Emerging market politics after the rules-based order
By Sam Wilkin, Director of Political Risk Analytics, WTW
As the rules-based order weakens, will the dramatic changes in emerging market politics over the past three decades such as democratization and market openness be reversed?
In this essay, we gaze into the crystal ball, estimating statistical models to predict the political futures of more than 100 countries worldwide, should the liberal international order continue to decline.
It is an ambitious task, and the model predictions should be taken with a grain of salt. That said, our analysis suggests that recent military coups in Africa may be symptomatic of the weakening order (in addition to recent economic and climate stresses), and further reversals of democracy could be possible in Latin America, East and South Africa, and parts of Southeast Asia. The models indicate that risks of conflict are high in Russia (perhaps obviously) and Southeast Asia.
The model forecasts are more optimistic about the order's economic tenets, but intellectual property rights could come under some threat in major BRICs economies, and investment freedom could be reduced in the wealthier parts of Latin America (such as Colombia, Peru, and Mexico).
Global spread of electoral democracy, 1975-2021
Global spread of GATT/WTO membership, 1975-2021
Source: WTW analysis of data from V-Dem (https://v-dem.net/), United Nations, and WTO. The threshold for electoral democracy is here set at 0.5 on the V-Dem polyarchy scale
During the 1990s, the number of countries adopting electoral democracy skyrocketed, as did the number of countries signing up to global trade agreements (see graphs). Partly, the surge in democracy was a result of the U.S. and USSR reducing support for allied authoritarian regimes.1 Partly, it was because the incentives the order offered for democratization (and disincentives for reversions) became increasingly effective.2
Measured by share of world population, the order’s success in promoting free trade, openness to foreign investment, and respect for intellectual property rights was even greater than its success in promoting electoral democracy. During the late-90s peak, bilateral investment treaties were signed at the rate of almost two a week.3
The order thrived based on what international relations scholars call a “club goods” strategy.4 Roughly speaking, countries that wanted the goods the order had to offer (foreign aid, foreign lending, market access, military equipment, security guarantees) had to join the club; to join the club they had to follow the order’s rules (holding elections, protecting human rights, maintaining international peace, signing up to global deals on trade, investment and intellectual property). Countries that violated the rules were (often) ejected; when they returned to compliance, they were (again, often) welcomed back in.5
This “club goods” strategy was extraordinarily effective during the 1990s, when the multilateral institutions and core states of the rules-based order (roughly, the U.S. and Canada, Western Europe and Japan) found themselves facing almost no competition in providing these goods to the emerging world. Prior to 1991, developing states had the option to turn to the Soviet block for aid or military assistance. After 1991, the rules-based order was, more or less, the only option (some scholars claim that the rules-based order only became a truly global order after the end of the Cold War).6
One might contend that, in a theoretical sense, the global rules-based order came to a sudden end in 2008, when the core states of the order once again found themselves facing competition in providing “club goods” (in 2005, Chinese foreign aid and lending matched the U.S. figure of one year before; in 2007, Chinese foreign aid and lending exceeded that of the U.S.; in 2016, Chinese foreign aid and lending briefly exceeded that of the entire G7 combined – see graphs).7 Arguably, once the order’s states and institutions again faced competition, the order was no longer genuinely global.
Share of global economic output, 1975-2022
Aid and lending to the emerging world, by donor, 2000-2021, US$bn
Source: All figures in nominal USD; “core states” are here defined as OECD DAC member states; WTW analysis of data from OECD Data Explorer; World Bank World Development Indicators; https://www.oecd.org/dac/development-assistance-committee/; Johns Hopkins China-Africa Research Initiative; and AidData at William & Mary, Global Chinese Development Finance Dataset, Version 3.0. Retrieved from https://www.aiddata.org/data/aiddatas-global-chinese-development-finance-dataset-version-3-0
In a more practical sense, the order did not simply end – many of its elements, from democracy to open markets, are yet on display in countries worldwide. That said, as we noted in the executive summary, when under pressure (e.g., from the rush to self-reliance triggered by the pandemic, or political controversies over rising migration) these commitments have proved far weaker and more subject to reversal than at the order’s peak.
Western countries, and particularly the U.S., have in some cases pushed back against these symptoms of decline. There has been a dramatic rise in the use of sanctions – all but unnecessary in the order’s heyday, when a simple threat to withdraw the “club goods” of aid and trade was usually enough to restore compliance (see graph). The rise in violent conflict we noted in the introduction has also, in some part, involved Western countries seeking to uphold the order’s core principles – for instance the coalition attacks on Yemen’s Houthi rebels who have sought to undermine freedom of navigation.
Number of sanctions programs in force globally, by type, 1949-2019
New sanctions designations, by target, 2022 & 2023
Source: WTW analysis of the Global Sanctions Database; https://www.castellum.ai/insights/2022-sanctions-year-in-review; https://www.castellum.ai/insights/2023-sanctions-year-in-review; China figure for 2023 includes secondary sanctions
But sanctions and violence are expensive – almost certainly too expensive to deploy on the global scale of the “club goods” strategy. Moreover, as Russia’s recent economic growth has shown, sanctions are less effective in a world where there are willing purchasers of Russian oil. Russia’s 2024 growth is projected by the IMF at 32% – higher than any G7 member country, despite Russia being the world’s primary sanctions target in 2022 and 2023 (see graph).
Of course, the order’s core members still have “club goods” to offer (and withdraw). But in a more competitive world, the West may be afraid to withdraw those goods even from club members seen to be flouting the rules (as South Africa’s contentious retention of its Africa Growth and Opportunity Act status may suggest).
Arguably, one symptom of the order’s decline has been a ‘coup epidemic’ affecting Southeast Asia and Sub-Saharan Africa. There have been eleven coups in nine countries since 2020, against an average of less than one per year over the prior thirty years.8 In years past, a coup might be followed by the withdrawal of aid or arms or market access and the quick restoration of democracy;9 as of this writing, in not one of the post-coup states have new elections been held.
Do these coups represent an aberration, or a return to normalcy following the exceptional period of the rules-based order? If we estimate a simple model predicting levels of electoral democracy in countries worldwide, the coups that have occurred since 2020 do appear to have, in many cases, returned countries to roughly the level of political rights we would expect, given those countries’ political histories and levels of economic development (see graph).
Levels of democracy in states experiencing recent military coups: comparing 2019, 2024 and a model prediction for 2024
Source: WTW analysis; for details of model and sources see the Appendix
Only Ivory Coast and Niger, despite coups, retain higher levels of political rights than the model expects. Guinea, Myanmar, Mali, Burkina Faso, and Benin moved closer to model predictions; seven other states were already authoritarian before their recent coups (as the model also expects).
The model is also broadly accurate in predicting (non-coup-related) reversions from democracy, between 2019 and 2024, in Afghanistan, Haiti, Georgia, Mozambique, Nigeria, Senegal, Serbia, Sri Lanka, Tanzania, Tunisia, and Zimbabwe; the reversions in El Salvador and Peru, however, are a surprise.10
Is this the pattern we can expect in the years ahead, not only for democracy, but for other elements of emerging market politics that were conditioned by the rules-based order?
In the maps that follow, we show the results of model-based estimates of how emerging market politics might change as the influence of the order declines. Please note that these models are approximations, making make no effort to distinguish correlation from causation (does development cause democratization or does democratization lead to more rapid development – there is evidence for both hypotheses).11 In addition, instead of reading these maps as forecasts, one could read them as a ‘glass half full’ indication of where the order’s influence remains strongest.
We developed simple models of several principles commonly associated with the rules-based order, using explanatory factors found in the academic literature. In general, the models perform fairly well, explaining the majority of variance between countries in 2023 (although the statistical fit of the conflict model is weaker).
The models and their components are:
Electoral democracy, predicted by level of development (mobile phones per capita),12 political history (years of democracy), reliance on oil and gas earnings, and EU membership;
Interstate conflict, predicted by military expenditure, country size (larger countries appear more likely to have hegemonic ambitions), underdevelopment (agriculture as a percentage of output), and external political pressures (as assessed by the Fragile States Index);
Trade freedom, predicted by level of development (income per capita), reliance on oil and gas earnings, and patenting activity;
Investment freedom, predicted by the same factors as trade freedom;
Intellectual property rights protection, predicted by level of development (share of population with access to the internet), patenting activity, reliance on oil and gas earnings, and country size (larger countries are less likely to protect intellectual property); and
Adoption of global standards (e.g., ISO 9000), predicted by level of development (share of population with access to the internet), high-technology exports, reliance on oil and gas earnings, membership of the OECD, and country size (larger countries are less likely to adhere to global standards).
The details of the models are available in the appendix.
Model predictions for changes in politics (democratization and conflict risk) in the absence of the rules-based order
Source: WTW analysis; for details of models and sources see the Appendix
Model predictions for changes in economic institutions (trade freedom, investment freedom, intellectual property protection, adherence to global standards) in the absence of the rules-based order
Overall, our findings suggest that democracy may be most at risk in Latin America, East and South Africa, and parts of Southeast Asia (or, to take a more optimistic view, the influence of the rules-based order continues to be strongest in those areas). There also appears to be some risk of democratic backsliding in Eastern Europe. Conflict risks are highest in most of the same places, but especially Russia (perhaps unsurprisingly) and parts of Southeast Asia.
By contrast, the models suggest that fewer countries are at risk of retreating from the order’s economic agreements.13 The model suggests protection of intellectual property rights could come under some threat in major economies including Russia (perhaps now a foregone conclusion), China, India and Mexico. Trade freedom is predicted to be at greatest risk in Southeast Asia and Central Asia; investment freedom in the wealthier parts of Latin America (such as Colombia, Peru, and Mexico).
These forecasts should be taken with a grain of salt. The model predicts, for instance, that intellectual property rights are at risk in Norway. The vast majority of the world’s oil-rich countries are (unlike Norway) authoritarian states, perform poorly on most economic policy indicators, and have suffered long-term economic stagnation or decline (a phenomenon known as the ‘resource curse’).14 Norway appears very unlikely to revert to this global norm, no matter what the model says.
By contrast, the model predicts the imminent outbreak of democracy in China (as well as in Vietnam and Thailand). Time will tell if these predictions based on global norms are simply naïve; or indeed if the model’s apparently naïve forecasts are symptomatic of a rising alternative order. To see model forecasts for individual countries, examine the country profiles that follow.
Any public discussion of politics is itself politics. There is no way to step outside of politics to provide an apolitical analysis of world events – a fact that has hamstrung leaders’ and policymakers’ ability to discuss many of the geopolitical changes taking place in our world today. One cannot say anything that might one day be used as an excuse for a rival state’s actions, even if that statement may be well-intended, important or even obvious – a lesson some Western policymakers have learned the hard way following the escalation of conflict in Ukraine.
Businesses also struggle with geopolitical transitions, although for different reasons. The complex institutions of business, from the organization of corporate functions to the fine print on legal contracts, are closely attuned to the political and geopolitical environment – a point that is usually not obvious until that political environment suddenly changes.
For example: during the colonial era, legal contracts tended to be well-suited to the harsh strictures of colonialism. If the price of a natural resource went up, for instance, foreign investors often enjoyed mind-boggling levels of profitability.
After World War II, there was a rapid transition from a colonial to a post-colonial world. The UN had 51 member states in 1945; by 1960, the figure had doubled to 100; and then by 1980, tripled to 154, as former colonies became independent. (Today there are 193 UN members, in part due to post-Soviet break-ups.)15
In some post-colonial revolutions, foreign investors lost everything (Iran, Cuba, China). In most cases, host governments seized the specific foreign assets that were misaligned with new geopolitical realities.16 As commodity prices spiked in the 1970s, and huge profits went to foreign investors, host governments often gave up on the time-consuming process of renegotiating contracts and leaned into nationalizations. By some estimates, by the close of the 1970s, nearly a fifth of all foreign investment overseas had been expropriated – including nearly all foreign oil and gas and mining investment in emerging markets, much of which was based on colonial-era legal contracts.17
Will the current geopolitical transition be any kinder to globalized business? The rules-based order (as the intrepid reporters from the Center for Investigative Journalism noted) made the world safe for economic globalization. Presumably, if globalized companies knew then what they know now (that the order had a shelf life), they might not have distributed their production locations, core markets, and vital functions around the world in the way they are currently distributed.
Largest catastrophe risk losses since 1980
Source: https://www.statista.com/statistics/268126/biggest-natural-disasters-by-economic-damage-since-1980/; https://www.ft.com/content/c4ea72b4-4b02-4ee9-b34c-0fac4a4033f5; for an alternate estimate see https://www.reuters.com/markets/europe/foreign-firms-losses-exiting-russia-top-107-billion-2024-03-28/
Adjusting supply chains to match the decline of the order will probably be easiest (although admittedly expensive and time consuming, in part because of unexpected threats, e.g. in the Red Sea). Developing new key markets and relocating offshored functions will probably be harder, as shareholders tend to react negatively to actions that, in the near term, diminish corporate value. In some cases (e.g., Ukraine and Russia) Western companies will linger too long and lose everything (see graph). In all cases, making the correct adjustments will require a level of geopolitical foresight that humanity does not possess, our best efforts at model-building notwithstanding.
It was possible to make money internationally before the rules-based order. It will be possible afterwards. But for many companies, the transition is likely to hurt.
The author thanks Kerry Boyd-Anderson, Stuart Ashworth, Cullen Hendrix, and Jonathan Panikoff for comments on early drafts of this essay. Each model of political and economic institutions ‘after the rules-based order’ is estimated on factors prevalent in the academic literature.
To develop each model, we used the standard approach of taking the most correlated variable, adding it to the regression, and then adding the next variable that was most correlated with the residuals, until additional variables no longer improved the fit.
The models are:
Democracy model (Freedom House’s political rights index is the dependent variable). Independent variables are:
Years of Democracy, from Our World in Data (countries that have been democracies for longer are more likely to continue to be democracies)
Oil and Gas Rents as a Percentage of Economic Output, from the World Bank World Development Indicators (countries relying more on oil and gas are less likely to be democracies)
Mobile Phone Subscribers as Percentage of the Population, from the World Bank World Development Indicators (countries where more people have phones are more likely to be democracies)
EU Membership (EU member States are more likely to be democracies)
Conflict model (where years of conflict over the past 10 years, calculated using data from the Uppsala Conflict Data Program, is the dependent variable). The independent variables are:
Intensity of External Intervention, as assessed by the Fragile States Index (countries experiencing greater external intervention are more likely to engage in conflict)
Population, from the World Bank World Development Indicators (countries with larger populations are more likely to engage in conflict)
Agriculture as a Percentage of Economic Output, taken to represent underdevelopment, from the World Bank World Development Indicators (countries that are more reliant on agriculture are more likely to engage in conflict)
Military Expenditure as a Percentage of Economic Output, from the World Bank World Development Indicators (countries that spend more on the military are more likely to engage in conflict)
Trade freedom model (trade freedom from the Heritage Foundation’s Index of Economic Freedom is the dependent variable) and Investment freedom model (investment freedom from the Heritage Foundation’s Index of Economic Freedom is the dependent variable). For both models the independent variables are the same:
Per capita economic output, from the World Bank World Development Indicators (countries with higher per capita output are likely to have greater trade and investment freedom)
Oil and Gas Rents as a Percentage of Economic Output, from the World Bank World Development Indicators (countries relying more on oil and gas are less likely to have trade and investment freedom)
Patent Applications per Capita, from the World Bank World Development Indicators (countries patenting more are likely to have lower levels of trade and investment freedom). This variable was incorrectly signed from a theoretical perspective, but we kept it in both models because it improved fit. The variable appeared to explain East Asian development approaches, which tend to have strong economic performance, high levels of patenting activity, and relatively low trade and investment freedom
Trade freedom model
Investment freedom model
Intellectual property rights protection model (where the average of strength of patent and trademark protection, from the Property Rights Alliance, is the dependent variable). Independent variables are:
Internet Users as a Percentage of the Population, from the World Bank World Development Indicators (countries with greater internet penetration are likely to have better intellectual property protections)
Oil and Gas Rents as a Percentage of Economic Output, from the World Bank World Development Indicators (countries relying more on oil and gas are likely to have worse intellectual property protections)
Share of Patent Applications that are Made by Residents, from the World Bank World Development Indicators (countries whose residents account for a greater share of patent applications are more likely to protect intellectual property rights)
Size of the Economy (countries with larger economies are less likely to protect intellectual property rights). This variable is somewhat atheoretical but improved model fit and we saw no reason to exclude it
Adherence to Global Standards Model (where the Global Quality Infrastructure Index score is the dependent variable). Independent variables are:
Internet Users as a Percentage of the Population, from the World Bank World Development Indicators (countries with greater internet penetration are more likely to adopt global quality standards)
Population, from the World Bank World Development Indicators (countries with larger populations are less likely to adopt global quality standards)
Oil and Gas Rents as a Percentage of Economic Output, from the World Bank World Development Indicators (countries relying more on oil and gas are less likely to adopt global quality standards)
Technology Exports as a Share of Total Exports, from the World Bank World Development Indicators (countries that are more reliant on high-tech exports are more likely to adopt global quality standards)
OECD member status (countries in the OECD are more likely to adopt global quality standards)
Unless otherwise noted above, all variables were from 2023 or the most recently available year. All variables were significant with the exception of Agriculture in the conflict model (which was significant using other numbers of conflict-years), and ICT exports in the standards model (which was correctly signed in all model specifications we tested).
In general, we attempted to avoid using analyst-assigned variables (e.g., Worldwide Governance Indicators), but broke this rule for the conflict model. We experimented with a number of other dependent variables, such as the Frasier Institute’s economic freedom indices, conflict deaths from the Uppsala database, and the US Chamber of Commerce Intellectual Property Rights assessments, but the above variables were appealing for their large country coverage and superior model fit.
1 https://academic.oup.com/jogss/article/3/1/2/4798911
2 http://pscourses.ucsd.edu/ps200b/Marinov%20Goemans%20Coups%20and%20Democracy.pdf
3 https://unctad.org/press-material/bilateral-investment-treaties-quintupled-during-1990s
4 Claire Provost and Matt Kennard, After Hegemony, 2023.
5 See the examples for Malawi and Ghana discussed in a prior edition of this Index, https://willistowerswatson.turtl.co/story/political-risk-index-winter-2022-2023-gated/page/3
6 https://www.youtube.com/watch?v=7kRtt4Jrd_Y
7 https://docs.aiddata.org/reports/belt-and-road-reboot/Belt_and_Road_Reboot_Full_Report.pdf
8 https://projects.voanews.com/african-coups/
9 Again, see the example for Malawi, https://willistowerswatson.turtl.co/story/political-risk-index-winter-2022-2023-gated/page/3
10 As is the reversion in Ukraine, caused by the escalation of conflict.
11 For one of the essentially infinite number of studies indicating that development causes democracy (a theory advanced by Seymour Martin Lipset in the 1950s), see https://www.imf.org/external/pubs/ft/wp/2012/wp12295.pdf. For democracy causing development, see for instance Daron Acemoglu, Suresh Naidu, Pascual Restrepo, James Robinson, “Democracy Does Cause Growth,” Journal of Political Economy, 2019.
12 For a discussion of such “people power” indicators see https://willistowerswatson.turtl.co/story/political-risk-index-spring-summer-2023-ungated/page/3
13 Although this may be partly because the statistical fit of these models is better.
14 A classic study is Michael Ross, “Does Oil Impede Democracy?” World Politics, 2001.
15 https://www.un.org/en/about-us/growth-in-un-membership
16 Stephen J. Kobrin, “Expropriation as an Attempt to Control Foreign Firms in LDCs: Trends from 1960 to 1979,” International Studies Quarterly, 1984.
17 Geiger, Linwood T. 1989. Expropriation and External Capital Flows. Economic Development and Cultural Change, 37: 535-556; Eaton, Jonathan and Mark Gersovitz. 1984. A Theory of Expropriation and Deviation from Perfect Capital Mobility. Economic Journal, 94: 16-40.
