Since the emergence of sustainability reporting, several studies examined the frameworks used by companies to disclose environmental and climate-related information. Prior research has analysed reporting practices using frameworks such as the GRI, the International Integrated Reporting Council framework, the KPMG reporting guidance and the Sustainability Accounting Standards Board (SASB) standards (Erin & Bamigboye 2021; Erin, Bamigboye & Oyewo 2021; Iazzi et al. 2021; Ionașcu et al. 2020; Lodhia, Kaur & Kuruppu 2023). Empirical evidence suggests that the choice of reporting framework influences both the extent and the quality of sustainability disclosures. Erin and Bamigboye (2021), for example, utilised the GRI framework to assess sustainability reporting practices across 408 global companies and found that GRI provides robust metrics for evaluating performance against the Sustainable Development Goals, making it a comprehensive reporting framework. However, the voluntary nature of sustainability reporting continues to result in variation in reporting practices across companies. Ionașcu et al. (2020) reported that only 56% of Western European real estate companies adopted the GRI framework, highlighting the discretion afforded to companies in selecting and applying sustainability reporting frameworks. Other studies have examined the adoption of alternative voluntary standards. Lashitew (2021) argued that voluntary adoption of SASB standards can enhance the credibility and transparency of sustainability disclosures. Similarly, Bochkay, Choi and Hales (2022) found that companies applying SASB standards often demonstrate stronger sustainability performance, including lower GHG emissions, and that adoption is influenced by company size, societal norms and the presence of institutional sustainability governance structures. Despite these potential benefits, the voluntary nature of these frameworks continues to allow selective disclosure and variation in reporting quality across companies.

A growing body of literature has focused specifically on GHG and carbon-related disclosures. Pitrakkos and Maroun (2020) employed content analysis to assess the quantity and quality of carbon disclosures among JSE-listed companies using a structured disclosure index developed from prior literature. Their findings indicated that companies operating in carbon-intensive industries tend to provide more extensive and higher-quality carbon disclosures, often supported by external assurance mechanisms. However, the study relied on broad carbon reporting constructs and did not explicitly assess company-level compliance with individual GRI 305 disclosure requirements. International evidence similarly highlights persistent weaknesses in GHG reporting. Bais, Nassimbeni and Orzes (2024), in a systematic review of GRI-related research, identified ongoing concerns regarding disclosure quality, including inconsistencies, selective reporting and the risk of greenwashing. These challenges are often attributed to the voluntary nature of sustainability reporting and misalignment between corporate objectives and broader public interest goals. Kamra et al. (2025) conducted a detailed assessment of GHG emissions disclosures by Indian companies claiming alignment with the GRI 305 Emissions standard. Using content analysis of sustainability and integrated reports of companies listed on the NIFTY 100 Index, the study found substantial gaps between reported disclosures and the specific requirements of GRI 305. Despite public claims of adherence, most companies failed to fully comply with the standard, particularly in relation to methodological transparency and completeness. These findings reinforce concerns about the reliability and comparability of voluntary GHG disclosures. A recent study by Shen et al. (2026) on listed Chinese companies and their suppliers’ carbon disclosure practices highlights persistent challenges associated with Scope 3 emissions disclosure. The authors found that Scope 3 reporting remains methodologically inconsistent and insufficiently standardised, despite increasing stakeholder and regulatory pressure for greater transparency regarding value-chain emissions. These findings further highlight the challenges associated with achieving comparable and reliable voluntary climate-related disclosures.

Legitimacy theory suggests that companies use corporate disclosure practices to gain, maintain or restore societal legitimacy (Deegan 2002; Dowling & Pfeffer 1975; Lindblom 1994). Within the environmental reporting context, companies may therefore disclose sustainability and climate-related information strategically to demonstrate alignment with societal expectations and stakeholder concerns. Cho and Patten (2007), for example, found that companies use environmental disclosures as a legitimising mechanism, particularly when environmental performance is subject to greater public scrutiny. Recent evidence by Grahn (2025) similarly found that stakeholder pressure and legitimacy considerations influence voluntary GHG disclosure practices, even among private companies. In the context of voluntary GHG reporting, legitimacy theory provides a useful lens for understanding why companies may selectively disclose emissions-related information while omitting methodological detail or more challenging categories such as Scope 3 emissions. Stakeholder theory emphasises that companies are accountable to a broad range of stakeholders, rather than shareholders alone (Freeman 1984). Within the sustainability reporting context, companies may therefore disclose GHG-related information in response to increasing pressure from stakeholders concerned with environmental performance and climate-related risks (Chithambo et al. 2022; Liesen et al. 2015). Chithambo et al. (2022), for example, found that shareholders, investors, and communities exert significant influence on corporate GHG disclosure decisions. Stakeholder theory, therefore, provides a useful perspective for understanding why companies may vary in the extent and completeness of voluntary GHG disclosures. In the South African context, GHG reporting by JSE-listed food producers remains largely voluntary, allowing companies considerable discretion regarding the extent and detail of emissions-related disclosures. Legitimacy and stakeholder theories therefore provide useful perspectives for interpreting differences in disclosure completeness and transparency among companies.

Sector-specific studies further illustrate the challenges associated with GHG reporting, particularly in the food and beverage industry. Hansen et al. (2022) examined emissions reporting and target setting among the top 50 global food and beverage companies using publicly available CDP data. Their results revealed that Scope 3 emissions accounted for approximately 88% of total reported emissions, yet disclosures in this category were frequently incomplete and inconsistent. Moreover, more than one-third of reported Scope 3 emissions were excluded from emissions reduction targets. The authors concluded that the sector urgently needs to improve both the quality and the scope of GHG reporting, particularly in relation to value-chain emissions. Similarly, Reavis et al. (2022) analysed GHG reporting and climate targets of the top 100 global food and beverage companies and found that while many companies disclose Scope 1 and Scope 2 emissions, fewer than half adequately assess or report Scope 3 emissions. Given the dominance of value-chain emissions in the sector, this gap significantly undermines the usefulness of disclosed information for climate-related decision making. Evidence from South Africa remains limited. Bimha and Nhamo (2017), using CDP data, examined the relationship between carbon disclosure and share price movements among JSE-listed companies and found no significant difference between companies that regularly disclosed emissions and those that did not. While informative, this study neither assessed the content or quality of GHG disclosures, nor did it evaluate alignment with specific reporting standards such as GRI 305. Evidence from the food and beverage sector consistently indicates uneven GHG disclosure practices, particularly in relation to Scope 3 emissions, despite their dominance in sectoral emissions profiles (Hansen et al. 2022; Reavis et al. 2022). While these studies provide valuable global insights, they rely primarily on CDP data or broad disclosure measures and do not assess company-level adherence to specific GRI 305 requirements. Moreover, empirical evidence from South African food producers remains limited. While much of the existing evidence examines food and beverage companies jointly, this study focused on food producers as a distinct subsector within the JSE classification, and evaluated company-level alignment with specific GRI 305 disclosure requirements.

Ethical clearance to conduct this study was obtained from the North-West University Economic and Management Sciences Research Ethics Committee on 2 September 2025. The ethical clearance number is NWU-00870-25-A4. This research examined and assessed publicly accessible annual integrated reports and sustainability documents. The findings were anonymised, with companies labelled as FP1, FP2, and so forth. Given that the annual reports are readily available on company websites, obtaining consent was unnecessary. As there were no interactions with individuals, the Protection of Personal Information Act 4 of 2013 did not apply.

In relation to S1a, apart from FP2, all companies in the sample disclosed their gross Scope 1 emissions (see Table 2). Accordingly, all but one company explicitly acknowledged GHG emissions through the disclosure of gross Scope 1 emissions. For example, FP1 noted ‘Material environmental aspects – Scope 1 direct emissions tCO₂e 178 019’ in their ESG and integrated reporting, while FP3 stated in its ESG report: ‘Scope 1 CO₂e emissions (direct – fossil fuels/non-renewable) (tonnes) 121 503’.

The remaining disclosure requirements were largely unmet with no companies reporting gases (S1b) or biogenic CO₂ (S1c), and only two reporting baselines (S1d), which limits comparability. FP3 stated: ‘In 2024, we recorded a 23% decrease in absolute GHG emissions from our 2019 baseline’. FP5 stated: ‘Our recent data verification exercise confirmed our GHG emissions for 2023 – our new baseline year for goal setting and performance monitoring’. Only FP5 disclosed emission factors and GWP sources (S1e) and S1f: ‘We quantify our GHG emissions of carbon footprint in accordance with the GHG Protocol Corporate Accounting and Reporting Standard and we consolidate our emissions under the “operational control” approach’. FP4 did, however, state: ‘No emissions from the other African operations are included in Scope 2 emissions, and comparative figures are presented’, which suggests that the disclosed figures may exclude emissions from other African operations. The reporting on S1g was uneven, with only FP1, FP4 and FP5 disclosing methodologies, assumptions, standards and calculation tools. The findings suggest that JSE-listed food producers disclose higher-level quantitative emissions information while the omission of supporting methodological information limits stakeholders’ ability to fully assess the reliability and comparability of the reported emissions data. Similar concerns regarding incomplete and inconsistent carbon reporting have been identified by Hansen et al. (2022) and Kamra et al. (2025). From a legitimacy perspective, the disclosure of aggregate Scope 1 emissions may help food producers demonstrate environmental accountability, while the omission of methodological detail limits stakeholders’ ability to fully assess the reliability and comparability of the reported emissions data. This supports the argument by Pitrakkos and Maroun (2020) that carbon disclosures may, in some instances, function as a legitimacy management mechanism rather than providing fully decision-useful information.

All sampled companies, except FP2, disclosed location-based Scope 2 emissions (S2a) (see Table 3). FP6 only partially fulfilled its obligations by reporting total Scope 2 emissions without specifying the location base: ‘Total carbon emissions – Scope 2 (tCO2e) 101 571’. FP3 disclosed: ‘Scope 2 CO2e emissions (indirect – electricity purchased) (tonnes) 23 795’.

None of the companies reported market-based Scope 2 emissions (S2b), limiting comparability, and akin to Category 1, none of them revealed the gases used in their calculations, resulting in S2c also being unmet. Only two companies disclosed baseline years: FP3 and FP5. FP5 addressed the rationale behind their Scope 2 emissions baseline year in their sustainability report, stating: ‘Baseline year 2023 used for Scopes 1 & 2 comparisons’. FP5 fully complied with S2e and S2f by including the following statements in their sustainability report: ‘The emission factor for purchased electricity aligns with the one reported by Eskom’ and ‘We quantify our GHG emissions of carbon footprint in accordance with the GHG Protocol Corporate Accounting and Reporting Standard and we consolidate our emissions under the “operational control” approach’. For S2g, FP1 fully complied with the requirement, stating: ‘Absolute gross greenhouse gas emissions expressed as metric tonnes of CO₂ equivalent and measured in accordance with the Greenhouse Gas Protocol for Scope 1, Scope 2 and Scope 3’. FP4 and FP5 also met this requirement. FP4 specified: ‘Scope 2 emissions were calculated using the electricity grid emission factor for South Africa of 1.01 kg CO2e/kWh, Government Gazette No 50071 (2024) (2023: 1.06 kg CO2e/kWh, Eskom (2021))’. Similar to Scope 1 disclosures, the sampled companies tend to prioritise disclosure of total Scope 2 emissions while not adequately providing supporting methodological information. This pattern is consistent with prior literature highlighting incomplete and selective GHG disclosures within voluntary sustainability reporting frameworks (Hansen et al. 2022; Kamra et al. 2025; Pitrakkos & Maroun 2020). From a stakeholder perspective, incomplete methodological disclosure may reduce the decision usefulness of emissions information for investors and other stakeholders seeking transparent and comparable climate-related data. The findings therefore reinforce concerns that voluntary reporting frameworks may encourage disclosure of high-level quantitative indicators without sufficient information to support meaningful evaluation and comparison.

Category 3 was the weakest disclosure category (see Table 4). Only FP1 and FP3 disclosed total Scope 3 emissions (S3a). FP1 stated that no material Scope 3 emissions were identified, stating: ‘Scope 3 tCO₂e N/A;’ ‘Our company is not responsible for any Scope 3 emissions, and our significant GHG emissions pertain to Scope 1 and 2 carbon emissions’. This disclosure was treated as a reported Scope 3 position in the coding, although no category-level breakdown was provided. Meanwhile, FP3 successfully measured and reported ‘Scope 3 CO₂e emission (indirect, not Scope 1 or 2) (tonnes) 56 749’. None of the companies in the sample adhered to S3b, S3c, S3e or S3f.

Only FP3 reported on S3d: ‘Our Scope 3 footprint covers from packaging materials, potable water use, waste management, staff travel and office paper’. However, this statement only partially complied with the requirement, as it referenced various categories but did not provide an official breakdown of Scope 3 emissions by category. FP1 fully complied with S3g, detailing the methodology and standards used for their disclosure: ‘Absolute gross greenhouse gas emissions expressed as metric tonnes of CO₂ equivalent and measured in accordance with the Greenhouse Gas Protocol for Scope 1, Scope 2 and Scope 3’. FP3 only partially addressed methodological disclosure by reporting: ‘We currently measure and report partial Scope 3 emissions, focusing on indirect emissions under our operational control’. This disclosure lacks specifics on the methodologies or standards applied. Overall, Category 3 achieved a score of 8 out of a possible 84 (7 × 12), confirming that Scope 3 emissions reporting remains highly limited among JSE-listed food producers. This pattern is consistent with prior findings in the food and beverage sector, where Scope 3 disclosures are frequently incomplete and inconsistent (Hansen et al. 2022). Given that Scope 3 emissions are often more difficult to measure and remain voluntary within the South African reporting context, the limited disclosure may reflect both methodological challenges and selective reporting practices associated with voluntary sustainability reporting frameworks. From a legitimacy perspective, firms may prioritise disclosure of more visible and readily measurable emissions categories while omitting complex value-chain emissions that are more difficult to quantify and verify.

Most companies reported on this category (see Table 5). Where intensity was disclosed, companies generally reported ratios and operational metrics (I1–I2), but gas-level detail remained absent (I4). FP5 stated: ‘GHG emissions intensity for Scope 1 and 2 emissions tCO₂e/PHW 0.029 (2023: 0.024)’.

Four of six (67%) of the sample complied with I2. FP1 reported: ‘Total direct and indirect emissions per man hour is at 0.021 tCO₂e/man hour worked (2023: 0.019 tCO₂e/man hour worked)’. Both FP1 and FP6 complied with requirement I3. FP6 reported both Scope 1 and 2 and their intensity ratios, while FP5 did not specify the individual gas types included in its intensity ratio. As was observed in Categories 1, 2 and 3, the absence of gas-level detail limited the comparability of reported intensity metrics (I4). The comparatively stronger reporting for Category 4 may indicate that the sampled companies perceive ratio-based disclosures as more useful for communicating environmental performance. However, the absence of gas-level detail, together with the broader lack of methodological transparency observed across Categories 1–3, limits the interpretability and comparability of these metrics across companies, reinforcing persistent concerns in the sustainability reporting literature about incomplete climate-related disclosure in voluntary reporting frameworks.

Most companies reported or mentioned reductions in GHG emissions because of initiatives (R1) (see Table 6). FP3 reported: ‘We installed a 0.5 MW solar and battery system’ and then continued by saying ‘year-on-year a 41% reduction in GHG emissions emanating from vessel freon usage’. However, the sampled companies omitted disclosure on R2.

FP3 and FP5 reported on R3, with the latter disclosing: ‘Our recent data verification exercise confirmed our GHG emissions for 2023 – our new baseline year for goal setting and performance monitoring purposes’. R4 demonstrated comparatively stronger adherence across the sample. FP1 partially reported ‘Coal carbon emissions reduction of 79 tCO₂e’ and ‘Electricity carbon emissions reduction of 465 tCO₂e’, but did not explicitly state the scope(s) to which these reductions relate, although they appear to correspond to Scope 1 and Scope 2 emissions. R5 was again largely unmet. Only FP5 fully adhered by reporting: ‘We quantify our GHG emissions of carbon footprint in accordance with the GHG Protocol Corporate Accounting and Reporting Standard and we consolidate our emissions under the “operational control” approach’ and ‘Our Scope 2 emissions result from the use of purchased electricity by the same operations. Emission factors are sourced from the Intergovernmental Panel on Climate Change (IPCC) 2006 Guidelines and the South African Department of Forestry, Fisheries and Environment’s (DFFE) Methodological Guidelines for Quantification of GHG Emissions.’

The reporting on GHG reduction initiatives suggests that sampled food producers are more willing to disclose visible reduction activities and outcomes than the methodological assumptions underpinning those reductions. Similar concerns regarding incomplete methodological disclosure and selective reporting have been identified in prior GHG reporting literature (Kamra et al. 2025; Pitrakkos & Maroun 2020). From a legitimacy perspective, the disclosure of reduction initiatives may help food producers signal environmental responsiveness and commitment to sustainability objectives. However, the omission of supporting methodological detail, gases included, and clearly defined reporting scopes reduces stakeholders’ ability to evaluate the credibility and comparability of the reported reductions.

When comparing the sampled companies’ scores (see Table 7), substantial variation is evident in the extent to which the six companies adhered to the GRI 305 reporting requirements. Overall compliance levels ranged from very low to moderate, with no company achieving full alignment across all five disclosure categories.

Table 7 shows that adherence was generally higher for Scope 1, Scope 2 and emissions intensity disclosures, while Scope 3 reporting was consistently weak across the sample. One company achieved the highest overall score (57%), while another failed to comply with any of the GRI 305 requirements included in the coding index. The remaining companies displayed mixed disclosure profiles, with stronger performance in selected categories and limited or no disclosure in others. These results highlight the absence of a consistent or standardised approach to GHG reporting among JSE-listed food producers, reinforcing the uneven application of voluntary GRI 305 guidance across companies. The findings therefore suggest that, within a voluntary reporting environment, sampled food producers may prioritise visible quantitative disclosures while providing less consistent methodological and Scope 3 information, limiting the overall comparability and decision usefulness of reported GHG data.