 is that the assignment into the treatment might not always be random due to the intentional placement of the program as well as self-selection in the program (Khandker et al., 2009). in other words, programs are cogenT Food & agRicUlTURe 5 located based on the needs of populations and individuals, who then self-select given program set-up as well as placement. The auto-selection process could be associated with observed characteristics as well as factors that have not been observed, or both (Rosenbaum & Rubin, 1983). When unobserved variables are integrated, the stochastic error term ( ε ) in estimated equation ( )1 will include variables that are also interrelated with the dichotomous treatment variable ( T ). in equation ( )1 , such unobservable factors cannot be measured and thus cannot be accounted for, resulting in non-observed selection bias. in other words, if the covariance between the treatment () T and the stochastic error term () ε is different from zero, that is, Cov T (,) ε ≠0 , then this violates an important ordinary least squares (olS) assumption, necessary to obtain unbiased estimates. as a result, the correlation over time between () T and () ε biases other estimates in equation ( )1 , particularly the estimate of () β , which is the causal effect of the treatment on potential outcomes ( Y ). The main challenge of a robust impact evaluation study is therefore to develop techniques and methods for eliminating or taking into account the selection bias (Khandker et al., 2009; Soviadan et al., 2022). Several different methods can be used in impact evaluation theory to address the fundamental question of the missing counterfactual. each of these methods carries its assumptions about the nature of potential selection bias in program targeting and participation, and the assumptions are crucial to developing the appropriate model to determine program impacts (heckman & Vytlacil, 2001, 2005; Khandker etal., 2009). Because in our study, the data before and after the implementation of the iTTPF are available for both treated and untreated groups of farmers, the double differences or difference-in-differences (dd) method under the parallel-trend assumption is the appropriate impact assessment technique to evaluate the impact of iTTPF adoption on the potential outcomes of farmers in Togo. dd methods assume that unobserved selection is present and that it is time-invariant. The treatment effect is assessed by taking the difference in potential outcomes across treated and untreated groups of farmers before and after the introduction and implementation of iTTPF. 3.3. Empirical specifications 3.3.1. DD methods of estimating Y t T Tt it i i it =++ + + ββ β β µ 01 2 3 (2) where Y it is the potential outcome; β 3 is the estimator of the dd, the coefficient of the interaction term between time t and the treatment variable Ti ; β 2 is the fixed difference between the mean outcome values for the treated and untreated groups in the absence of any treatment; β 1 is the measurement of changes in results’ mean values in the absence of any treatment; β 0 is the constant; and µ it is the stochastic error term. The variables t and Ti are included separately to take into account the potential effect of the passage of time and an effect from being included in the treated group (a priori, not zero if the treatment is not random). The estimates of the coefficients in this regression lead to the following results: Y t T Tt it i i it =++ + + ββ β β µ 01 2 3 DD EY Y T EY Y T i ii i ii = −= () − ][ −= ()      10 10 10 (3) DD = +++ −+ () − ][ +−       [( ) ( ) ( )] ββββ ββ ββ β 012 3 02 01 0 (4) DD = +− ][     ββ β 31 1 (5) DD = β 3 The dd equation can also be rewritten as follows: Y time Treatment time Treatment it it =++ + () + ββ β βµ 01 2 3* (6) time =0 before the treatment and time =1 after the treatment Treatment =0 for the untreated group and Treatment =1 for the treated group The dd estimate is obtained following these iterations: E Y time Treatment it= = () =+++ 11 012 3 , ββββ (7) E Y time Treatment it= = () = + 01 02 , ββ (8) E Y time Treatment it= = () = + 10 01 , ββ (9) E Y time Treatment it= = () = 00 0 , β (10) now define: 6 M. K. SoViadan eTal. EY time Treatment EY time treatmentD it it         11 01 11 , ,   3 (11) EYtime Treatment EYtime treatmen tD it it         10 00 21 , ,  (12) Therefore DD D D=−= 12 3 β (13) 3.3.2. Parallel trends assumption β 3 is the dd estimate and thus the real impact of improved technology adoption in traditional poultry farming on the potential outcomes of farmers in Togo. β 3 concerns production, productivity, turnover and profit.2 The dd method is based on the paralleltrends assumption which states that endogeneity and unobserved heterogeneity in program participation may be present but that such factors are time-invariant. This impact evaluation study is therefore based on the following requirements: (i) the selection bias is invariant over time that is changes in potential outcome variables due to the intervention are not a function of the initial conditions that influenced program participation, (ii) there were no other programs introduced concurrently, and no time-persistent shocks and (iii) the potential outcomes would not have been different over time in the treated group compared to the untreated group if the program had not been introduced (please refer to the graphical results of the parallel trends assumptions tests in the appendix). 3.4. Study area and data collection This study adopted a farm household survey design. The survey was conducted between July and october 2020 in the five regions of Togo (see Figure 1). documentation and field visits allowed us to identify the different districts and localities of the five major rural areas involved in this investigation. The target population size represented the total number of farmers in Togo. The sample size for this study was determined using Fellegi’s (2003) sampling technique with a 95% confidence level. From a population of 3,738,430 farmers, 400 farmers were then selected as the core sample for this study. Baseline data collected from the Ministry of agriculture helped in identifying 86 farmers who benefited in 2014 from a subsidy for the adoption of iTTPF. This grant was awarded to them through PniaSan and PaSa implemented by the government in Togo (the Ministry of agriculture) with help from Fao and financial support from the World Bank in which they voluntarily participated. The total sample of 400 respondents was broken down by region according to the weight of each region in the national agricultural population. The 86 farmers exposed to iTTPF were distributed in the five regions of the country and by district. They were therefore considered as the beneficiaries and were part of the overall sample. Three hundred and fourteen (314) non-beneficiary farmers, randomly selected from the population using a baseline dataset, constituted the rest of the sample and were also stratified according to the weight and distribution of farmers subsidized by the district. Key socioeconomic variables, institutional characteristics, livestock ownership, production costs, income and expenditure were all collected. on September 20, 2017, this study received approval and financial support from the german academic exchange Service (daad) for its implementation. The daad is a joint organization of the universities and other institutions of higher education in the Federal Republic of germany. Supported from public funds, the daad promotes international academic cooperation, especially through the exchange of students and academics. This research project has been funded by the daad under the reference numbers 57377171, 57423580 and 57520399. at the beginning of the study, the research proposal was submitted to the daad selection committee in germany comprising a panel of independent academics for consideration and approval. The committee was transparent in its functioning and was independent of the researchers, the sponsors and any other stakeholders. Before fieldwork and data collection from respondents in Togo, authorization was obtained from the Ministry of agriculture, livestock and Rural development (MaedR) through its advisory and Technical Support institute (icaT), under reference number 0325/icaT/ dRh/dciFS. The article does not include any animal studies conducted by any of the authors. We did not collect any confidential or private information about the farmers. all individual participants in the survey provided verbal consent to be interviewed after being informed about the purpose of the study. Upon the study’s completion, the researchers submitted a final report to the daad, encapsulating a summary of the study’s findings, conclusion and recommendations. cogenT Food & agRicUlTURe 7 after performing the statistical and econometric analyses, the t-test results are presented in Table 2, the average impact results are shown in Table 3, and Tables 4, 5, 6 and 7 contain the heterogeneous impact results. 4. Results and discussions 4.1. Descriptive statistics Table 1 compares the different technology components adopted by treated and untreated farmers in Figure 1. Map of the study area (togo) Source: Author’s own design 8 M. K. SoViadan eTal. Table 4. estimates of heterogeneous impact controlling for the amount of subsidy received by program beneficiaries Potential outcomes Heterogeneous impact evaluation difference-in-differences Standard grant Sub-standard grant loss rate of poultry −0.61*** −0.63*** (0.0192) (0.0234) Hatching rate of eggs 0.30*** 0.26*** (0.0182) (0.0221) Annual sale of poultry 0.0192*** 0.0282*** (0.0012) (0.0023) turnover 0.1610*** 0.3206*** (0.0102) (0.0416) Profit 0.0922*** 0.1862*** (0.0062) (0.0250) Note: Standard errors are in parentheses. Annual sale of poultry, turnover and profit are estimated at the 10,000th scale. Source: Authors’ computation based on field data. ***Statistical significance at the 1% level. their respective poultry farming systems after program implementation. Table 2 shows the p-values of Student’s t-tests on the potential outcome variables before and five years after the implementation of the program. except for the p-value of poultry loss rate before program implementation, all the remaining are less than 1%. These findings suggest that farmers were not assigned to the program randomly. Results also indicate that after five years of the implementation of the program, there is a significant difference between the means of potential outcomes variables such as poultry loss rate, hatching rate of eggs, farm size, annual sale of poultry, turnover and profit for both the treated and untreated groups of farmers. The impact evaluation of the adoption of iTTPF on Table 3. estimates of the average impact of the program Potential outcomes difference-in-differences (Atet) loss rate of poultry −0.62*** (0.0156) Hatching rate of eggs 0.28*** (0.0144) Annual sale of poultry 0.0228*** (0.0017) turnover 0.2241*** (0.0277) Profit 0.1294*** (0.0166) Notes: Standard errors are in parentheses. Atet: average treatment effect on the treated. Annual sale of poultry, turnover and profit are estimated at the 10,000th scale. Source: Authors’ computation based on field data. ***Statistical significance at the 1% level. Table 2. Student test (t-test) on potential outcomes of both the treated and untreated groups of farmers before and five years after the implementation of the program Potential outcomes Before the program After the program Student test (t-test) Student test (t-test) loss rate of poultry 0.55 56.38*** Hatching rate of eggs 3.30*** −22.93*** Farm size −12.57*** −12.12*** Annual sale of poultry −12.33*** −14.47*** turnover −11.23*** −8.54*** Profit −11.24*** −8.37*** Source: Authors’ computation based on field data. ***Statistical significance at the 1% level. Table 1. Comparison of different technology components among the treated and untreated farmers Program poultry categories technology components untreated farmers treated farmers traditional poultry farming improved traditional poultry farming Poultry farming systems Smallholder and backyard systems Semi-intensive systems Chickens guinea fowl ducks turkeys Poultry farm Free-range rearing with or without traditional poultry dormitories Semi-modern poultry housing featuring five compartments equipped for both indoor and outdoor management practices Poultry breeds local or indigenous poultry breeds local and improved local poultry breeds Breeding protocol traditional breeding methods based on local knowledge and cultural practices Semi-modern breeding methods, integrating modern, commercially oriented approaches with traditional knowledge Poultry feed ration locally available feed resources Balanced poultry feed ration available at lower cost Breeding equipment traditional poultry farming equipment Semi-modern equipment such as feeders, waterers, nesting boxes, incubators, hatcheries, brooders, scales, etc. Health management and disease control traditional health management Modern health management: hygiene practices, prophylaxes and vaccination protocols incubation natural incubation natural and artificial incubation using incubators, hatcheries, and brooders Advisory and technical support services no or infrequent advisory and technical support services Comprehensive advisory and technical support services, including training, capacity building and expert monitoring and evaluation Financial resources Very low equity limited working capital Source: Authors’ conceptualization based on field data. cogenT Food & agRicUlTURe 15 Food and agriculture organization of the United nations. (2014). Food and agriculture organization of the United nations in collaboration with the international Fund for agricultural development and Kyeema Foundation. n°16. Decision tools for family poultry development. FAO animal production and health guidelines (Vol. 16). Rome, italy. Food and agriculture organization of the United nations. (2015). Secteur Avicole Togo. Revues nationales de l’élevage de la division de la production et de la santé animales de la FAO (N°9). Poultry sector in Togo. national livestock reviews by the Fao animal Production and health division. n°9. Rome, italy. Food and agriculture organization of the United nations. (2017). The future of food and agriculture – Trends and challenges. https://www.fao.org/3/i6583e/i6583e.pdf Foster, a. d., & Rosenzweig, M. R. (1995). learning by doing and learning from others: human capital and technical change in agriculture. Journal of Political Economy, 103(6), 1176–1209. https://doi.org/10.1086/601447 Foster, a. d., & Rosenzweig, M. R. (2010). Microeconomics of technology adoption. Annual Review of Economics, 2(1), 395–424. https://doi.org/10.1146/annurev.economics.102308.124433 Fuglie, K., gautam, M., goyal, a., & Maloney, W. F. (2020). Harvesting prosperity: Technology and productivity growth in agriculture. World Bank. https://doi.org/10.1596/978-1-4648-1393-1 gallardo, R. K., & Sauer, J. (2018). adoption of labor-saving technologies in agriculture. Annual Review of Resource Economics, 10(1), 185–206. https://doi.org/10.1146/ annurev-resource-100517-023018 gao, y., Zhao, d., yu, l., & yang, h. (2020). influence of a new agricultural technology extension mode on farmers’ technology adoption behavior in china. Journal of Rural Studies, 76(March), 173–183. https://doi.org/10.1016/j. jrurstud.2020.04.016 gollin, d., Morris, M., & Byerlee, d. (2005). Technology adoption in intensive post-green revolution systems. American Journal of Agricultural Economics, 87(5), 1310–1316. https:// doi.org/10.1111/j.1467-8276.2005.00824.x gong, B. (2020). new growth accounting. American Journal of Agricultural Economics, 102(2), 641–661. https://doi. org/10.1002/ajae.12009 heckman, J. J. (1992). haavelmo and the birth of modern econometrics: a review of the history of econometric ideas by Mary Morgan. Journal of Economic Literature, 30(2), 876–886. heckman, J. J. (1997). instrumental variables: a study of implicit behavioral assumptions used in making program evaluations. The Journal of Human Resources, 32(3), 441–462. https://doi.org/10.2307/146178 heckman, J. J., ichimura, h., & Todd, P. (1997). Matching as an econometric evaluation estimator. Review of Economic Studies, 65(2), 261–294. https://doi.org/10.1111/1467937X.00044 heckman, J. J., & Vytlacil, e. (2001). Causal parameters, structural equations, treatment effects and randomized evaluations of social programs [Manuscript]. University of chicago and american Bar Foundation. heckman, J. J., & Vytlacil, e. (2005). Structural equations, treatment effects, and econometric policy evaluation. Econometrica, 73(3), 669–738. https://doi.org/10.1111/j.14680262.2005.00594.x imbens, g. W., & angrist, J. d. (1994). identification and estimation of local average treatment effects. Econometrica, 62(2), 467–475. https://doi.org/10.2307/2951620 issahaku, g., & abdulai, a. (2020). adoption of climate– smart practices and its impact on farm performance and risk exposure among smallholder farmers in ghana. Australian Journal of Agricultural and Resource Economics, 64(2), 396–420. https://doi.org/10.1111/1467-8489.12357 Khandker, S. R., Koolwal, g. B., & Samad, h. a. (2009). Handbook on impact evaluation. Quantitative methods and practices. The Word Bank. https://doi. org/10.1596/978-0-8213-8028-4 Khonje, M., Manda, J., alene, a. d., & Kassie, M. (2015). analysis of adoption and impacts of improved maize varieties in eastern Zambia. World Development, 66, 695– 706. https://doi.org/10.1016/j.worlddev.2014.09.008 Kolavalli, S., Birner, R., & Flaherty, K. (2012). The comprehensive Africa agriculture program as a collective institution (iFPRi discussion Paper 1238). international Food Policy Research institute (iFPRi). Kuan, K. K. y., & chau, P. y. K. (2001). a perception-based model for edi adoption in small businesses using a technology–organization–environment framework. Information & Management, 38(8), 507–521. https://doi. org/10.1016/S0378-7206(01)00073-8 Kumaresan, a., Bujarbaruah, K. M., Pathak, K. a., chhetri, B., ahmed, S. K., & haunshi, S. (2008). analysis of a village chicken production system and performance of improved dual purpose chickens under a subtropical hill agro-ecosystem in india. Tropical Animal Health and Production, 40(6), 395–402. https://doi.org/10.1007/ s11250-007-9097-y lee, d. (2005). agricultural sustainability and technology adoption: issues and policies for developing countries. American Journal of Agricultural Economics, 87(5), 1325– 1334. https://doi.org/10.1111/j.1467-8276.2005.00826.x Magothe, T. M., okeno, T. o., Muhuyi, W. B., & Kahi, a. K. (2012). indigenous chicken production in Kenya: i. current status. World’s Poultry Science Journal, 68(1), 119– 132. https://doi.org/10.1017/S0043933912000128 Magruder, J. R. (2018). an assessment of experimental evidence on agricultural technology adoption in developing countries. Annual Review of Resource Economics, 10(1), 299– 316. https://doi.org/10.1146/annurev-resource-100517023202 Mahanty, S., doron, a., & hamilton, R. (2023). a policy and research agenda for asia’s poultry industry. Asia & the Pacific Policy Studies, 10(1–3), 63–72. https://doi. org/10.1002/app5.377 Mamo, n., Bhattacharyya, S., & Moradi, a. (2019). intensive and extensive margins of mining and development: evidence from Sub-Saharan africa. Journal of Development Economics, 139(2), 28–49. https://doi.org/10.1016/j.jdeveco.2019.02.001 Mcclaughlin, e., elliott, S., Jewitt, S., Smallman-Raynor, M., dunham, S., Parnell, T., clark, M., & Tarlinton, R. (2024). UK flockdown: a survey of small-scale poultry keepers and their understanding of governmental guidance on highly pathogenic avian influenza (hPai). Preventive Veterinary Medicine, 224, 106117. https://doi.org/10.1016/j. prevetmed.2024.106117 16 M. K. SoViadan eTal. Mendola, M. (2007). agricultural technology adoption and poverty reduction: a propensity-score matching analysis for rural Bangladesh. Food Policy. 32(3), 372–393. https:// doi.org/10.1016/j.foodpol.2006.07.003 oecd/Fao. (2023). OECD-FAO agricultural outlook 2023–2032. oecd Publishing. https://doi.org/10.1787/08801ab7-en Robertson, T. S., & gatignon, h. (1986). competitive effects on technology diffusion. Journal of Marketing, 50(3), 1– 12. https://doi.org/10.1177/002224298605000301 RoPPa. (2013). Ten years after the Maputo declaration on agriculture and food security: An assessment of progress in West Africa: CASE OF TOGO. european centre for development Policy Management. https://www.roppaafrique.org/iMg/pdf/togo_rapport_final-kf.pdf Rosenbaum, P. R., & Rubin, d. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41–55. https://doi.org/10.1017/ cBo9780511810725.016 Rubin, d. B. (1974). estimating causal effects of treatment in randomized and nonrandomized studies. Journal of Educational Psychology, 66(5), 688–701. https://doi. org/10.1037/h0037350 Schulz, d., & Börner, J. (2023). innovation context and technology traits explain heterogeneity across studies of agricultural technology adoption: a meta-analysis. Journal of Agricultural Economics, 74(2), 570–590. https://doi. org/10.1111/1477-9552.12521 Soviadan, M. K., enete, a. a., okoye, c. U., & Kubik, Z. (2023). determinants of farmers’ participation in the agricultural sector support project for the adoption of improved technology in traditional poultry farming: evidence from rural Togo. Journal of Agriculture and Environment for International Development (JAEID), 116(2), 87–108. https://doi.org/10.36253/jaeid-12642 Soviadan, M. K., Kubik, Z., enete, a. a., & okoye, c. U. (2022). assessing the adoption rates of improved technology in traditional poultry farming: evidence from rural Togo. African Journal of Agricultural and Resource Economics, 17(3), 206– 223. https://doi.org/10.53936/afjare.2022.17(3).14 Tambo, J. a., & Mockshell, J. (2018). differential impacts of conservation agriculture technology options on household income in Sub-Saharan africa. Ecological Economics, 151(april), 95–105. https://doi.org/10.1016/j. ecolecon.2018.05.005 Wong, J. T., de Bruyn, J., Bagnol, B., grieve, h., li, M., Pym, R., & alders, R. g. (2017). Small-scale poultry and food security in resource-poor settings: a review. Global Food Security, 15(april), 43–52. https://doi.org/10.1016/j. gfs.2017.04.003 Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MiT Press. World-Bank. (2017). International Development Association (Project paper. Report no: Pad2219). https://documents1. worldbank.org/curated/en/598011492394445446/pdf/ Togo-PP-03282017.pdf yadav, g. S., debnath, c., datta, M., ngachan, S. V., yadav, J. S., & Babu, S. (2013). comparative evaluation of traditional and improved farming practices in Tripura. Indian Journal of Agricultural Sciences, 83(3), 310–314. cogenT Food & agRicUlTURe 17 Appendix: Graphical results of parallel trends assumptions tests on potential outcomes of treated and untreated groups of farmers before and after treatment