{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T06:03:44Z","timestamp":1789538624925,"version":"build-2803163510"},"update-to":[{"DOI":"10.12688\/f1000research.161343.1","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000}},{"DOI":"10.12688\/f1000research.161343.2","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000}}],"reference-count":40,"publisher":"F1000 Research Ltd","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["f1000research.com"],"crossmark-restriction":false},"short-container-title":["F1000Res"],"abstract":"<ns4:p>Background Orthodox microeconomic theory establishes a positive link between employee wages and productivity in competitive markets. Nevertheless, this perspective often overlooks gender and treats the workforce as a homogeneous category labelled labor, potentially obscuring gender-based inequalities in wage determination. This article addresses the gender wage gap by analyzing, for the first time, at least in an emerging economy such as Colombia, the impact of manufacturing firm\u2019s productivity on wages explicitly including gender. Method First, we use the GMM two-equation system proposed by Wooldridge (2009) to obtain consistent and unbiased estimates of output elasticities and TFP, respectively. Secondly, we explore the wages-productivity linkage by gender, implementing a dynamic random effect generalized least squares model (GLS) with panel data to address endogeneity issues. Results Our main findings reveal, among others, that firms with a higher proportion of female workers (female-intensive firms) generally have higher productivity than those with a higher proportion of male workers (male-intensive firms). Conclusions The effect of female-intensive firm\u2019s productivity on wages is lower than that of male-intensive firm\u2019s productivity, consistent with gender-based wage disparities in the Colombian manufacturing industry.<\/ns4:p>","DOI":"10.12688\/f1000research.161343.3","type":"journal-article","created":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T11:09:06Z","timestamp":1788260946000},"page":"306","update-policy":"https:\/\/doi.org\/10.12688\/f1000research.crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Productivity, Real Wages, and Gender. A study in Colombian Manufacturing"],"prefix":"10.12688","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6582-4129","authenticated-orcid":false,"given":"Andr\u00e9s Mauricio","family":"G\u00f3mez-S\u00e1nchez","sequence":"first","affiliation":[{"name":"Department of Economics Popayan Cauca Universidad del Cauca"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7772-7302","authenticated-orcid":false,"given":"Zoraida","family":"Ram\u00edrez-Guti\u00e9rrez","sequence":"additional","affiliation":[{"name":"Department of Accounting Popayan Cauca Universidad del Cauca"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Isabel Cristina","family":"Rivera-Lozada","sequence":"additional","affiliation":[{"name":"Department of Economics Popayan Cauca Universidad del Cauca"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"2560","published-online":{"date-parts":[[2026,9,1]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"2411-2451","DOI":"10.3982\/ECTA13408","article-title":"Identification Properties of Recent Production Function Estimators.","volume":"83","author":"D Ackerberg","year":"2015","journal-title":"Econometrica."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"255-283","DOI":"10.2307\/2937787","article-title":"The Fair Wage-Effort Hypothesis and Unemployment.","volume":"105","author":"G Akerlof","year":"1986","journal-title":"En. Q. J. Econ."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"75-100","DOI":"10.1257\/jep.14.4.75","article-title":"Gender differences in pay.","volume":"14","author":"F Blau","year":"2000","journal-title":"J. Econ. Perspect."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"115-143","DOI":"10.1016\/S0304-4076(98)00009-8","article-title":"Initial conditions and moment restrictions in dynamic panel data models.","volume":"87","author":"R Blundell","year":"1998","journal-title":"J. Econ."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"1739-1773","DOI":"10.1111\/ecoj.12214","article-title":"Wage adjustment and productivity shocks.","volume":"126","author":"M Carlsson","year":"2016","journal-title":"Econ. J."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"115-158","DOI":"10.1080\/10438599800000031","article-title":"Research, innovation and productivity: An econometric analysis at the firm level.","volume":"7","author":"B Crepon","year":"1998","journal-title":"Econ. Innov. New Technol."},{"key":"ref9","doi-asserted-by":"publisher","first-page":"1-21","DOI":"10.1257\/mic.5.3.1","article-title":"Detecting learning by exporting.","volume":"5","author":"J De Loecker","year":"2013","journal-title":"AEJ: Microeconomics."},{"key":"ref10","doi-asserted-by":"publisher","first-page":"1072-1089","DOI":"10.1002\/smj.1955","article-title":"Does female representation in top management improve firm performance? A panel data investigation.","volume":"33","author":"C Dezs\u00f6","year":"2012","journal-title":"Strateg. Manag. J."},{"key":"ref84","doi-asserted-by":"publisher","first-page":"455-473","DOI":"10.1525\/sp.2002.49.4.455","article-title":"Wages of virtue: The relative pay of care work.","volume":"49","author":"P England","year":"2002","journal-title":"Social Problems"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"135-158","DOI":"10.1016\/j.red.2012.10.009","article-title":"Trade and market selection: Evidence from manufacturing plants in colombia.","volume":"16","author":"M Eslava","year":"2013","journal-title":"Rev. Econ. Dyn."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"591-594","DOI":"10.1016\/j.jpolmod.2008.04.003","article-title":"Did wages reflect growth in productivity?","volume":"30","author":"M Feldstein","year":"2008","journal-title":"J. Policy Model."},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1080\/01603477.2020.1774393","article-title":"The illusions of calculating total factor productivity and testing growth models: from Cobb-Douglas to Solow and Romer.","volume":"43","author":"J Felipe","year":"2020","journal-title":"J. Post Keynes. Econ"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"1091-1119","DOI":"10.1257\/aer.104.4.1091","article-title":"A grand gender convergence: Its last chapter.","volume":"104","author":"C Goldin","year":"2014","journal-title":"Am. Econ. Rev."},{"key":"ref13","article-title":"Topics on firms\u2019 estrategic desicions in a developing country: the case of manufacturing in Colombia.","author":"A G\u00f3mez S\u00e1nchez","year":"2020"},{"key":"ref14","article-title":"Gender and firm\u2019s productivity in Colombian manufacturing.","author":"A Gomez-Sanchez","year":"2025"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.14867593","author":"A G\u00f3mez S\u00e1nchez","year":"2025","journal-title":"Zenodo."},{"key":"ref16","article-title":"ICT Effects on Firms' Export Decisions: Evidence for Colombian Manufacturing.","volume":"20","author":"A G\u00f3mez-S\u00e1nchez","year":"2022","journal-title":"International Review of Entrepreneurship."},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.3386\/w27054","author":"B Hansen","year":"2020","journal-title":"Information and the persistence of the gender wage gap: Early evidence from California's salary history ban (No. w27054)."},{"key":"ref83","doi-asserted-by":"publisher","first-page":"295-321","DOI":"10.1080\/13545701.2015.1090009","article-title":"Gender equality, economic growth, and women\u2019s agency: The \"endless variety\" and \"monotonous similarity\" of patriarchal constraints.","volume":"22","author":"N Kabeer","year":"2016","journal-title":"Feminist Economics"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"317-341","DOI":"10.1111\/1467-937X.00246","article-title":"Estimating production functions using inputs to control for unobservables.","volume":"70","author":"J Levinsohn","year":"2003","journal-title":"Rev. Econ. Stud."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"1-33","DOI":"10.60758\/laer.v31.109","article-title":"The dynamic linkages between exporting and importing in Colombian manufacturing.","volume":"31","author":"J Llopis","year":"2022","journal-title":"Latin American Economic Review."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"45-59","DOI":"10.1016\/j.regsciurbeco.2009.09.005","article-title":"Firms' location decisions and minimum wages.","volume":"40","author":"I M\u00e9jean","year":"2010","journal-title":"Reg. Sci. Urban Econ."},{"key":"ref24","article-title":"La productividad de las empresas y los salarios de los trabajadores.","volume":"13","author":"J Messina","year":"2019"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.18235\/0010911","article-title":"Ethnic and gender wage gaps in Ecuador.","volume":"No. 679","author":"H \u00d1opo","year":"2009","journal-title":"Inter-American Development Bank Working Paper"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"25-46","DOI":"10.5709\/ce.1897-9254.129","article-title":"Age and gender effects of workforce composition on productivity and profits: Evidence from a new type of data for German enterprises.","volume":"8","author":"C Pfeifer","year":"2014","journal-title":"Contemporary Economics."},{"key":"ref44","first-page":"545-564","article-title":"The Decision to Export in Colombia: An Empirical Model of Entry with Sunk Costs.","volume":"87","author":"M Roberts","year":"1997","journal-title":"Amer. Econ. Rev."},{"key":"ref28","doi-asserted-by":"publisher","first-page":"51-67","DOI":"10.1080\/00343400701654186","article-title":"Research and development, spillovers, innovation systems, and the genesis of regional growth in Europe.","volume":"42","author":"A Rodr\u00edguez-Pose","year":"2008","journal-title":"Reg. Stud."},{"key":"ref29","doi-asserted-by":"publisher","first-page":"588","DOI":"10.12688\/f1000research.147674.1","article-title":"The gap that Matilda will bridge: a look at the Colombian case.","volume":"13","author":"I Rivera-Lozada","year":"2024","journal-title":"F1000Res."},{"key":"ref30","doi-asserted-by":"publisher","first-page":"62-82","DOI":"10.1108\/aea-01-2023-0009","article-title":"The dynamic links between product and process innovations and productivity for Colombian manufacturing.","volume":"32","author":"J Sanchis Llopis","year":"2024","journal-title":"Applied Economic Analysis."},{"key":"ref31","doi-asserted-by":"publisher","first-page":"31-65","DOI":"10.21789\/24222704.1221","article-title":"Faldas por pantalones: las obreras y la brecha salarial en la industria colombiana en 1945.","volume":"4","author":"D Santos","year":"2017","journal-title":"Tiempo y econom\u00eda."},{"key":"ref82","doi-asserted-by":"publisher","first-page":"1225-1243","DOI":"10.1016\/j.worlddev.2011.05.001","article-title":"Gender wage inequality and economic growth: Is there really a puzzle?","volume":"28","author":"S Seguino","year":"2000","journal-title":"World Development"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"303-322","DOI":"10.1007\/s10290-010-0049-7","article-title":"Higher wages in exporting firms: self-selection, export effect, or both? First evidence from linked employer-employee data.","volume":"146","author":"T Schank","year":"2010","journal-title":"Rev. World Econ."},{"key":"ref34","first-page":"1-17","article-title":"Investment in human capital.","volume":"51","author":"T Schultz","year":"1961","journal-title":"Am. Econ. Rev."},{"key":"ref35","first-page":"433-444","article-title":"Equili-brium Unemployment as a Worker Discipline Device.","volume":"74","author":"C Shapiro","year":"1984","journal-title":"Am. Econ. Rev."},{"key":"ref87","doi-asserted-by":"publisher","first-page":"115-120","DOI":"10.2307\/1927538","article-title":"Laws of Production and Laws of Algebra: The Humbug Production Function.","volume":"56","author":"A Shaikh","year":"1974","journal-title":"Rev. Econ. Stat"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"4164","DOI":"10.3390\/su15054164","article-title":"Relationships between average wages in the manufacturing sector and economic indicators of the manufacturing sector in the region of Visegrad Group countries.","volume":"15","author":"L Suh\u00e1nyi","year":"2023","journal-title":"Sustainability."},{"key":"ref88","doi-asserted-by":"publisher","first-page":"326-365","DOI":"10.1257\/jel.49.2.326","article-title":"What Determines Productivity?","volume":"49","author":"C Syverson","year":"2011","journal-title":"Journal of Economic Literature."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"197-236","DOI":"10.1590\/S0104-83332002000100008","article-title":"Desempe\u00f1o laboral de hombres y mujeres: opinan los empresarios.","author":"R Todaro","year":"2002","journal-title":"Cadernos pagu."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"19-36","DOI":"10.1016\/j.chieco.2019.03.005","article-title":"Does gender structure affect firm productivity? Evidence from China.","volume":"55","author":"M Tsou","year":"2019","journal-title":"China Econ. Rev."},{"key":"ref40","doi-asserted-by":"publisher","first-page":"112-114","DOI":"10.1016\/j.econlet.2009.04.026","article-title":"On estimating firm-level production functions using proxy variables to control for unobservables.","volume":"104","author":"J Wooldridge","year":"2009","journal-title":"Econ. Lett."}],"container-title":["F1000Research"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/f1000research.com\/articles\/14-306\/v3\/xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/f1000research.com\/articles\/14-306\/v3\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/f1000research.com\/articles\/14-306\/v3\/iparadigms","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T05:24:11Z","timestamp":1789536251000},"score":1,"resource":{"primary":{"URL":"https:\/\/f1000research.com\/articles\/14-306\/v3"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":40,"URL":"https:\/\/doi.org\/10.12688\/f1000research.161343.3","relation":{"has-version":[{"id-type":"doi","id":"10.12688\/f1000research.161343.2","asserted-by":"subject"},{"id-type":"doi","id":"10.12688\/f1000research.161343.1","asserted-by":"subject"}],"has-review":[{"id-type":"doi","id":"10.5256\/f1000research.207118.r516938","asserted-by":"subject"},{"id-type":"doi","id":"10.5256\/f1000research.207118.r516938","asserted-by":"object"}]},"ISSN":["2046-1402"],"issn-type":[{"value":"2046-1402","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,1]]},"assertion":[{"value":"Indexed","URL":"https:\/\/f1000research.com\/articles\/14-306\/v3#article-reports","order":0,"name":"referee-status","label":"Referee status","group":{"name":"current-referee-status","label":"Current Referee Status"}},{"value":"10.5256\/f1000research.207118.r516938, Davide Villani, Joint Research Centre, European Commission, Seville, 16 Sep 2026, version 3, indexed","URL":"https:\/\/f1000research.com\/articles\/14-306\/v3#referee-response-516938","order":0,"name":"referee-response-516938","label":"Referee Report","group":{"name":"article-reports","label":"Article Reports"}},{"value":"10.5256\/f1000research.177354.r400226, Ronald M. Hernandez, Universidad Senor de Sipan (Ringgold ID: 203395), Chiclayo, Lambayeque, 29 Aug 2025, version 1, 1 approved with reservations, 1 not approved","URL":"https:\/\/f1000research.com\/articles\/14-306\/v1#referee-response-400226","order":1,"name":"referee-response-400226","label":"Referee Report","group":{"name":"article-reports","label":"Article Reports"}},{"value":"<b>Zoraida Ram&iacute;rez-Guti&eacute;rrez<\/b>; \n<i>Posted: 13 Oct 2025<\/i>; Thank you for your comments and feedback on our manuscript. Next, we provide detailed responses to each of your observations and describe how we have addressed them in the revised version of the paper.  Comment 1: The literature review needs to be reinforced with a broader theoretical approach, clearly defining the frameworks that influence productivity. Reply 1: We have substantially expanded the literature review to integrate a broader set of theoretical frameworks explaining the determinants of productivity. Beyond neoclassical and human capital approaches, we incorporated efficiency wage theory, industrial organization, and firm heterogeneity (Syverson, 2011), as well as feminist economics and labor market segmentation (Blau &amp; Kahn, 2000; Kabeer, 2016). These additions clarify how productivity is shaped by both firm-level strategies and institutional factors, providing a stronger conceptual foundation for the empirical analysis.  Comment 2: Provide more detailed evidence on how productivity influences employees\u2019 wages. It is stated that this view varies from country to country\u2014please illustrate this claim with examples. Reply 2: We enriched the review with empirical examples showing how the productivity\u2013wage nexus differs across contexts. For instance, Tsou &amp; Yang (2019) find that in China, productivity gains from female workers are stronger in small private and foreign firms than in public enterprises. Pfeifer &amp; Wagner (2014) report that in Germany, female-dominated firms appear less productive under OLS but outperform male firms under GMM, illustrating the importance of methodology and institutional settings. We also added Colombian evidence (RUES, Emicr&oacute;n, WCP) showing that women-led firms display higher productivity yet receive lower wages, reinforcing the relevance of structural and cultural factors.  Comment 3: Consolidate the theoretical framework. In addition to mentioning economic theories, I consider that a gender perspective is important to explore its relationship with productivity. Reply 3: The theoretical framework was consolidated by integrating gender economics as a core dimension. We discuss how occupational segregation, unequal access to training, and promotion barriers (Blau &amp; Kahn, 2000; Seguino, 2000; England, 2005; Kabeer, 2016) influence productivity and its transmission to wages. This gender-sensitive approach is now presented alongside traditional theories (human capital, efficiency wages, firm heterogeneity), allowing us to examine how structural biases interact with productivity within female- and male-intensive firms.  Comment 4: Justify, in the results section, why random effects models were used instead of fixed effects models. Review the tests applied and strengthen the rationale for their selection and use. Reply 4: We have clarified the methodological rationale for our choice of the dynamic random effects model and have reported the specification tests that support this selection. We opted for a dynamic random effects (RE) GLS model instead of a fixed effects (FE) model for three main reasons. First, the Hausman test did not reject the null hypothesis of no systematic differences between estimators, indicating that the random effects estimator is consistent and more efficient. Second, the random effects specification allows us to include time-invariant variables, such as geographic location and industrial sector, which are crucial to our analysis and would be omitted under the fixed effects specification. Third, the dynamic structure of our model captures wage persistence and controls for unobserved heterogeneity correlated with lagged wages, consistent with the methodology of Blundell &amp; Bond (1998). This strategy provides robust estimates while addressing simultaneity and endogeneity issues (Wooldridge, 2009).  Comment 5: The Discussion section needs to be reformulated. It is currently supported by only five citations and should focus on the changes generated by productivity and gender, the biases present in this study, and the potential practical implications that can be drawn. Reply 5: We have reformulated the Discussion section to include a broader set of references and a clearer focus on three aspects: (i) the implications of productivity differences by gender and their impact on wage transmission (Tsou &amp; Yang, 2019; Pfeifer &amp; Wagner, 2014; Rivera-Lozada et al., 2024); (ii) the main sources of potential bias, such as firm classification by workforce composition and sectoral coverage; and (iii) the practical relevance of aligning wages with productivity in female-intensive firms. We now discuss policy measures gender-disaggregated cost systems, incentives for inclusive promotion practices, and training for women in high-productivity sectors that can help address structural discrimination while enhancing firm efficiency and competitiveness.","URL":"https:\/\/f1000research.com\/articles\/14-306\/v3#referee-comment-14774","order":2,"name":"referee-comment-14774","label":"Referee Comment","group":{"name":"article-reports","label":"Article Reports"}},{"value":"10.5256\/f1000research.197171.r478891, Naeem Akram, Ministry of Economic Affairs Government of Pakistan, Islamabad, 08 May 2026, version 2, 3 approved with reservations, 1 not approved","URL":"https:\/\/f1000research.com\/articles\/14-306\/v2#referee-response-478891","order":3,"name":"referee-response-478891","label":"Referee Report","group":{"name":"article-reports","label":"Article Reports"}},{"value":"<b>Zoraida Ram&iacute;rez-Guti&eacute;rrez<\/b>; \n<i>Posted: 16 Jul 2026<\/i>; Dear Reviewer:  Thank you for your thorough and constructive review of our manuscript. Your comments have led to substantive improvements in the econometric strategy, the terminology, the interpretation of results, and the framing of our conclusions. Below, we provide a detailed response to each of your five comments and describe the specific changes made in the revised version of the paper.  \n<b>Comment 1:<\/b> \n<i>The central conclusion of the manuscript is that lower wage responsiveness in female-intensive firms constitutes evidence of gender wage discrimination. However, the empirical specification primarily identifies correlations rather than causal discriminatory mechanisms. Without worker-level information or stronger identification strategies, it is difficult to isolate discrimination from structural labor market sorting. The authors should therefore moderate the language throughout the manuscript and frame the findings more cautiously as evidence consistent with gender-based disparities rather than definitive proof of discrimination.<\/i>  \n<b>Reply 1:<\/b> We fully accept this observation and have moderated the language throughout the manuscript. The terms 'gender wage discrimination' and 'labor exploitation' have been systematically replaced with formulations such as \n<i>'gender-based wage disparities'<\/i>, \n<i>'patterns consistent with gender-based inequality'<\/i>, and \n<i>'evidence suggestive of structural gender disparities'<\/i>. The following specific changes were made: (i) Abstract: \u201cwhich could indicate gender wage discrimination\u201d, replaced by \u201cwhich is consistent with gender-based wage disparities in the productivity-wage transmission mechanism\u201d. (ii) Introduction: \u201csuggesting wage discrimination in the Colombian manufacturing industry\u201d replaced by \u201csuggesting a productivity-wage gap that may reflect structural gender disparities\u201d. (iii) Discussion and Conclusions: All references to \u201cdiscrimination\u201d have been qualified with language indicating that our results are correlational and that causal identification would require worker-level data or quasi-experimental strategies. We have also added an explicit caveat in the Discussion section acknowledging that our firm-level panel data, while enabling a robust characterisation of the productivity-wage link by gender composition, does not allow us to disentangle wage discrimination from sector-specific sorting, compensating differentials, or unobserved firm-level heterogeneity. This is presented as a limitation and a direction for future research, consistent with Blau and Kahn (2000) and Seguino (2000). Lastly, regarding your suggestion of estimating a dynamic panel GMM wage model, while System GMM (Blundell &amp; Bond, 1998) is a formally valid alternative, our specification already addresses the initial conditions problem, the primary econometric motivation for GMM, through the inclusion of pre-sample means of the dependent variable (\n<i>pre_lw2013<\/i>), following the parametric approach proposed by Wooldridge (2005). This strategy controls for the correlation between &nbsp;\n<i>lw<\/i>\n<i>t-1<\/i>&nbsp;and the individual fixed effect \n<i>u<\/i>\n<i>i<\/i>&nbsp;without the instrument proliferation concerns that arise in System GMM when T is moderate (T\u0304 \u2248 7 in our sample), and which can invalidate the Hansen overidentification test. The CRE\/Mundlak robustness checks (Table A2) further validate our approach by confirming that the main coefficients are stable across alternative treatments of unobserved heterogeneity. We therefore consider the current specification, dynamic RE GLS with pre-sample means and CRE robustness, to be both econometrically rigorous and more transparent than a GMM alternative in this context.  \n<b>Comment 2:<\/b> \n<i>The classification of firms according to the proportion of female workers is not equivalent to female ownership, female management, or female leadership. Yet the manuscript occasionally uses these concepts interchangeably. A firm with a high female labor share may simply belong to sectors traditionally characterized by lower wages or labor-intensive production processes. Consequently, the terminology 'female firms' and 'male firms' is conceptually imprecise and potentially misleading. The manuscript would benefit from replacing these labels with terms such as 'female-intensive firms', 'firms with higher female workforce participation', or 'female-majority workforce firms'.<\/i>  \n<b>Reply 2:<\/b> We agree entirely with this observation. The terms \n<b><i>female-intensive firms<\/i><\/b> and \n<b><i>male-intensive firms<\/i><\/b> have been adopted systematically throughout the revised manuscript, replacing all instances of \u201cfemale firms\u201d and \u201cmale firms\u201d. This applies to the abstract, introduction, literature review, methodology, results, discussion, and conclusions, as well as to all table and figure captions. We have also added an explicit definitional note in the Methodology section explaining the basis for our classification: \n<i>Following Tsou and Yang (2019) and Pfeifer and Wagner (2014), we classify firms based on their workforce gender composition rather than ownership or management structure. This approach is consistent with our data source (EAM), which does not report ownership by gender. Accordingly, we refer to firms where women represent 50% or more of the workforce as female-intensive firms, and to those where men predominate as male-intensive firms. This classification reflects labor force composition, which may capture broader organizational practices, sectoral norms, and institutional barriers, rather than leadership structure.<\/i> We further note that this classification is consistent with the literature's existing operational approaches. Tsou and Yang (2019) and Pfeifer and Wagner (2014), among others, classify firms by the proportion of female employees precisely because workforce composition, rather than leadership, is the relevant dimension for studying productivity and wage determination at the firm level. As Reviewer 3 correctly identifies, sectoral sorting (i.e., female-intensive firms concentrating in traditionally lower-wage industries) is a plausible alternative explanation; we address this by including industry fixed effects in all specifications and by presenting results disaggregated by sector in Table 2.  \n<b>Comment 3:<\/b> \n<i>The justification for preferring random effects over fixed effects is not convincing. The manuscript argues that fixed effects would eliminate industry and time controls, which is not correct. Time dummies can be included in fixed-effects models, and sectoral heterogeneity can be treated in alternative ways. Moreover, the decision not to report or rely on a Hausman test weakens confidence in the specification choice. Since unobserved firm heterogeneity is likely correlated with productivity and wage determination, the random-effects estimator may be inconsistent. The authors should: (i) provide a formal econometric justification for the RE assumption, (ii) conduct robustness checks using fixed effects, (iii) consider correlated random effects, or (iv) estimate a dynamic panel GMM wage model. At minimum, the paper should demonstrate that the main findings are robust across alternative specifications.<\/i>  \n<b>Reply 3:<\/b> This is the most substantive econometric comment in the review, and we have responded to it comprehensively. We conducted three complementary specification tests, the results of which are now reported in the revised Methodology section and in Tables A1\u2013A3 of the Appendix.  A. Serial Correlation Test (Wooldridge, 2002). As a first diagnostic, we applied the Wooldridge (2002) test for first-order serial correlation in panel data to the static version of our model. The results confirm strong and highly significant serial correlation in wages for all three productivity specifications:  \n<b>Specification<\/b> \n<b>F statistic<\/b> \n<b>p-value<\/b> \n<b>Interpretation<\/b>  TFPwm (preferred) 519.62 &lt; 0.001 Strong serial correlation  TFP (aggregate) 553.74 &lt; 0.001 Strong serial correlation  APL 544.17 &lt; 0.001 Strong serial correlation  This finding does not represent a problem for our specification; rather, it validates the dynamic model. The strong persistence in wages, consistent with wage rigidity documented in industrial organization literature (Carlsson et al., 2016), provides direct empirical justification for including the lagged wage (\n<i>lw<\/i>\n<sub>(t-1)<\/sub>) as a regressor. Our dynamic model absorbs this autocorrelation, ensuring consistent estimation. B. Hausman Test (FE vs RE). We report the Hausman specification test for all three subsamples of the preferred TFPwm model. Results are shown in Table A1 of the Appendix:  \n<b>Subsample<\/b> \n<b>chi2<\/b> \n<b>d.f.<\/b> \n<b>p-value<\/b>  Full sample (Industry) 9,454.58 39 &lt; 0.001  Female-intensive firms 1,745.65 36 &lt; 0.001  Male-intensive firms 5,860.63 39 &lt; 0.001  The test formally rejects RE in all three subsamples. However, we must flag a critical limitation that makes this result uninformative in the present context: when the lagged dependent variable (\n<i>lw<sub>(it-1)<\/sub><\/i>) is included as a regressor, as required by our dynamic model, the within (FE) estimator is also inconsistent. This is the well-known Nickell (1981) bias, arising because the within-group transformation creates a mechanical negative correlation between (\n<i>lw<sub>(it-1)<\/sub><\/i>)and the transformed error that does not vanish as N \u2192 \u221e but only as T \u2192 \u221e (Wooldridge, 2010, Ch. 11). The empirical signature of this bias is evident in our own results. The coefficient on (\n<i>lw<sub>(it-1)<\/sub><\/i>) under FE is approximately 0.33-0.38, while under RE it is 0.59-0.67, a discrepancy of approximately -0.29 that is entirely consistent with downward Nickell bias and inconsistent with the idea that RE is capturing spurious persistence. The Hausman test in this case is comparing two inconsistent estimators and is therefore uninformative (Wooldridge, 2010). We report it to be transparent, but do not rely on it for our specification choice.  C. Correlated Random Effects - CRE\/Mundlak (1978) Following the Reviewer\u2019s suggestion and the recommendation of the modern panel data literature, we estimate Correlated Random Effects (CRE) models following Mundlak (1978) and Chamberlain (1980). This approach is the formally valid alternative to the Hausman test in dynamic panels (Wooldridge, 2010). The CRE model relaxes the strict RE orthogonality assumption by modelling the correlation between firm-level heterogeneity (\n<i>ui<\/i>) and time-varying regressors, augmenting the RE model with firm-level time-averages (\n<i>X\u0304i<\/i>) of all lagged time-varying variables (productivity, skills, market concentration, mark-up, exports, innovation, age, and size). The joint Wald test on the Mundlak terms provides the relevant specification test. The results, reported in Table A2, show three key findings: (i) Joint Wald tests on Mundlak terms. chi2(8) = 1,320.30 (full sample, p &lt; 0.001), 395.15 (female-intensive, p &lt; 0.001), and 977.63 (male-intensive, p &lt; 0.001). The orthogonality condition is rejected, confirming that between-firm heterogeneity is correlated with the regressors, as expected in any sample of heterogeneous manufacturing firms. (ii) Key coefficient of interest (ltfp_wm\n<sub>(t-1)<\/sub>) in CRE. The within-firm productivity-wage elasticity differs markedly by gender composition. For \n<i>female-intensive firms<\/i>, the coefficient is 0.087* in the CRE model. larger than the RE estimate of 0.057* and remaining statistically significant. For \n<i>male-intensive firms<\/i>, the CRE coefficient falls to 0.018 (non-significant), indicating that the RE estimate of 0.062 for this group was largely driven by between-firm structural differences. This finding strengthens the paper's core argument: the productivity-wage link for female-intensive firms is robust even after fully controlling for firm-level heterogeneity, while for male-intensive firms the relationship is primarily structural (between-firm) rather than dynamic (within-firm). (iii) Qualitative conclusions unchanged. Wage persistence, the negative SME effect, the positive role of skills and innovation, and the Bogot&aacute; location premium all remain robust across RE and CRE specifications. Regarding the Reviewer\u2019s suggestion to estimate a \n<i>dynamic panel GMM<\/i> (Blundell-Bond) model, we note that the CRE approach already addresses the core concern, namely, that unobserved heterogeneity may be correlated with regressors, while maintaining a more tractable and interpretable framework for our nine-specification comparative analysis. The system GMM estimator would be a natural extension and is noted as a direction for future research.  \n<b>Comment 4:<\/b> \n<i>Several coefficient interpretations appear problematic or exaggerated. For example, the manuscript interprets estimated coefficients near 0.05 or 0.06 as indicating that a 1% increase in productivity raises wages by 5\u20136%, which appears inconsistent with standard log-log interpretations. Similarly, the difference between female and male labor elasticities (0.136 versus 0.132) is statistically significant but economically modest. The manuscript should distinguish more carefully between statistical significance and substantive economic relevance. A clearer discussion of the economic magnitude of the estimated effects is necessary.<\/i> \n<b>Reply 4:<\/b> The Reviewer has identified an important error in our presentation of results, and we are grateful for this correction. In a log-log specification, which is our model, where both the dependent variable (\n<i>lw<\/i>, real wages) and all continuous regressors are in natural logarithms, the estimated coefficients are elasticities, not percentage-point effects multiplied by 100. A coefficient of 0.062 on \n<i>ltfp_wm<sub>(t-1)<\/sub><\/i>\n<sub> <\/sub>means that a 1% increase in gender-weighted TFP is associated with a 0.062% increase in real wages, not a 6.2% increase. This error has been corrected throughout the manuscript. Every quantitative statement in the Results and Discussion sections now uses the correct elasticity interpretation. Regarding the economic magnitude of the elasticity differential (0.062 for male-intensive vs 0.057 for female-intensive firms, a gap of 0.005 percentage points per 1% TFP increase): while this difference appears numerically small in elasticity terms, we have added a paragraph in the Discussion section contextualizing it in economic terms. Given the average TFP level and wage levels in the sample (expressed in Colombian pesos), this differential implies a non-trivial cumulative wage gap when compounded over the eight-year panel period, particularly for the high-productivity sectors identified in Table 2. We now explicitly acknowledge the distinction between statistical significance and economic magnitude and frame the elasticity gap as one piece of evidence, alongside the descriptive analysis, the sectoral breakdown, and the wage persistence results, supporting the hypothesis of gender-based wage disparities.  \n<b>Comment 5:<\/b> \n<i>Although lagged explanatory variables are included, simultaneity between wages and productivity is still plausible. Higher wages may themselves raise productivity through efficiency wage mechanisms, while more productive firms may pay higher wages. The current framework does not convincingly address this bidirectional relationship. Additional discussion and robustness checks are needed.<\/i>  \n<b>Reply 5:<\/b> We agree that the bidirectional relationship between wages and productivity is theoretically important, it is in fact central to efficiency wage theory, which we discuss in the Literature Review, and we have strengthened our treatment of simultaneity in the revised manuscript. Our empirical strategy addresses this concern through three channels, which are now explained explicitly in the Methodology section:  (i) Lagged regressors. All explanatory variables, including productivity, are entered with a one-period lag. This eliminates contemporaneous feedback from wages to productivity: the TFP measure in period t\u22121 cannot have been affected by wage decisions in period t. While this does not rule out lagged reverse causality (wages at t\u22121 affecting TFP at t, which then affects wages at t), the dynamic structure of the model captures this channel through the lagged wage term \n<i>lw<sub>(it-1)<\/sub><\/i> rather than through the productivity regressor. (ii) Pre-sample mean of wages (pre_lw2013). Following Blundell and Bond (1998), we include the firm's pre-sample average wage as a regressor. This controls for the time-invariant component of wage-setting behavior, including persistent managerial practices that may simultaneously determine productivity levels and wage policies, reducing the scope for omitted variable bias. (iii) Serial correlation test. As reported in Reply 3 above, the Wooldridge (2002) test applied to the static model confirms strong serial correlation (F &gt; 500, p &lt; 0.001 for all specifications), which the dynamic model is specifically designed to absorb. The inclusion of \n<i>lw<sub>(it-1)<\/sub><\/i> ensures that the residuals of the estimated model are serially uncorrelated, which would not be the case if a meaningful reverse-causality channel from wages to productivity were left unaddressed. We have added a paragraph in the Methodology section explicitly acknowledging the efficiency wage mechanism, that higher wages can incentivize effort and raise productivity, and explaining how our dynamic specification with lagged regressors and the pre-sample mean addresses the resulting endogeneity. We also note in the Discussion that residual endogeneity remains a limitation of firm-level panel approaches without natural experiments, and that future work with quasi-experimental variation (e.g., minimum wage reforms, trade liberalization episodes) could strengthen causal identification.  We hope that these substantive revisions fully address your concerns.  We remain available to address any further questions.  Yours sincerely,  Andr&eacute;s Mauricio G&oacute;mez-S&aacute;nchez Zoraida Ram&iacute;rez-Guti&eacute;rrez Isabel Cristina Rivera-Lozada \n<i>Universidad del Cauca, Popay&aacute;n, Colombia<\/i>","URL":"https:\/\/f1000research.com\/articles\/14-306\/v3#referee-comment-16676","order":4,"name":"referee-comment-16676","label":"Referee Comment","group":{"name":"article-reports","label":"Article Reports"}},{"value":"10.5256\/f1000research.197171.r478893, Sara Caria, University of Modena and Reggio Emilia, Reggio Emilia, 14 May 2026, version 2, 3 approved with reservations, 1 not approved","URL":"https:\/\/f1000research.com\/articles\/14-306\/v2#referee-response-478893","order":5,"name":"referee-response-478893","label":"Referee Report","group":{"name":"article-reports","label":"Article Reports"}},{"value":"<b>Zoraida Ram&iacute;rez-Guti&eacute;rrez<\/b>; \n<i>Posted: 16 Jul 2026<\/i>; Dear Reviewer,  Thank you for your careful reading of our manuscript and for your constructive comments. Your observations are well-founded and have led to meaningful improvements in the paper. Below we provide a detailed response to each of your points, indicating the specific changes made in the revised version.  \n<b>Comment 1: <\/b>\n<i>The paper overlooks completely any reference to informality, which represents a high share of the labor market in Colombia, as in the rest of Latin America. High levels of informality affect not only informal workers but also outcomes in the formal sector through several interconnected channels. Women are far more exposed to informality than men, which should at least be mentioned as a potential part of the explanation. There is a consolidated literature showing that high informality contributes to weaker productivity\u2013wage pass-through, greater labor market segmentation, and lower wage growth even among formal workers.<\/i>  \n<b>Reply 1:<\/b> We fully agree with this observation. Informality is a structural feature of the Colombian labor market that our study did not address, and we appreciate the Reviewer bringing this gap to our attention. Colombia exhibits one of the highest informality rates in Latin America, with approximately 57% of the total labor force in informal employment during the study period (DANE, 2020; La Porta &amp; Shleifer, 2014). Critically, this burden is not gender-neutral: female informality rates consistently exceed male rates across all Colombian regions and sectors (ILO, 2018; Perry et al., 2007). This has two direct implications for our analysis. First, the EAM survey, our primary data source, covers only formal manufacturing establishments with ten or more workers. Informal workers, who are disproportionately female, are therefore excluded from the sample. This structural exclusion may lead to an underestimation of the full extent of gender-based wage disparities in Colombian manufacturing, as the most economically vulnerable female workers do not appear in our data. Second, and as the Reviewer correctly highlights, high economy-wide informality can suppress wages even among formal workers through labor market segmentation. Informal employment serves as an outside option that weakens workers' bargaining power, particularly for women, whose concentration in informal sectors reduces their leverage in formal wage negotiations (Maloney, 2004; Perry et al., 2007; Ohnsorge &amp; Yu, 2022). This channel is consistent with our finding of a weaker productivity\u2013wage pass-through in female-intensive firms, and represents a complementary, rather than competing, explanation to the structural gender-based disparities we document. In this sense, we have added a sixth limitation paragraph in the Discussion section, explicitly acknowledging the role of informality, its gender dimension, and its implications for the interpretation of our findings. Five new references were added to the bibliography: DANE (2020), ILO (2018), La Porta &amp; Shleifer (2014), Maloney (2004), Ohnsorge &amp; Yu (2022), and Perry et al. (2007). The new paragraph reads as follows: \n<i>A sixth limitation concerns the role of labor market informality, which this study does not explicitly account for. Colombia, like most Latin American economies, exhibits persistently high levels of informal employment, approximately 57% of the labor force during the study period (DANE, 2020; La Porta &amp; Shleifer, 2014). Critically, informality is not gender-neutral as women are systematically more exposed to informal employment than men, with female informality rates consistently exceeding male rates across all Colombian regions and sectors (ILO, 2018; Perry et al., 2007). <\/i> \n<i>This gender-differentiated exposure to informality has two direct implications for our findings. First, the EAM survey covers only formal manufacturing establishments with ten or more workers, meaning that informal workers (who are disproportionately female) are excluded from our sample. This may lead to an underestimation of the true extent of gender-based wage disparities in Colombian manufacturing. Second, high economy-wide informality can suppress wages even among formal workers through labor market segmentation, that is, informal employment acts as an outside option that weakens workers\u2019 bargaining power, particularly for women, thereby contributing to a weaker productivity\u2013wage pass-through in female-intensive firms independently of firm-level discrimination mechanisms (Maloney, 2004; Perry et al., 2007; Ohnsorge &amp; Yu, 2022). Incorporating informality, through linked employer\u2013employee datasets that span both formal and informal employment, represents a direction for future research on gender wage gaps in Colombian manufacturing<\/i>.  \n<b>Comment 2: <\/b>\n<i>The classification of firms based on the share of female employees is translated into concepts such as female ownership, management, or leadership, in a way that is unclear and sometimes such classifications overlap. A high share of female workers within a firm may simply reflect its presence in sectors that are typically lower-paying or more labor-intensive, rather than any gender-related characteristics of firm control or decision-making. Please revise the terminology, explain the different concepts and use them consistently.<\/i> \n<b>Reply 2:<\/b> We fully agree with this observation. This concern was also raised by Reviewer 3 (Comment 2) and was addressed comprehensively in the current revision. Specifically, the terms \n<i>female firm'<\/i> and \n<i>male firms<\/i> have been replaced throughout the entire manuscript, including the abstract, introduction, literature review, methodology, results, discussion, and conclusions, with the terms \n<b><i>female-intensive firms<\/i><\/b> and \n<b>male-intensive firms<\/b>, which refer exclusively to firms where women represent 50% or more (or less) of the total workforce, respectively. A definitional note has been added to the Methodology section to make this classification explicit and transparent: \n<i>Following Tsou and Yang (2019) and Pfeifer and Wagner (2014), we adopt the term female-intensive firms to describe establishments where women represent 50% or more of the total workforce, and male-intensive firms for those where men predominate. This classification reflects labor force composition, not ownership or management structure, a distinction that, as we argue, captures broader organizational practices and sectoral norms that may be invisible in leadership-based classifications.<\/i> Regarding the Reviewer\u2019s point about sectoral sorting , that a high share of female workers may simply reflect the firm's presence in lower-paying or more labor-intensive sectors rather than any gender-related decision-making characteristic, we address this empirically by including 23 industry fixed effects in all specifications (Table 3). These dummies absorb sector-level wage and productivity differences, ensuring that the estimated differential in the productivity-wage link between female-intensive and male-intensive firms is identified within, not across, sectors.  \n<b>Comment 3: <\/b>\n<i>Some coefficient interpretations in the manuscript appear overstated. For instance, coefficients around 0.05\u20130.06 are interpreted as implying that a 1% increase in productivity leads to a 5-6% increase in wages, which is not consistent with the standard interpretation of log-log models. In addition, while the difference between female and male labor elasticities (0.136 versus 0.132) is statistically significant, its economic magnitude may be not so relevant. The manuscript would benefit from more clearly distinguishing between statistical significance and economic relevance, and from providing a more careful discussion of the practical size of the estimated effects.<\/i> \n<b>Reply 3:<\/b> The Reviewer is entirely correct, and we are grateful for identifying this error. This issue was also raised by Reviewer 3 (Comment 4) and has been fully corrected in the current revision. In a log-log specification, where both the dependent variable (real wages) and all continuous regressors are expressed in natural logarithms, the estimated coefficients are elasticities. A coefficient of 0.062 on \n<i>ltfp_wm<sub> (t-1)<\/sub><\/i> therefore means that a 1% increase in gender-weighted TFP is associated with a 0.062% increase in real wages, not a 6.2% increase. All quantitative statements in the Results and Discussion sections have been corrected accordingly, and a clarifying note on the interpretation of elasticities has been added to the Methodology section. Regarding the economic magnitude of the elasticity gap between female-intensive (0.057%) and male-intensive (0.062%) firms, a difference of 0.005 percentage points per 1% TFP increase, we have added a paragraph in the Discussion section that contextualizes this differential in economic terms. Although numerically small in elasticity units, this gap is meaningful when compounded over the eight-year panel period and expressed in Colombian pesos at average wage and productivity levels, particularly in the high-productivity sectors identified in Table 2. We explicitly acknowledge the distinction between statistical significance and economic relevance throughout the revised text.  \n<b>Comment 4: <\/b>\n<i>Justification for the selection of random effects could be strengthened.<\/i> \n<b>Reply 4:<\/b> We agree that the justification for the random effects (RE) specification required strengthening. This concern was also raised by Reviewer 3 (Comment 3) and has been addressed comprehensively in the current revision through three complementary specification tests, which are now reported in the Methodology section and in Tables A1-A3 of the Appendix. A. Serial Correlation Test (Wooldridge, 2002). We apply the Wooldridge (2002) panel serial correlation test, which reveals strong first-order autocorrelation in wages in the static model (F &gt; 519, p &lt; 0.001 for all three productivity specifications). This finding validates our dynamic specification: the strong wage persistence empirically justifies including the lagged wage \n<i>lw\u209c\u208b\u2081<\/i> as a regressor, and our dynamic model absorbs this autocorrelation, ensuring consistent estimation. B. Hausman Test (FE vs RE). Table A1. We report the Hausman specification test for all three subsamples of the preferred TFPwm model (full sample chi2(39) = 9,454.58; female-intensive chi2(36) = 1,745.65; male-intensive chi2(39) = 5,860.63; all p &lt; 0.001). However, in dynamic panel models with a lagged dependent variable, the within (FE) estimator is also inconsistent due to the Nickell (1981) bias: the within-group transformation creates a mechanical negative correlation between \n<i>lw\u2090\u209c\u208b\u2081<\/i> and the transformed error that does not vanish as N \u2192 \u221e but only as T \u2192 \u221e. With T\u0304 \u2248 7 periods in our sample, this bias is non-trivial, as evidenced by the large discrepancy between FE (0.33\u20130.38) and RE (0.59-0.67) coefficients on the lagged wage (a gap of approximately -0.29) that is consistent with textbook Nickell bias. The Hausman test therefore compares two inconsistent estimators and is uninformative in this context (Wooldridge, 2010, Ch. 11). C. Correlated Random Effects-CRE\/Mundlak (Table A2). As the definitive specification test, we estimate Correlated Random Effects (CRE) models following Mundlak (1978) and Chamberlain (1980), which is the formally valid alternative to the Hausman test in dynamic panels. The CRE model relaxes the strict RE orthogonality assumption by including firm-level time-averages of all time-varying regressors. Joint Wald tests on the Mundlak terms are highly significant (chi2(8) = 1,320.30, 395.15, and 977.63 for the full, female-intensive, and male-intensive samples respectively), confirming correlated heterogeneity. Crucially, however, the key coefficient of interest, the within-firm productivity-wage elasticity for female-intensive firms, remains statistically significant and is even larger in the CRE model (0.087*) than in the baseline RE estimate (0.057*), while for male-intensive firms the within-firm effect is not significant (0.018, p &gt; 0.10). This asymmetric result strengthens our core argument and confirms that the main findings are robust to controlling for firm-level unobserved heterogeneity. In addition, Table A3 presents robustness checks with alternative classification thresholds (54.5% and 60%) for defining female-intensive firms. Results for male-intensive firms are virtually identical across all thresholds (0.061-0.062***), and coefficients for female-intensive firms remain stable in magnitude (0.057* to 0.047), with the slight loss of significance at higher thresholds reflecting reduced sample sizes rather than structural instability.  We hope that these revisions, particularly the new treatment of labor market informality, fully address your concerns and improve the quality of the manuscript. We remain available to address any further questions.  Yours sincerely,  Andr&eacute;s Mauricio G&oacute;mez-S&aacute;nchez Zoraida Ram&iacute;rez-Guti&eacute;rrez Isabel Cristina Rivera-Lozada \n<i>Universidad del Cauca, Popay&aacute;n, Colombia<\/i>","URL":"https:\/\/f1000research.com\/articles\/14-306\/v3#referee-comment-16677","order":6,"name":"referee-comment-16677","label":"Referee Comment","group":{"name":"article-reports","label":"Article Reports"}},{"value":"The author(s) declared that no grants were involved in supporting this work","order":7,"name":"grant-information","label":"Grant Information"},{"value":"This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.","order":0,"name":"copyright-info","label":"Copyright"}]}}