natbib - Same author same year citations with a and b references in harvard-style (or anything other than apalike)

I have a problem where I didn't find any solution even after hours of googling. I have two references where three of the authors are the same (BibTeX file with Mendeley):

\documentclass[a4paper, 12pt]{article}
\usepackage[nottoc]{tocbibind}
\usepackage{natbib}
\usepackage{url} \let\harvardurl\url
\bibliographystyle{agsm}
\usepackage{filecontents}

\begin{filecontents*}{Mendeley.bib}
@article{Attanasio2015a,
title = {Human Capital Development and Parental Investment in India},
year = {2015},
journal = {NBER Working Paper Series},
author = {Attanasio, Orazio and Meghir, Costas and Nix, Emily},
volume = {21740}
}
@article{Attanasio2015b,
title = {{Estimating the Production Function for Human Capital: Results from
Randomized Control Trial in Colombia}},
year = {2015},
journal = {NBER Working Paper Series},
author = {Attanasio, Orazio and Cattan, Sarah and Fitzsimons, Emla and Meghir,Costas and Rubio-Codina, Marta},
volume = {20965}
}
\end{filecontents*}

\begin{document}
..... \cite{Attanasio2015a} \cite{Attanasio2015b}
\bibliography{Mendeley}
\end{document}


Now, when referencing those two sources, I want them to appear as

Attanasio et al. (2015a) and Attanasio et al (2015b).

However, when using agsm as bibliographystyle I get the full list of authors instead of

XY et al. (2015a) / XY et al. (2015b).

I already read here that apparently agsm does not support this sort of citation. But he same problem occurs with other Harvard styles like dcu and kluwer and even chicago. The only thing so far that works is using apalike.

I'm fairly new to LaTex but I feel like there must be a solution for this problem? I've never seen this way of citation in a journal article and I don't believe the authors use apalike all the time.

• In the post you linked, other comments state that this works only if all authors are the same, which is not the case here. Could you clarify I feel like there must be a solution for this problem? I've never seen this way of citation in a journal article ? – BambOo Aug 20 '18 at 20:25
• Off-topic: It's utterly wrong to use the @article entry type for the two entries at hand. You -- or, presumably, Mendeley -- should be using the entry type @techreport. – Mico Aug 21 '18 at 0:04
• Please clarify what you mean by "I don't believe the authors use apalike all the time". Which authors? Are you free to use the apalike bibliography style? – Mico Aug 21 '18 at 0:14

It's not clear from your write-up if you have to satisfy any particular formatting guidelines for the bibliographic entries. All you've stated, really, is that the citation call-outs must be of the author-year type.

I already read here that apparently agsm does not support this sort of citation.

Not just "apparently" -- it simply does not. It's not a bug, it's a "feature".

But [t]he same problem occurs with other [h]arvard styles like dcu and kluwer ...

Indeed.

Assuming you're not free to use the apalike bibliography style -- Aside: What do you mean by "I don't believe the authors use apalike all the time"? Which authors? -- you could try the elsarticle-harv and abbrvnat bibliography styles. In the answer below, I use the elsarticle-harv style.

By the way, as I've already noted in a comment, it is utterly wrong to use the @article entry type for the two working papers at hand. The NBER working paper series is not a "journal", at least not in the usual (academic) sense of the word. You (and/or Mendeley?!) should be using the entry type @techreport, with type = {Working Paper}, institution = {National Bureau of Economic Research}, and number instead of volume for the numbers of the working papers.

Finally, be careful to encase words such as "India" and "Colombia" in curly braces, to prevent BibTeX from converting them to lowercase if "sentence style" is in use. (The opposite of "sentence style" is "title style".) If Mendeley doesn't place curly braces around "India" and "Colombia", it is your duty and responsibility to do so by hand. The excuse "I use bib entries exactly as they are provided by Mendeley" is quite lame, and it won't get you very far. And, do check diligently if the information provided by Mendeley is actually correct (and complete). If you just assume that the information is correct, you're going to get a rude awakening sooner or later -- probably sooner. I've already mentioned the issue of the incorrect entry type. Another problem I noticed in the the two entries you posted is that the title field in one of the entries is missing a word ("a", in case you're curious).

\RequirePackage{filecontents}
\begin{filecontents*}{Mendeley.bib}
@techreport{Attanasio2015a,
title  = {Human Capital Development and Parental Investment in {India}},
year   = 2015,
type   = {Working Paper},
institution = {National Bureau of Economic Research},
author = {Attanasio, Orazio and Meghir, Costas and Nix, Emily},
number = 21740,
}
@techreport{Attanasio2015b,
title  = {Estimating the Production Function for Human Capital: {Results} from a Randomized Control Trial in {Colombia}},
year   = 2015,
type   = {Working Paper},
institution = {National Bureau of Economic Research},
author = {Attanasio, Orazio and Cattan, Sarah and Fitzsimons, Emla and Meghir, Costas and Rubio-Codina, Marta},
number = 20965,
}
\end{filecontents*}

\documentclass[a4paper, 12pt]{article}
\usepackage[nottoc]{tocbibind}
\usepackage[hyphens,spaces,obeyspaces]{url}

\usepackage[round]{natbib}
%\bibliographystyle{apalike} % is there something wrong with this bib style?
\bibliographystyle{elsarticle-harv}
% or: \bibliographystyle{abbrvnat}

\begin{document}
\cite{Attanasio2015b} and \cite{Attanasio2015a} \dots

\cite{Attanasio2015b,Attanasio2015a} \dots
\bibliography{Mendeley}
\end{document}