Smallholder-centered agricultural strategies have had limited success in Malawi, prompting interest in anchor-enterprise models in which agribusinesses partner with surrounding smallholder farmers. This note examines whether the benefits of one such anchor enterprise — Pyxus Agriculture Malawi Limited (PAM), a Growth Poles Project partner that sources groundnuts from smallholder farmers around its Chilanga Farm in Kasungu District — reach beyond the farmers and workers it engages directly. Using a local economy-wide impact evaluation (LEWIE) model calibrated on 2025 household and business survey data together with firm administrative records, we trace the full circuit of income that PAM's local spending sets in motion.
PAM's 2025 injection of MWK 623 million — MWK 165 million in net outgrower payments for groundnut sales and MWK 458 million in wages — generated an estimated MWK 1,152 million in real household income across the zone of influence (95% CI: 1,090–1,214), corresponding to a real income multiplier of 1.85. In other words, for every kwacha PAM spent locally, a further MWK 0.85 of real income was created as that spending circulated through local markets, amounting to MWK 529 million in secondary or spillover income. Roughly 74 percent of this spillover accrued to households with no direct contractual relationship with the enterprise, reaching them through the wages and profits of local businesses that serve the PAM workforce and outgrowers. The poorest 40 percent of households captured about 34 percent of the secondary income (MWK 178 million). PAM's operations and spending also supported approximately 6,780 additional labor days beyond direct employment at Chilanga Farm. These gains came with only moderate price pressure: PAM's operations raise non-tradable prices by approximately 7.4 percent and local wages by 4.6 percent.
The multiplier is large and statistically robust, falling to 1.62 only under the most conservative assumption — treating all local activities as non-tradable — and otherwise close to the central estimate of 1.85. Its comparatively high level reflects the specific conditions of the zone of influence around Chilanga Farm and forest: high local retention of household spending, a high value-added share in groundnut production that channels outgrower payments directly into household income rather than into purchased inputs, and strong circulation of income through nearby village and trading-center markets. The principal brake on the multiplier is leakage through non-farm intermediate inputs, as local businesses source between 51 and 70 percent of their inputs from outside the zone.
Ex-ante simulations map the levers available to PAM and its partners to amplify its local economic impact further. Of the five scenarios modeled, a seed-multiplication ("nursery") package that lowers outgrower input costs and local sourcing of a workforce meal program — generate the most total income relative to the direct injection (MWK 117.6 million total from MWK 63.5 million and MWK 28.2 million from MWK 15.2 million, respectively). The wage channel yields somewhat less total income relative to the direct injection (MWK 193.3 million total from MWK 114.4 million) but distributes it more broadly, with non-PAM households capturing nearly half of the spillover income. A complementary 10 percent productivity improvement in local non-tradable businesses would raise the overall multiplier from 1.85 to 2.04 while substantially reducing the associated price pressure (from 7.4 to 4.3 percent), underscoring that investments in local market infrastructure, business capacity, and supply-chain logistics could simultaneously amplify income gains and reduce leakage.
Authors: Justin D. Kagin and J. Edward Taylor
Read and download the project note here (PDF 510 KB)
