The ARC breeding programme Economic benefits for SA’s pear industry

Consequently, investment in agricultural research in SA has been declining. Against this backdrop, the MSc study at the University of Limpopo by Constance Rasweswe aimed to estimate the economic benefits emanating from the ARC’s pear-breeding programme from 1990 to 2016. Her estimation of the benefit-to-cost ratio (BCR), marginal internal rate of return (MIRR) and the adoption rate of the outputs of the breeding programme were used to measure the economic impact of the programme. This study found that the overall adoption rate of the ARC bred pear cultivars was 5.3%, but for blush cultivars the adoption rate was 25.1%. And the BCR was 1.63:1, while the MIRR was 10.49%, with a 19-year lag period. This study concluded that the pear-breeding programme is economically viable and that the initial investment yielded notable return. For every R100 invested, an increase of R10.49 in the value was realised for the pear industry. These benefits cease after 19 years, implying that the pear-breeding programme is economically viable.

Introduction

Pear production is the second largest fruit commodity in SA (12 319 ha or 23%), and exports to various countries amount to 212 149 tonnes. Of the South African pear exports, 32% go to Europe, 20% to the Far East and Asia, 18% to the Middle East, 16% to Russia, 6% to the UK, 3% to Africa, 3% to the US and Canada, and 2% to the Indian Ocean Islands. The pear industry is valued at R3 789 million and offers employment to 13 181 permanent workers with 52 725 dependents (HORTGRO, 2018). Although SA is small in terms of global pear production, our export volumes are ranked fourth in the world (HORTGRO, 2018).

The ARC Infruitec-Nietvoorbij Institute in Stellenbosch plays a key role in maintaining the competitiveness of the South African pear industry by creating novel cultivars for the industry. To date, the ARC breeding programme has released three blush cultivars through conventional breeding techniques: “Rosemarie”, “Flamingo” and “Cheeky”. Despite the success of the deciduous fruit industry, investment in agricultural research in SA has been declining (ARC, 2016). Government funding to the ARC started declining in the 1990s and by 2008, it had plummeted to 55% of the total funding allocated to the ARC in 2001 (Flaherty et al., 2010). Inadequate evidence of the benefits of research investment and the long lag period associated with these returns – especially for long-term breeding projects, curbed efforts to change the downward trend of funding. Against this backdrop, our study attempted to quantify the economic value of the ARC’s pear-breeding programme and to assess and stimulate the potential for further investment.

Previously, no study specifically considered the economic impact of funding invested in the ARC pear-breeding programme in SA. Studies by Badiu et al., (2015) and Manning (2009) focused on introducing improved technologies for breeding programmes, but very little investigation occurred on the economic gains and the adoption rates of the cultivars released. Our study outlined the value that the agricultural industry gained from every rand invested in the ARC pear-breeding programme, in other words, the rate of return (ROR) on past investments and the lag period of research and development (R&D) investment. The study hypothesised that the ARC’s pear-breeding programme did not have any significant economic benefits, as it has only released three cultivars to a highly populated industry.

Impact assessment of various other breeding programmes

The BCR and MIRR have been calculated using different approaches. Ruttan (1982) gave an extensive review of agricultural projects in the US carried out prior to 1979, and reported that returns to agricultural research ranged from 30 to 60%. Bottomley and Thirtle (1987) and Ruttan (1982) found the returns to these projects were in excess of 20%. However, recent studies reported lower returns than reported by Ruttan (1982). For instance, Khatri et al., (1996) showed that the ROR to public sector agricultural R&D in SA is over 40%. Regardless of the methodologies used in these studies, all ROR estimates were reliable and practical (Thirtle et al., 1998) and ROR of various studies conducted in the ARC are listed in Table 1.

Data collection techniques/process

Although the Western Cape is the main pear-producing area in SA, this study was conducted in all nine provinces. In ascertaining the economic benefit of the ARC’s pear-breeding programme, a number of variables were identified and data was collected through various methods. The data on R&D costs were obtained from the pear-breeding programme’s annual progress reports submitted to industry. However, as the research funding structures and recording methods have changed over the years and policy-driven financial records are destroyed after 10 years, obtaining a full set of data was hampered. Wherever financial record data could not be obtained, the missing information was forecast using the @Risk statistical package (Palisade Corporation, Newfield, NY, US) to determine a full series of R&D data from the period 1990 to 2016.

Data on the quantity of pears produced in SA were obtained from agricultural statistics published annually by the Department of Agriculture, Forestry and Fisheries (DAFF, 1990 – 2016). Total area used for pear production data was obtained from HORTGRO Key Deciduous Fruit Statistics records while data on area planted were obtained from HORTGRO, from 1993 – 2016. Data prior to 1993 were obtained from Deciduous Fruit Board Tree Survey reports amongst archives at HORTGRO, Paarl. And records on pear prices were obtained from agricultural statistics, which were acquired from the DAFF. Income received from royalties for the ARC bred pear cultivars was obtained from CULDEVCO (Pty) Ltd.

Results and discussion

As one of the pear-breeding project objectives is to produce cultivars that are likely to be adopted in the market, the study also determined the rate of adoption of ARC cultivars. The study used the Cobb-Douglass production function approach to determine the ROR on the ARC’s pear-breeding programme, following a methodology used by Thirtle and Bottomley (1989) and Townsend and Van Zyl (1998).
The lag structure (which indicates the response on initial investment and duration) compiled by this study shows R&D having a small impact from year 0. The data imply that there was no lead time, and indicates that initial investment immediately started generating benefits (Fig. 1). The reason may be that the “Rosemarie” cultivar was released in 1990 (the same year this study period started) and therefore, started to generate revenue. The 19-year lag period is quite reasonable due to the long-term nature of pome fruit breeding. It takes 25 – 27 years from the initial crosses made to the time a cultivar is commercially exploited. The investment benefit peaked at year nine to 10, after which it dwindled until the nineteenth year.

The MIRR of the ARC’s pear-breeding programme was found to be 10.49%. This ROR is relatively low but shows that pear-breeding has in fact generated profits for the pear industry over the years. For every R100 increase in R&D investment, the output increased by R10.49. Notably, this ROR is greater than that of livestock research, (estimated to be 5%) but lower than the ROR in public sector research (44%) that focuses on field crops and horticulture sub-sectors (Khatri et al., 1996).

Our results further show that overall only 5.36% of the ARC cultivars are adopted in the industry. However, for blush cultivars, the adoption rate of ARC blush cultivars is much higher at 25.1% (HORTGRO, 2018). The low overall rate is probably due to the ARC only having released three blushed cultivars to the industry. The area planted to “Flamingo” and “Rosemarie” cultivars was relatively stable over the years and the new blush cultivar, “Cheeky®” has been doing well in the market and it is expected to boost the ARC cultivar adoption rate. The area under production for “Cheeky®” has been increasing since it was released to the industry in 2009. The BCR was estimated to be 1.63:1, despite the low adoption rate and indicates that the research benefits outweigh its costs considerably; also that the ARC could expect R1.63 in benefits for each R1 of its input cost. Therefore, the BCR confirms the viability of the ARC’s pear-breeding programme and justifies further investment in this programme. Furthermore, promising new selections in an advanced phase of evaluation could lead to the release of more new cultivars, which will further enhance the BCR.

Summary, conclusion and recommendations

The aim of this study was to investigate if the ARC pear-breeding programme is benefitting the national pear industry economically. This was done by evaluating farmers’ rate of adoption of the ARC’s pear varieties, analysing the costs and benefits of the programme (benefit-to-cost ratio), determining the ROR on the pear-breeding programme and using secondary data for a period of 26 years to assess lag structure. We showed that the adoption rate of ARC cultivars was 5.36% overall, but for blush cultivars the adaption rate was much higher at 25.1%. The benefit-to-cost ratio is 1.63:1 and the marginal internal ROR of the pear-breeding programme is 10.49%. This implies that for every R100 invested, the industry gets R10.49 in returns. The R&D lag distribution showed that the benefits from R&D investment are spread over 19 years with no lead time. The benefits peak on year nine to 10, then start lessening until they cease after 19 years. Though low, this ROR is positive and significant. It indicates that pear-breeding has in fact, generated profits for the pear industry over the years. And the BCR and ROR values obtained in this study confirm the economic ROR of the ARC’s pear-breeding programme. However, further assessment of these values is inhibited by the absence of comparable values for other public investments, as well as data limitations. The results of studies determining the lag length between R&D investment and yield effects, and measuring RORs against agricultural R&D both in SA and elsewhere in the world, compare well with the results of this study. The lag structure for pear breeding (19 years) is shorter than that of peach and nectarine breeding (21 years), but much longer than grape breeding (seven years). The returns to R&D are lower than those of grape, peach and nectarine breeding (Townsend and Van Zyl, 1998; Tsvakirai, 2018).

In, conclusion, based on these findings, the continuation of this programme is vital for the sustainability of the pear industry. Therefore, it makes economic sense for the industry to increase funding of this programme. This is in opposition to the initial hypothesis that the ARC’s pear-breeding programme did not have any significant economic benefits, as it has only released three cultivars to an already flooded cultivar industry. It is also anticipated that values obtained in this study may change over time as new cultivars from the pear-breeding programme are released. Therefore, the data should be revised every five years. In addition, this would allow the ARC to undertake long-term basic research in support of efforts to breed high-value crops, which would continually fetch higher export prices and unlock a concomitant positive spin-off to the economic growth of the country. The data limitations experienced in this study highlight the need for establishing mechanisms for capturing, saving, retrieving and sharing useful information. This will enable the evaluation of the effects of technological interventions in SA and analysis of long-term agricultural productivity.

Acknowledgments

The authors would like to express their gratitude for the supply of valuable data and time towards this study to: (i) Mariette Kotze from HORTGRO (ii) Dr Leon von Mollendorff from CULDEVCO.

Taaibos Human may be contacted at: humant@arc.agric.za.

Featured Image: Rosemarie cultivar