Further and better limits of expectation: Schenoni and Mainwaring’s ‘US hegemony and regime change in Latin America’

A couple of years ago I felt moved to respond to Kurt Weyland’s article ‘Limits of US influence: The promotion of regime change in Latin America’. A provocative intervention, it made the rather eyebrow-raising claim that – in that region – ‘US influence has been fairly limited for many years and has diminished over time’. ¹ Suffice to say, I found that piece ‘flabbergasting’ and ‘rather disingenuous’.² My response took as its starting point Max Paul Friedman’s 2003 assertion that ‘a work of history that strongly argues the merit of US policies in the region [Latin America] or claims these have been designed principally to protect and promote freedom and democracy’ was, by then, ‘unusual’.³ While neither Weyland’s piece nor the article under discussion here are works of history, they mark different approaches to what ultimately constitutes a whitewashing of US foreign policy in the region. Weyland argued that the United States ‘has not managed to use its overwhelming capability advantage to successfully shape regime trajectories’ because ‘a liberal-democratic superpower faces much greater constraints’ than ‘authoritarian or totalitarian great powers’.⁴

In a 2019 article, Luis L. Schenoni and Scott Mainwaring took a different – and much more positive – route to a similar conclusion.⁵ I do not intend to go through everything I found problematic or debatable in this article, but rather to provide what I see as a representative sample of the kinds of assertion, categorisation or assumption that underpin the argument and methodology. Let us begin with a key statement from the introduction: ‘Typically … hegemonic influence entails some agency on the ground through the application of economic or political tools to affect the balance of power between pro-regime groups and the opposition.’⁶

This may well be typical when speaking of other hegemons, or of the United States’ influence in other parts of the world – I would not claim to be sure – but this is far from the only way of operating. In Latin America, for several decades, right-wing potential coup-plotters were emboldened by the knowledge that at worst the United States would grudgingly accept their seizure of power, while very often they would be actively rewarded (more so during particular administrations than others, clearly). This is a demonstration effect, a form of signalling that is not encapsulated in Schenoni and Mainwaring’s definition of hegemonic influence above. Their article notes that ‘by applying or threatening sanctions and offering or promising rewards … the hegemon can influence the domestic balance’.⁷ This is obviously true, but the hegemon can also have a tremendous impact by not applying or threatening sanctions – for example, its early knowledge of plots against Salvador Allende. Frustratingly, the authors clearly understand this: ‘a hegemon can also influence regime outcomes without necessarily displaying great agency’, while ‘even the hegemon’s acquiescence can have effects’.⁸ These two caveats need far greater attention and weight, and given such, would radically alter the conclusions, I suspect.

This relates to a second point from the opening sections – the assertion that ‘scholars need to disentangle hegemonic and horizontal influences before claiming or rejecting their presence’.⁹ In many complex cases where longstanding local conflicts intersected with Cold War framings, I would argue that such disentangling is impossible to achieve fully. Even in the relatively small, well-defined case of the Guatemalan coup of 1954, there are several competing interpretative frameworks to be found in the secondary literature.¹⁰ The argument rows back from the very strong claim in their abstract – that there is ‘reason to doubt’ the attribution of ‘responsibility to the US for the waves of democratic breakdowns’ in the region – to the less bullish contention that ‘the US was not primarily responsible for the authoritarian waves’.¹¹ There is a huge difference between not being responsible at all and not being primarily responsible: one can be equally responsible without being primarily responsible. In my own work on the Cold War, I have largely come to see the United States and local elites as co-authors of authoritarianism.¹² Again, a potentially crucial get-out clause passes the reader’s eye all too rapidly: ‘hegemons could be simply intervening in an already moving process or triggering such a process involuntarily’.¹³ This is absolutely true – though neither of those caveats removes responsibility. The United States was the most powerful ‘hegemon’ in the world throughout this period, let alone in Latin America. Its military power is well understood, of course, and after the end of the gold standard, it effectively became the issuer of the world’s reserve currency. Its hegemony was discursive too though; the US shaped and fostered the global language of anticommunism in pursuit of its own interests. If we are to take its democratic pretensions seriously, as the authors do here, the United States surely had an oversized duty to exercise caution in precisely such situations: ‘We didn’t start it’ or ‘We didn’t know it would happen exactly like that’ simply won’t do.

On a related note, I’m also unclear as to why ‘when hegemonic mechanisms are present, the first in a sequence of regime changes should start soon after a modification in the hegemon’s foreign policy’.¹⁴ This feels like tipping the scales, as instances of actively biding one’s time or structural change happening at a slower pace are removed from consideration. So, while I agree that ‘timing … and sequence … in addition to geographical location, can strengthen the inference that the hegemon played a key role in the process’ they might also have little or no effect, for myriad reasons.¹⁵

Moving into the analysis, the authors return to their stronger claim – that it is too ‘simplistic’ to state that the US bore ‘considerable responsibility’ for democratic breakdowns. As I argued in my response to Weyland’s article, the weight of the historiography is against such a view. I have absolutely no doubt that the US bore considerable responsibility by any reasonable definition of the word ‘considerable’. In the opening of the analytical section, there are also several fairly wild assertions that cannot go unchallenged. First, where the US wobbled in supporting democracy it was, we are told, due to ‘its assessment of strategic risks’.¹⁶ If this is a euphemism for the swivel-eyed paranoia that saw Soviet invasion plans everywhere from Guatemala to Chile, then that is fine, but it is better to say so clearly. If not, the utter absurdity of some of this strategic risk assessment ought to be noted.

Second, ‘Washington never in principle preferred authoritarian regimes to democracy’.¹⁷ This is rather infuriating. Who is ‘Washington’? Is it just that the US government never issued a blanket statement saying, ‘we do not necessarily prefer authoritarian forms of government’? What does ‘in principle’ even mean, when very, very often ‘in practice’ they did? This all acts as soft soap for the dramatic revelation that ‘at most, the US favoured coups in 12 out of 27 democratic breakdowns’.¹⁸ Only 12? Thank goodness. Perhaps I am being overly credulous here, but if a hegemon approves of 12 mostly right-wing coups, then – absent intensive efforts to signal the contrary – it is saying extremely clearly that most right-wing coups are likely to win its approval. In Panama (1948), Venezuela (1948) and Argentina (1966), the article argues that ‘Washington remained neutral when coups did not provide for a clear Cold War security advantage’.¹⁹ So, where there was no perceived strategic gain, the US did not oppose democratic breakdown? This doesn’t sound like a principled preference for democracy.²⁰ Conversely, ‘the resumption of pro-democracy US diplomacy’ – dated to Kennedy’s presidency (but never mind the Bay of Pigs) – correlates with ‘the hesitancy of militaries to retain power’ in Argentina and Peru.²¹ Does that mean the US is influencing transitions or responding to them? It is hard to infer, though later the authors do offer the possible explanation that the US was engaging in ‘mere credit taking’.²² It’s certainly plausible. Two further gripes here: first, as we shall see below, I think 12 is probably a rather conservative count; second, there is no distinction here in terms of scale. The US favoured a lot of coups in very large countries, while sometimes paying little attention to – or even expressing mild disappointment with – coups in smaller places. Tens of millions of people falling under authoritarian rule for decades is one n, hundreds of thousands for a few years just another n.

On to those n, then, where we can find some choice inclusions. To make the extremely bold claim that the US opposed all democratic breakdowns since 1980, the authors categorise both Haiti (1995 and 1999) and Honduras (2009) as moments of pro-democratic intervention. I assume that the reference to Haiti (1999) is an error, though the US was certainly part of Operation Uphold Democracy in 1994–95.²³ However, by 2000 the US was actively undermining democracy, having turned against Jean-Bertrand Aristide. The integrity of that year’s election was significantly weakened by the relentless propaganda of the Bush administration, which took a very active role in the overthrow of the elected president – first, through sanctions and blockades and second, through the co-planning and facilitating of the 2004 coup.²⁴ Obviously, this should not be counted as opposing a democratic breakdown. In the case of Honduras, President Obama’s initial condemnation of the coup was rapidly undermined by Secretary of State Clinton who equivocated on both condemnation and potential sanctions, opting ultimately for small cuts to aid and infrastructure investment, while maintaining huge commitments to funding both the Honduran military and police. She then pushed the Organization of American States to accede to President Manuel Zelaya’s removal, despite its strong initial response. Clinton’s actions were out of step with the position of the US ambassador, who had no hesitation in describing the military overthrow as ‘an illegal and unconstitutional coup’.²⁵ Clinton’s own memoir suggests that her speculations about Zelaya’s possible future actions eased her conscience, such as it is: ‘Certainly the region did not need another dictator’, a statement, by her own admission, based entirely on hearsay.²⁶ All told, this absolutely cannot be described as opposing a democratic breakdown. It adds further evidence to a more compelling case: that the United States government is a strong supporter of democracy in Latin America when (and only when) its preferred candidates win.

Here, we might look at another deeply contentious case: Argentina in 1946. The US government certainly supported elections, but they made it very clear that only one outcome was acceptable. It was not the one that they got – infamously, Spruille Braden lost.²⁷ Bolivia (1956) is another head-scratcher, again claimed here as a demonstration of US support for a democratic restoration. In fact, 1956 marks no major qualitative shift in Bolivia’s democratic status; rather, it was an externally directed political shift away from the previously capacious Revolutionary Nationalist Movement (MNR) coalition towards a much more narrowly pro-US construct, achieved through base economic pressure. The fact that a subsequent coup against this government was quite widely celebrated may be hard to fit into a quantitative model, but that does not mean it is irrelevant. What it may suggest, though, is that the definition of democracy used here is extremely thin gruel – the barest bones of Robert Dahl’s procedure over substance. Then there is the – at best – asinine and – at worst – grossly offensive description of US policy in Nicaragua and El Salvador in the 1980s as ‘indirect … military action’ to ‘support democracy’.²⁸ Ronald Reagan could not have said it better himself.

This brings us to a rather starker admission: ‘the US was not a major actor in most democratic transitions before 1977’.²⁹ So in the period in which the US backed (at least) 12 coups, it was not a major pro-democracy actor. First, this further undermines the claim that it did not necessarily prefer authoritarian governments to democracies, because it clearly preferred many authoritarian governments and was neutral towards democratic transitions in this era. Second, it prompts one to ask: in this period would support for democratic restoration not have been fundamentally more useful or meaningful than a wave that soon took on its own momentum? All this is pointing towards a characterisation of the United States as being in favour of democracy in Latin America only if (a) it costs very little effort and (b) it results in the victory of a preferred candidate. When we come to the third wave of democratisation, the article repeats Huntington’s conclusions that the US played a critical role in eight cases, a contributing role in three and a decisive role in three. But let me ask: did El Salvador become a democracy in 1984? Did Guatemala in 1986? Did Honduras in 1982? Did Peru in 1995? We are once more floating in Dahl’s thin gruel. Even these contentious successes are limited to small- and medium-sized polities. The United States played no significant role in the three largest democratisation processes in the region: Argentina in 1983, Brazil in 1985 and Mexico in 1988, in what, by the way, is very generously designated a democratic transition but was, in fact, a slam-dunk stolen election.³⁰ Bolivia, meanwhile, was pressured into holding elections, but suffered three years of subsequent coups. This is, nonetheless, categorised as a US success story – at least ‘probably’!³¹

Another major problem here is that there is very little attempt to complicate the notion of the ‘US government’. A pro-democracy sentiment expressed in an embassy is taken as indicative of a pro-democracy sentiment tout court, yet we know very well that sections of the US state apparatus acted in tension – and even outright opposition – throughout the Cold War. That the article bolsters its argument with contributions from enormously self-interested informants – a smattering of US ex-diplomats, who must surely be keen to burnish their resolutely liberal-democratic credentials – merely adds to the impression that this is ultimately either an exercise in contrarian clickbait or (more worryingly) an attempt to overwrite history with an algorithm. A separate issue is the exclusion of other possible factors; the correlation between a US rhetorical turn to democracy with democratisation itself is thus presented as logically causal. Yet, the fact that only five breakdowns are identified after 1978 ought to take into account other factors: the growth of social movements in the region; the near hegemony of global human rights discourse after the 1970s; a self-reinforcing interaction between these two factors; and the variance in regional solidarity and pro-democracy diplomacy. Moreover, there is no real mention of an absolutely enormous shift in the tools available to the US to exercise hegemony after 1978 – the Washington Consensus was constructed on the back of brutal economic pressure and allowed the United States to achieve its economic goals without funding too many more military or clandestine operations. Is this positive support for democracy? I would not say so.

Beyond what I see as demonstrably dubious conclusions to the specific line of enquiry here, I also feel compelled to ask a broader question about the utility of these political science methods. Crisp-set qualitative comparative analysis (CsQCA), the article notes, ‘is a data analysis technique suitable for medium-n contexts in which the researcher knows enough about each case to have an informed scholarly judgment’.³² I fear a different technique may be required. If ‘basic force’ and ‘force activation’ models³³ are telling us that the United States was essentially a peripheral actor in Latin America, they should be thrown out with the bathwater. Similarly, if a ‘hegemonic mechanisms’ model³⁴ which gives equal and mutually cancelling importance to (a) the United States Embassy making vague pro-democracy noises in El Salvador during a murderous civil war it was directly supporting and (b) overt backing for a coup in, say, Guatemala in 1954, then how helpful, really, is it down here on earth? Somewhere, a part of the US apparatus was always talking up democracy, to be sure, but I cannot see this constitutes a ‘notable finding’.³⁵

Notes

1. Weyland, ‘Limits of US influence’, 135. ⮭
2. Booth, ‘The limits of expectation’, 2. ⮭
3. Friedman, ‘Retiring the puppets’, 621. ⮭
4. Weyland, ‘Limits of US influence’, 159. ⮭
5. Schenoni and Mainwaring, ‘US hegemony and regime change’. ⮭
6. Schenoni and Mainwaring, ‘US hegemony and regime change’, 269. ⮭
7. Schenoni and Mainwaring, ‘US hegemony and regime change’, 271. ⮭
8. Schenoni and Mainwaring, ‘US hegemony and regime change’, 271. ⮭
9. Schenoni and Mainwaring, ‘US hegemony and regime change’, 271. ⮭
10. Most obviously those of Piero Gleijeses in Shattered Hope and Stephen Schlesinger and Stephen Kinzer in Bitter Fruit. ⮭
11. Schenoni and Mainwaring, ‘US hegemony and regime change’, 270. ⮭
12. See Booth, ‘Rethinking Latin America’s Cold War’. ⮭
13. Schenoni and Mainwaring, ‘US hegemony and regime change’, 270. ⮭
14. Schenoni and Mainwaring, ‘US hegemony and regime change’, 271. ⮭
15. Schenoni and Mainwaring, ‘US hegemony and regime change’, 271. ⮭
16. Schenoni and Mainwaring, ‘US hegemony and regime change’, 272. ⮭
17. Schenoni and Mainwaring, ‘US hegemony and regime change’, 272. ⮭
18. Schenoni and Mainwaring, ‘US hegemony and regime change’, 272. ⮭
19. Schenoni and Mainwaring, ‘US hegemony and regime change’, 273. ⮭
20. This point stands – indeed, it is reinforced – if we expand the scope of enquiry to include US Cold War policy in Africa and Asia. ⮭
21. Schenoni and Mainwaring, ‘US hegemony and regime change’, 273. ⮭
22. Schenoni and Mainwaring, ‘US hegemony and regime change’, 276. ⮭
23. Though the US was also intimately involved in the social and political breakdown that led to Operation Uphold Democracy; see, for example, Weiner, ‘Haitian ex-paramilitary leader’. ⮭
24. See, for example, Weisbrot, ‘America’s subversion’. ⮭
25. See, for example, Lakhani, ‘Did Hillary Clinton stand by?’ ⮭
26. Clinton, Hard Choices, 266. ⮭
27. See, inter alia, MacDonald, ‘The politics of intervention’. ⮭
28. Schenoni and Mainwaring, ‘US hegemony and regime change’, 273. ⮭
29. Schenoni and Mainwaring, ‘US hegemony and regime change’, 274. ⮭
30. Schenoni and Mainwaring, ‘US hegemony and regime change’, 274. ⮭
31. Schenoni and Mainwaring, ‘US hegemony and regime change’, 279. ⮭
32. Schenoni and Mainwaring, ‘US hegemony and regime change’, 274. ⮭
33. Weyland, ‘Limits of US influence’, 138. ⮭
34. Schenoni and Mainwaring, ‘US hegemony and regime change’, 269. ⮭
35. Schenoni and Mainwaring, ‘US hegemony and regime change’, 283. ⮭

Declarations and conflicts of interest