Acemoglu et al. (2004) solidified the claim of institutions as a long-term determinant of growth. Operationalizing the European settler’s mortality rates as an institutional proxy, they empirically supported the significant role of inclusive institutions in promoting sustainable economic growth (Acemoglu et al. 2001). They later provided empirical evidence for democracy as a foundation for growth, highlighting that democratization has increased long-run GDP per capita by 20 percentage points (Acemoglu et al. 2019).

Effectively, it showcases an integrated structure between institutions as a vehicle to initiate growth and democracy as its ideological foundation. It manifested in the form of a governance system, backed by theoretical rationale and empirical evidence.

The World Bank (2017) supported this premise. Classifying governance as “rules-based” and “deals-based” bargains, it asserts the importance of inclusivity to ensure prolonged economic development. The “rules-based” system, which emphasized accountability, equitable power distribution, and a less personalized form of governance, is claimed to be a more sustainable political and economic structure. Particularly, it advocates the predictability and adaptability of such a system to ensure effective transition and long-term consensus for the rules of governance.

Other scholars have taken alternative approaches to support the finality of democracy. Sen (2013) argued for the intrinsic value of democracy. He affirmed the role of political, economic, and social freedom as instruments of growth. Aligned with the idea of inclusive institutions and “rules-based bargains”, he stated the importance of transparency and protective security as necessary tools to sustain growth.

However, China, Rwanda, and Singapore raises questions about the generalization of democratic institutions as the sole mechanism for sustainable growth. These countries have distinct historical and sociological characteristics. Rwanda is a former Belgian colony. The colony categorized Hutu and Tutsi into distinct racial and hierarchical identities that resulted in a long-standing resentment between the two factions. The conflict culminated in 1994, when roughly 800,000 Tutsi were massacred by the ruling Hutu, driving most of them into exile in Uganda.

Paul Kagame, a Tutsi, led the Rwandan Patriotic Front (RPF), which ended the genocide during the same year. His military intervention is seen as a successful initiative to bring stability to the region. In 1998, Paul Kagame became the president of the RPF, and then became the national president in 2000. He has ruled since.

His governance is full of contradictions. Although he was successful in stabilizing Rwanda’s political and economic environment, the regime that followed is unambiguously authoritarian. Opposition is often suppressed and political opponents imprisoned. Already ruling for more than 20 years, his last election in 2024 still resulted in him winning 99% of the total votes (Kupemba 2024).

Despite such conditions, Rwanda’s economic performance excelled. From 1994 to 2024, Rwanda’s average GDP per capita growth reached a substantial value of 4.85%. With such an average, Rwanda has not only surpassed the economic performance of its neighbours, but also that of most countries in the world. Referenced from the World Development Indicator, the average GDP per capita growth from 1994 to 2024 has only reached 1.02% in Sub-Saharan Africa, -0.04% in Central Africa, 0.7% in Southern and Eastern Africa, and 1.8% in the world. Furthermore, control of corruption in Rwanda has been one of the best performing, showcasing a value of 0.67 in the World Bank’s Control of Corruption Estimate Index.

Figure 1 World’s GDP per capita growth (1994-2024)
Source: World Development Indicators (n.d.)

Figure 2 Rwanda GDP per Capita Growth (1994-2024)
Source: World Development Indicators (n.d.)

If Rwanda became a significant exception to a democratic form of governance, China and Singapore scaled it by multitudes. Both are considered economic miracles of the 20^th century. They have a centralized approach to political governance and experienced a sustained economic growth at an unprecedented level.

China has existed for millennia. It is one of the largest by region and the most populous in the 21st century. At one point in history, they dominated the global landscape. They had significant transformations over the years, experiencing multiple cultural and political revolutions. From the era of the Han dynasty, to the golden age of the Ming and Qing, to the period of modern communist China, they have endured. Thus, unlike Rwanda, they are a wounded giant attempting to restore its peak (Ho 2014).

Singapore tells a different story. It became independent after separating from Malaysia in 1965. In its entire history, it has been governed by a single party. For much of its time, power is also very much concentrated in the figure of Lee Kuan Yew, the founding father of Singapore and the longest serving prime minister from 1965 to 1990. But in 30 years, the country became one of the higher-performing economies in the world. Then in 2024, it had the 2nd highest per-capita GDP in the world in terms of purchasing power parity.

Their story highlights evidence against the necessity of democratic reforms as a basis for growth. Notably, all three share a common structure. Despite extractive political institutions, their economic institutions are inherently inclusive. Acemoglu & Robinson (2013) categorize this combination as unsustainable, yet, it held.

The limited success of autocratic political institutions does not necessarily disprove effectiveness of democracy: it disproves its generalizability. These cases raise enough evidence to highlight the conditionalities for democracy. That is, the institutional and cultural heterogeneity of countries should be taken into consideration when attempting reform.

Mansfield & Snyder (2005) strengthen this argument by highlighting the importance of sequencing in democratic reforms. Without strong political institutions and the rule of law, political participation and elections can backfire, resulting in unintended consequences. Zakaria (1997) further supported the argument by cautioning against the danger of constitutional liberalism with democracy, arguing that such misinterpretation often leads into ineffective prescriptions for development.

Cultural heterogeneity also becomes a distinct factor that influences this conditionality. Lipset (1959) argues that pre-requisites for democratic governance must first be fulfilled. He highlighted that aspects such as level of education, class structure, and wealth are determinants of whether democratic governance would succeed. Chua (2002) extends his to the Asian context, underscoring that the success of Singapore should also be attributed to the concept of ‘Asian democracy’. Thus, he firmly asserts the role of cultural factors towards forms of governance. While Xu (2011) argued that China, with its “regionally decentralized authoritarian” have been effective in promoting reforms given its unique political system, highlighting its adaptability while detaching itself from the idea of Western democracy.

Altogether, this empirical evidence converges into a single idea: the success of democratic reform itself is conditional. Thus, for the case of Rwanda, China, and Singapore, democratic reform should be viewed as a tool rather than an absolute destination. Evidently, China, Singapore, and Rwanda have simply followed a path shaped by its own constraints.

This iterative process was often ignited during a crisis, such as one shown during the Great Depression. In the worsening condition, policymakers were puzzled to find a reliable metric of its economic impacts. It was during this period that Simon Kuznets, a Russian-American economist and statistician, contributed to the creation and formalization of Gross National Product (GNP). It was defined as the total value of output by a nation’s population, regardless of location. His framework later provided the foundation for Gross Domestic Product (GDP), measured as the sum of consumption, government expenditure, investment, and net exports. Adopted as an international standard for economic performance, it has become the primary lens for economic measurement (Vanham 2021).

Yet, the establishment of GDP as an international metric did not come without challenges. Kuznets himself argued that using GDP as a measure of national welfare will overlook other aspects, including the quality of life, environmental degradation, and income inequality. Therefore, despite providing a foundation for GDP, he expressed a profound disagreement with simplifying the overall definition of economic performance as a measure of productive output (Vanham 2021).

This tension illustrates that a metric could be utilized with numerous interpretations. While GDP provides a consensus, Kuznets showcased that a singular definition of growth should be challenged. More importantly, disagreement regarding definitions should be taken as a corrective process to enhance our understanding of economic progress.

Such a sequence illustrates a recurring pattern that ideas, regardless of how well-formed, are constrained by what they can explain at a particular time. History acts as a record that keeps track of ideas and how they were exposed, revised, and advanced. Thus, with the reach contestation, our understanding of how capital, labour, growth, institutions, and technology will be shaped.

After the Great Depression, the idea that “supply creates its own demand”, known as Say’s Law, was rebutted . The world needed a new intellectual axiom, and at the end of the Great Depression, “The General Theory” by John Maynard Keynes, emerged. It argued that instead of wages, employment is decided by the aggregate demand for goods and services. Rebuking the classical theory, he asserts that government intervention is a necessary tool to help the economy reach full employment.

Building on the Keynesian framework, the Harrod-Domar model was one of the earliest to be introduced. It shows that economic growth is strictly a byproduct of the saving rate and the capital-to-output ratio (Boianovsky 2018). The idea became a landmark for capital fundamentalism, emphasizing capital formation as the main driver of economic growth. The model became enormously influential during the late 1940s to the mid-1950s, being one of the first models of the post-war growth theory.

Criticism emerged shortly after its birth. Comments were made highlighting fundamental problems with the model’s assumptions, defined as the knife-edge instability (Hagemann 2009). Particularly, the model assumes that the warranted rate of growth and the natural rate of growth are exogenously given without mechanisms to align them. This assumption produces an extensive period of rising or falling unemployment. In addition, fixed proportions of labour and capital add to the problem, leaving the model fundamentally unsustainable.

In 1956, the Solow-Swan model became a direct response to the Harrod-Domar model by asserting the roles of technological advancement and the substitutability of labour and capital as factors of production (Hagemann 2009). The model empirically proved the idea in 1957 through the “Solow residual”. It illustrates that a large portion of growth cannot be attributed to capital accumulation or population growth, implying the significance of total factor productivity towards income increase. The discovery effectively iterates the notion that capital accumulation is the determining factor of economic growth.

This sequence is not simply an empirical finding or a theoretical advancement, it was an acknowledgement. From a historical standpoint, it validated the claim that definitions of growth and its measurements are rarely final. More importantly, our understanding of growth will only continuously evolve.

The Solow-Swan began by scrutinizing the instability of its predecessors, only to create room for other sources of evaluation. It validated the limitation of capital as a factor of growth, but left the specification of total factor productivity for interpretation. Thus, theoretical evolutions will be limited by our understanding of current phenomena. Each solution will only expand the question, and not conclude it indefinitely.

After dominating growth theory for decades, the Solow-Swan model was eventually displaced by the Endogenous Growth Theory. The new theory improved upon the model, adding human capital and knowledge spillovers to the construct of total factor productivity (Lucas 1988; Romer 1994). Other perspectives, such as the Adaptive Efficiency Model, argues that growth is a byproduct of institutional adaptation, highlighting the conditionalities imposed by evolving institutions on economic growth (North 1994; Simpson et al. 2025).

Together, they define growth as a byproduct of several endogenous factors. They illustrate the limitations of past models as a predictive tool for growth. They did not replace the Solow-Swan model, but our understanding of growth has changed permanently since they emerged.

These models support the claim that defining growth should come as an evolutionary process. The statement that “the economy grows because of fast capital formation” is deceptive in its simplicity and only partially true. Growth is a phenomenon influenced by many factors, and attributing one factor without the other can only produce an incomplete understanding of its nature.

Acknowledging capital formation as a driver of growth is crucial. However, human capital, knowledge, and institutions determine how capital should be accumulated and will be deployed. The Solow-Swan model confirmed the idea that capital and labour can only explain a fraction of growth, and it has proved the limitation of capital accumulation as its primary explanatory factor.

The consequence of misalignment extends beyond theory. The Harrod-Domar model dominated post-war growth discussion, but its failure in creating a sustainable growth mechanism had severe consequences. It undermined alternative growth factors, and overemphasized its own importance and creates a disproportionate burden on developing countries, implying barriers of equitability (King & Levine 1994).

The same goes for the Solow-Swan. After being the primary reference for understanding growth, it failed specify total factor productivity at the empirical level. Without theoretical improvement, technological advancement could only be understood more broadly, significantly limiting understanding of growth factors, and thus, its policy implications.

The past dynamics of GDP and GNP also display a similar conflict. We agreed that growth is measured as an increase in productive output. But such a definition is not final. There will always be room to incorporate other definitions of growth, allowing alternative measurements to take place, and in effect, other models to emerge.

Therefore, the question of defining and measuring economic growth is not a question of empirical specification, but a clash of consensuses. GDP is a construct of economic growth, but not the entirety of it. The same way that capital formation is proven to be only a factor of growth, and not its whole.

Economic growth is not a problem awaiting a final solution, it is a moving object. The progression from Harrod-Domar to Solow to Endogenous Growth Theory is not a story of convergence but a record of succession, for each exposed the blind spots of its predecessor. Disagreement about definition and measurement is therefore not a weakness but a necessary mechanism by which the discipline advances. The right response to any consensus on growth is not to defend it, but to ask what it cannot see.

Comparisons between countries often focus on how the quality of goods and services in developing countries is worse than that in richer nations. Why do advanced countries seem to have better things than the rest of the world? We think it’s due to a simple “no-brainer” reason: some countries are just much richer than others. That is the simple yet best way to explain how the US has better infrastructures, or Japan with its public transport, or why Switzerland is so damn expensive.

We know that Indonesia’s living standards remain far behind those of advanced economies. But, examining how Indonesia have progressed over time compared to other nations should provide a clearer perspective on whether it’s growth trajectory has been typical, faster, or slower than expected.

Using GDP per capita, we can now see how Indonesia compares with the advanced countries. Figure 1 shows the 2024 GDP per capita of several advanced economies. The numbers range from the United States, one of the richest countries in the world with a GDP per capita of $84,534, to Japan, another wealthy nation with about $32,487 per person.

To make it concrete, imagine that the average person in the United States earns about $84,534 in a year, while the average Japanese earns around $32,487. Now compare that with Indonesia, where the average income per person is only about $4,925 per year.

Figure 1. GDP Per Capita Comparison of Indonesia to Advanced Countries

The gap is not marginal. The average American earns roughly seventeen times more than the average Indonesian. While even Japan, with lowest income level among other advanced economies in the figure, has an income per person about six times higher than Indonesia. The United States has more than twice the income per person of Japan. This highlights just how far Indonesia still has to go before reaching that category.

The difference in living standards then should therefore not come as a surprise. These income differences also translate into differences in the nation’s ability to provide public goods for its people. Governments in richer countries can collect far more tax revenue per citizen, allowing them to finance better infrastructure, public services, and social protection systems.

If we compare to countries that started from a similar income level and faced similar development challenges, Indonesia did not exactly win the race either. Figure 2 compares the growth of GDP per capita between Indonesia and several countries that had similar income levels in the past, including China, Vietnam, Peru, and Thailand. In 1985, Indonesia’s GDP per capita stood at around $511. At the time, this figure was actually higher than China ($295) and Vietnam ($238), though still below Peru ($767) and Thailand ($774). However, the trajectories of these countries over the following decades diverged quite significantly.

Figure 2. GDP per Capita Trends of Indonesia and Comparable Economies (1985–2024)

Thailand’s GDP per capita grew from around $774 in 1985 to more than $7,300 in 2024, and Peru increased from roughly $767 to over $8,400 during the same period. Meanwhile, China increased its GDP per capita from just under $300 in 1985 to more than $13,000 by 2024

Of course, Indonesia has also experienced significant growth as well. Its GDP per capita rose from around $511 in 1985 to approximately $4,925 in 2024. This represents nearly a tenfold increase over four decades. But, when placed alongside these comparable countries, Indonesia’s progress appears more stagnant.

Comparisons with countries from similar starting points suggest that Indonesia’s growth, while substantial, has been more moderate than that of some of its peers. Indonesia is progressing. We’re becoming richer, but in a rate slower than those who started poorer. This distinction matters. It should shift our perspective from asking “why is Indonesia poor”, into “what prevented Indonesia from growing at faster rate similar to China, Vietnam, Peru, and Thailand”. And it is a question that Indonesia has yet to answer.

Amid the expansion of higher education, the composition of unemployment by educational attainment has also shifted. Since the late 2000s, the proportion of university graduates among the unemployed has increased significantly, consistently reaching double digits in the 2010s and 2020s. In contrast, the share of senior secondary (SMU) graduates among the unemployed has changed only slightly over time, averaging around 28 % since 1986.

This pattern suggests a structural shift in the composition of unemployment within the Indonesian labour market. That is, the proportion of university graduates’ unemployment has risen relative to the total open unemployment pool. Such evidence signals an absorption failure in the labour market for university graduates, potentially highlighting a labour market mismatch during extensive higher education expansion initiatives.

This evidence presents a conundrum. Isn’t more quality education supposed to be good for the economy? Isn’t it supposed to increase the value of human capital, boost productivity, and thus create more jobs? Why does that not seem to be the case, especially when the Indonesian economy is growing?

First, we have to understand the proposed mechanism for which education could increase wage and employment outcome. Theoretically, Becker (1964) highlights that education is a form of capital investment. Just like how firm invests in machines, individual invests in skills. Both are expected to produce future returns. Becker models it with the following equation.

In the model, Y represents income after education, X represents income without additional education, while C represents the cost of education, and r illustrates its rate of return. The mechanism he proposed is that when people acquire education, they expect some sense of return. When they spend time and resources to obtain an education, including tuition, time spent studying, and foregone earnings, they expect future benefits. This increases the expectation of workers’ productivity by firms, resulting in higher wages and lower unemployment. Therefore, in an equilibrium, education should improve labour market outcomes.

Spence (1973) highlights the function of education as a signal of workers’ ability. Education may not primarily increase skills, but it may signal a worker’s ability to employers. Therefore, in an equilibrium, high-ability workers are predicted to have better education and wages based on observed signals taken by firms and hirers.

However, Indonesia presents a case where these models are not functioning properly. That is, Becker and Spence argued that education is supposed to produce a positive signal of workers’ ability and enhance labour market outcomes. Evidently, there seems to be a discrepancy of the proposed mechanism resulting in what seemed to be a labour market failure.

Such failures have been studied in extensive mismatch literature. Duncan and Hoffman (1981) argued that overeducation could happen when the education level increases more rapidly than the skill requirements for available jobs. In such cases, workers may obtain education levels that exceed the qualifications required by their jobs. This leads to a schooling surplus, where individuals are hired below their education level, resulting in an inefficient allocation of human capital.

Structurally, Pritchett (2001) highlights evidence that educational attainment in developing countries does not directly contribute to growth as predicted by foundational economic models. He argued that mismatches and institutional arrangements were sometimes unaccommodating of growth, resulting in counterproductive educational initiatives. In particular, he highlights the case where countries’ demand for growth for higher-skilled labour failed to meet the pace of their expansive educational initiatives. He also highlighted that their ability to transmit knowledge and skills effectively through educational programs also plays a role. This signals a potential case when rapid education expansion will not translate into employment opportunities.

A related structural explanation is presented by Dani Rodrik (2016), who argued that a misaligned transformation, illustrated by premature deindustrialization, will hinder growth for developing countries without a comparative advantage in manufacturing. He argued that manufacturing will produce an unconditional labour productivity convergence, meaning that it will produce growth quickly if they industrialized. He also showcased that manufacturing can absorb low-skilled workers more effectively, in contrast to non-industrialized and service sectors that require highly skilled labour. He illustrates the case of some African and Latin American countries that have been enduring the consequences of premature deindustrialization, showing signs of an increase in informality and petty services despite growing investment. These underline the central role of manufacturing in driving catch-up growth, and the danger of prematurely transitioning into a high-skilled sector.

For the demand side, Van De Werfhorst (2009) presents a case where education has evidently transformed into a positional good. He highlights that educational pursuits were also significantly driven by social-class pressure, underscoring the phenomenon of credential inflation amidst the loss of educational value. As a result, demand for education steadily rises even without rewarding job market consequences.

Altogether, these combinations may signal that Indonesia is experiencing structural challenges, facing a mechanism failure in the current labour market ecosystem. With a growing number of educated populations and a rising share of university graduates in unemployment composition, Indonesia’s labour market failed to absorb a significant number of the educated workforce. This raises an important policy question: What is the current strategy for Indonesia’s government for higher education expansion, and where will it take us?

Indonesia’s grand development strategy for education is embedded within a broad national development strategy, emphasizing structural transformation. The National Long-Term Plan (RPJPN 2025-2045) acknowledges the weaknesses of the current manufacturing climate, the signs of premature deindustrialization, and poor research and development capability. The document emphasized the importance of higher education initiatives to increase productivity, along with downstreaming as a strategic initiative for growth and development.

On paper, Indonesia’s long-term development plan presents a coherent framework for sustainable growth. However, its success is highly dependent on short-term implementations. And so far, the alignment between the short-term trajectory of labour market development and its long-term vision has been questionable at best.

Indonesians are enduring problems regarding labour mismatch, including overeducation and undereducation. As discussed, unemployment amongst university graduates having also been rising. This pattern is shows that there is a clear sign of labour market misalignment amidst HE expansion.

If we examine the composition of undergraduate students in 2025, most are enrolled in the degree that would presumably directly contribute to the service and non-manufacturing sectors. Primarily, enrolment are concentrated towards education, social, economics, and healthcare sectors.

Furthermore, if we take a look at the employment composition, the primary sector, being agriculture, forestry, and fishing, along with service sectors (retail, accommodation, and education), still dominate the employment share of the job market. More importantly, they are growing at a similar rate to the secondary sector, signalling labour market absorption failure in manufacturing industry. This highlights that the manufacturing sector failed to grow at a targeted rate despite evident policy prioritization.

In response to the current labour market situation, the government tend towards a more interventionist approach. Welfare programs such as Makan Bersama Gratis (MBG) signal massive boosts for the service sector, particularly the food and agricultural sectors. However, whether such an initiative will translate into a meaningful development consequence, especially in the middle of rising unemployment, remains an open question.

Another massive program, Koperasi Merah Putih, also became a source of scrutiny. With a promise to initiate a top-down institutional design for more than 80,000 rural cooperatives, the government also proclaimed it as a substantial program to help ease the problem of rising unemployment. Through loans channelled by Bank Himbara, the government presented Koperasi Merah Putih as a vehicle for rural development. However, being concentrated in the service sector, the prospect of Koperasi Merah Putih becoming a proper vehicle for substantial labour market development is still highly doubtful. Not to mention, with the current institutional design, one being a top-down initiative that contrasts the autonomous and bottom-up approach of cooperatives, a question of the government’s ability to monitor and incentivize “workers” is still debated. All of which are oriented towards primary and tertiary sectors.

Other programs, such as TKDN and Danantara, signal commitment to strengthening the national manufacturing industry. However, amidst criticism of protectionist tendencies, rising costs, and perceived inefficiency, its ability to generate employment opportunities and growth in the manufacturing sector would be highly dependent on technological adoption and utilisation. Yet, most of the current technology was obtained during the previous administration. Evidence also shows that despite prioritization for manufacturing, labour market absorption for the sector remains stagnant, signalling a transition barrier towards high-skilled jobs in the manufacturing sector for Indonesians.

This evidence provides a mixed perception of Indonesia’s development strategy. Tension arises due to the lack of coherence between the long-term plan and short-term execution. The government take notice of the importance of strengthening the Indonesian manufacturing sector for sustainable growth. But, the problem regarding labour market absorption for the educated remains, for which most do not gravitate towards manufacturing. As a result, growth of education will significantly outpace the demand for quality jobs, resulting in a labour market mismatch and rising unemployment for the educated.

All in all, while the higher education expansion initiative is a necessary investment to prepare for future growth, it is also imperative that industrial sequencing and short-term alignment are also oriented to reduce unemployment and increase the workforce’s productivity. Ultimately, an alignment between industrial and education policy would be a crucial determinant in addressing the long-run consequences of higher education expansion. Whether it proves to be a worthwhile investment or a wasted opportunity will depend on the evolution of these policies in the years ahead.

This is the moment of realization that macroeconomics is not just a list of slogans, but a discipline about mechanisms — about how outcomes are produced, and how those outcomes depend on institutions, expectations, constraints, and assumptions.

And even when the policy and its mechanism “works”, the bigger question we should be asking is: Inflation on what? Inflation for whom? or even better, who wins and who loses? The rich? The poor? People living paycheck to paycheck? People who barely even know what an interest rate is, but still feel its effects every month? These are not side questions. They go straight to the heart of how the economy works. This is where frontier macroeconomics begins.

Let’s take a step back and understand what do we really mean by macroeconomics. At its core, macroeconomics is a study of the performance of the national and global economies. Traditionally, the two headline questions were growth and cycles: why living standards rise in the long run, and why economies boom and slump in the short run.

But the modern version of macro doesn’t stop there. It asks: growth for whom? cycles for whom? That shift happened for a practical reason. When real economies went through episodes like jobless recoveries after the 2008 global financial crisis, decentralization reforms that shifted fiscal responsibilities, and the massive income shocks like the COVID-19 pandemic, it became harder to pretend that the “average household” is the whole economy. The distribution of income, wealth, and risk shapes how policies transmit through the economy — and ultimately, whether they work at all. Inequality has therefore become an increasingly central theme of modern macroeconomics.

Frontier macroeconomics starts with its bottom-up approach: microfoundations. Instead of treating the economy as a single machine moved by a few aggregate relationships, modern macro asks a more disciplined question: what decisions are households, firms, banks, and governments actually making, and how do those decisions add up to aggregate outcomes?

A useful place to begin is the Solow–Swan model1, the classic in growth theory. It is powerful, but deliberately simple. Many of its key drivers are treated as given — or, in economic terms, exogenous. In its standard form, the model starts with a production function in which aggregate output (Y) is produced using capital and labor.

The backbone of the model is capital accumulation: not all output is consumed today, and the portion that is saved is reinvested to build the capital stock for the future. This process is what drives growth over time. A simple way to think about it is “saving an egg today so you can have more chickens tomorrow.”

One of the most important elements in the Solow model is the saving rate. The model assumes households save a fixed fraction of income. That simplicity is useful, but it also leaves an important question unanswered: why do households save that amount in the first place?

Modern macroeconomics pushes directly on that gap. Instead of taking saving as given, microfoundations ask: how does saving emerge from household decisions? How do people choose between consuming today and saving for tomorrow? Since households do not simply spend everything at once, theorists model that trade-off explicitly through an intertemporal utility maximization problem. In other words, saving is no longer imposed from outside the model; it becomes the result of optimization.

These optimization behavior of not only households, but also firms, governments, and other agents marks the transition from older growth models to the broader architecture of modern macro. The diagram below gives a broad roadmap of how these models evolve and connect to the benchmark frameworks used in modern macroeconomics. What follows is a quick tour of some of the most important ones.

• Neoclassical Growth Model (NGM) / Ramsey-Cass-Koopmans234

  This is the classic micro-founded growth model. It explains how households choose consumption and saving over time, and how those choices shape capital accumulation and long-run growth. It is used as the foundational benchmark for modern dynamic macro models.

• Overlapping Generations (OLG) Models56

  People do not live forever. The OLG model introduces finite lives: people are born, work, save, retire, and die. It is designed to study how different generations interact. It is especially useful for analyzing pensions, aging populations, public debt, social security, and intergenerational redistribution.

• New Keynesian (NK) Model7

  Prices do not adjust flexibly. The New Keynesian model takes the dynamic structure of modern macro and adds sticky prices, nominal rigidities, and imperfect competition. Its main purpose is to explain why monetary policy has real effects in the short run. It is widely used to study inflation, interest rates, business cycles, and stabilization policy.

• Heterogeneous-Agent (HA) Model89

  The HA model drops the idea that everyone in the economy is basically the same. Instead, households differ in income, wealth, risk exposure, and constraints. These models are used to study how inequality, incomplete markets, and household-level shocks shape aggregate outcomes.

• Heterogeneous-Agent New Keynesian (HANK) Model10

  HANK combines heterogeneous households with a New Keynesian monetary framework. Its main contribution is showing that monetary policy does not affect all households in the same way. It is increasingly used as a modern benchmark for monetary policy analysis, especially in central banks, because it captures distributional effects and more realistic transmission channels.

• Search and Matching Models111213

  This models focus on labor market frictions. Workers do not instantly find jobs, and firms do not instantly find suitable workers. These models are used to explain unemployment, vacancies, wage bargaining, job creation, and labor market dynamics.

One application of the OLG models is the study of pension reform in Indonesia. Kudrna, Piggott, and Poonpolkul (2022)14 show that extending pensions can raise welfare, especially for older informal workers, but also creates long-run fiscal trade-offs as the population ages. This shows how policy affects different generations and fiscal sustainability over time.

In essence, frontier macroeconomics is not a rejection of traditional macroeconomics. It is when the old questions are taken more seriously. Growth still matters. Business cycles still matter. But modern macro asks those questions with sharper tools: by making behavior explicit, by taking frictions seriously, and by recognizing that households and firms are not all the same. The field has evolved from simple aggregate relationships towards models built upon optimization (why households choose to consume, save, or work), expectations (what do people believe about inflation, wages, or policy tomorrow), and constraints (how firms finance investments, how household borrow, and where markets fail).

This matters most when the economy is hit by shocks or policy changes. A tax reform does not affect capital owners and workers in the same way. An interest rate hike does not hit rich households, indebted households, and hand-to-mouth households equally. A recession is not just a fall in output; it is also a change in hiring, job finding, and household risk. Frontier macroeconomics matters because it gives us a way to study those mechanisms directly, rather than hiding them behind one “average” household.

That said, frontier macroeconomics still faces important critiques. Even sophisticated models depend heavily on assumptions, and microfoundations do not automatically make them true. Expectations may be modeled too neatly, data are often hard to identify cleanly, and policy conclusions can change depending on which frictions the model emphasizes. In light of this, frontier macro should not be seen as a one-fits-all framework. Rather, macroeconomic models are useful precisely because they help isolate specific mechanisms and clarify what drives economic outcomes in different contexts. As Dani Rodrik puts it in Economic Rules: “The correct answer to almost any question in economics is: It depends.”

References

Reformasi was expected to be a turning point for Indonesian democracy. Law Number 22 of 1999 on Regional Government and Law Number 25 of 1999 on Fiscal Balance between Central and Regional Governments became significant milestones, effectively laying the foundation for the independence of subnational governments in domestic political affairs. For a moment, the pillars of change seemed established, and the promise of a new Indonesia seemed to be within reach.

Twenty-seven years have passed since, yet the dream of a new Indonesia still feels long overdue. Stalled by institutional failures, the steady erosion of democratic principles, and overly concentrated power structures, Indonesia has failed to deliver on its promise of equitable development. And the very heart of it, decentralization, becomes a subject of paradoxical scrutiny.

The 2001 Decentralization, or “The Big Bang Decentralization”, was enacted to redistribute the political, fiscal, and administrative capacity to local agents. It was legally anchored on Law Number 22 of 1999 and Law Number 25 of 1999, effectively increasing the autonomy for municipalities and city-level governments. Although theoretically aligned with Oates’ (1972) framework, it is conditional on the existence of a robust and democratic local institution.

Evidently, the post-decentralization consequences of Reformasi have produced mixed socio-economic results. Several indicators have highlighted alarming economic conditions, as evidenced by a steady increase in the Gini Index. In addition, Freedom House’s score has also indicated a steady erosion of democratic quality in Indonesia, plateauing in the late 2000s.

Several pieces of evidence have also highlighted the nuanced results of the post-decentralization period. Although granted higher authority to obtain local revenues, changes in the proportion of locally sourced revenue (PAD) in respect to central transfer remain dormant for most local governments. Furthermore, the widening of corruption distribution by elected officials also signals institutional failure, producing a flawed check-and-balance procedure that was once thought to be superior in a decentralized system.

With the creation of a profound legal basis and an apparent shift towards the democratization of Indonesia’s political culture, we then ask: what went wrong? Have the institutional and social conditionalities for the Indonesian landscape yet been met for an effective decentralization? Are we moving too fast, too soon?

Historically, Indonesia is not the only one experiencing a disappointing turn. Numerous decentralizing nations, particularly in Asia and Latin America, also encountered mixed results. Using cases from the Philippines and Pakistan, Guess (2005) highlights that the implementation problem by supervisory institutions often determines the success of decentralization. He specifies aspects that include fiscal and political support, individual capacity of policymakers, modification of cultural constraints, and the strength of institutions as determining factors of regional decentralization’s effectiveness.

Malesky and Hutchinson (2016) emphasize that the deficiency in implementing decentralization is caused by the resistance of authoritarian elites, who fear losing control. Using examples from Cambodia, Myanmar, Thailand, and Vietnam, they highlighted that the ability to choose the local government as a training ground is highly dependent on the central government’s political will. Thus, authoritarian governments will attempt to restrict such educational initiatives.

The study from Faguet and Shami (2022) illustrates a contrasting comparison between two decentralizing nations, with Bolivia decentralizing under a democracy and Pakistan under an authoritarian regime. Both were differently designed and had different outcomes, highlighting the cultural and institutional conditionalities of decentralization.

In Bolivia, decentralization was initially enacted as a response to the decline of the MNR (Movimiento Nacionalista Revolucionario), providing an opportunity for the then-government to gain loyalty from Bolivia’s rural voters. In addition, decentralization also offers a response to the regional pressures that demanded a greater fiscal and political autonomy from La Paz. The Bolivian decentralization was enacted with several mechanisms, including the enforcement of ownership of infrastructure and responsibility for the provision of primary services by local governments, twenty percent transfer of national tax revenues to municipalities, allocation of funds to municipalities on a strict per capita basis, creation of Oversight Committees to enhance accountability by civic organizations, the creation of new municipalities, and expansion of existing municipalities. The result was immediate and significant, dramatically shifting resource allocation to the smaller municipalities. Bolivian subnational governments increased their responsiveness to local needs, and decentralization became a lasting national characteristic.

On the other hand, Pakistan’s decentralization was designed to increase the legitimacy of Musharraf’s leadership by initiating the creation of new local governments and extending some existing services. By overthrowing an elected government, Musharraf’s decentralization prioritized the exclusion of the existing political establishment. The proposed mechanism lacked the institutional enforcement necessary for an effective decentralization, resulting in an architectural system characterised by poor service responsibilities and reporting, along with an ambiguous procedure of fiscal flows. The decentralization gained little support from the Pakistani grass-roots and significant opposition from existing political parties. In effect, Musharraf’s decentralized system only lasted for 7 years before being abolished in 2008, along with his resignation.

For Indonesia, the mechanism of decentralization is built upon the pillars of political, administrative, and fiscal aspects, uniformly applied to all subnational regions. Firstly, decentralization forces a delegation of authority to local governments through local elections for subnational government heads and the Regional House of Representatives (DPRD) at the district level. Secondly, the authority on the provision of most local public services is transferred from central to subnational governments. Lastly, local governments gained increased authority for managing expenditures and are encouraged to collect local taxes through the mechanism of regionally owned revenue (PAD).

Prior to decentralization, Indonesian districts function under the strict control of the central government. Under the framework of Law No. 5 of 1974, regional governments were predominantly reliant on the central government. Mostly, revenue sources are obtained from the Subsidy Fund (SDO) and Instruction Funds (Dana Instruksi), with minimal revenue sharing from natural resources. Thus, the 1999 laws replaced the former model of dependency with an entirely new framework for self-governance and independence.

Apparently, the uniform approach to Indonesian decentralization, although equitable in spirit, has failed to consider the asymmetrical economic and social landscape of affected districts. With significant disparities in fiscal capacity, public infrastructure, and human resources quality, Indonesian decentralization rested on the conceptual axiom that greater autonomy, regardless of institutional readiness, would produce positive social and economic consequences for treated districts. An increasingly contested idea based on recent evidence.

By such a mechanism, Indonesian decentralization has yet to take into account the unique adaptability of regions. The implementation rests on the assumption that greater independence allows districts to exercise control over resources, that independence and autonomy should be an unambiguous strength and a driving factor for better economic growth. However, such an approach undermines the importance of addressing different constraints of regional government.

Between 1996 to 1999, Indonesian districts are structurally unequal in several foundational aspects. Using the dataset obtained from the World Bank’s Indonesia Dataset for Economic and Policy Research, it is shown that the socio-economic conditions vary significantly. The result indicated that aspects such as regional human development index (HDI), fiscal capacity, physical infrastructure, and government service delivery, differ severely.

It highlighted the widespread inequality in areas such as regional economic conditions, fiscal capacity, and service delivery. Furthermore, it is indicated that although the human development index is moderately unequal, subnational advancement in aspects of human capital, as measured by the percentage of net secondary school enrolment, varies significantly among districts.

Not to mention, other measurements also indicated severe variation in the quality of provision of basic needs, which include access to electricity and sanitation. Moreover, health service indicators, as illustrated by the percentage of child births attended by health workers, showed significant inequality. With such a condition, it highlights the disparities in the local government’s ability to provide basic services, indicating evidence of unequal service delivery.

Other regional economic indicators also showcased similar results. Proxied by percentages of roads already covered with asphalt, the physical infrastructure of Indonesian districts was severely unequal prior to decentralization. This aspect is especially important for the Indonesian development agenda, underlining the existence of a structural barrier for equitable subnational economic activities.

Some indicators showed more staggering results. The poverty gap indicated a coefficient variation of more than 0.9, highlighting the existence of extreme inequality among regions. Furthermore, fiscal independence, proxied by the PAD share over total revenue, exceptionally reached over a 1-point value of the coefficient variation. These aspects revealed a foundational discovery for Indonesian districts prior to decentralization, that institutional capacity varies extremely amongst regions. Thus, their adaptability towards the uniform “Big Bang Decentralization” is highly influenced by these extraordinarily different prerequisites.

Based on such evidence, the Indonesian decentralization sparked an important perspective, one that highlights the potential dilemma of undertaking a uniform approach to distributing power. Particularly, the Indonesian case highlights the need to anticipate structural constraints. That is, one that takes into account the institutional adaptability of affected regions.

A simultaneous, uniform, and significant increase in fiscal resources and the local government’s authority may result in unintended consequences if implemented in a landscape characterised by severe inter-regional inequalities. One unambiguous premise following the 2001 power distribution is that local governments uniformly gained abundant opportunities but were strained by an evidently unique self-imposed limitation. However, for the Indonesian case, increasing central transfers or electing local government officials do not seem to tackle these limitations. Thus, it reveals such structural constraints that are possibly unaddressed.

By instituting the 2001 decentralization reform, Indonesia has created a framework for equitable power distribution. But, behind the legal hardware, lie such deeper complexities. One that is intangible and only seemingly permanent.

Indonesia did not decentralize into a neutral landscape. Rather, it unfolded across regions with severe administrative, fiscal, and service inequality. Yet, even though power can be distributed simultaneously among regions, capacity differs in character.

It is not that the decentralization is wrong, but it has revealed a foundational principle amongst Indonesian regions. The success of autonomy and independence is conditional. It may expose the fragility of some and accelerate the opportunity of others.

The lesson is that institutional sequencing matters. Power can be transferred, but not the institutional ability that comes with it. When power moves faster than institutions, the promise of empowerment may strain and risk the benefits of the transformation it brings. Two decades later, decentralization remains a discussion, and with it, its paradoxical manifestation.

The idea of cooperative-based economics has long been debated in the Indonesian development agenda. Historically, ideas of cooperatives gained national momentum after Indonesia’s independence in 1945. Amongst many of its advocates, the first vice president of Indonesia, Mohammad Hatta, strongly believed in cooperatives as a suitable foundation of Indonesia’s economic progress. In his studies, such as “Cooperatives in Indonesia” and “Cooperatives and the People’s Economy”, he argued that cooperatives would be a fitting vehicle to foster rural development by providing small farmers more authority over their economic future.

However, cooperatives gradually became one of the sources of national divide in the 1960s. They were viewed as increasingly politicized and were often pressured to support certain parties. Thus, once viewed as an indispensable foundation to Indonesian economic development, cooperative ideas became increasingly contested.

The slowdown continued after the presidency of Soekarno. In an attempt to further curb left-wing political influence, Soeharto’s regime further dismantled cooperative ideas, resulting in a significant drop in the number of cooperatives from 1966 to 1967. Existing cooperatives were then replaced with Koperasi Unit Desa (KUD) and were often viewed as a top-down initiative to increase the central government’s presence in rural areas. Such an approach effectively differed from the core principle of 1945’s cooperatives by detaching itself from the pillars of autonomy and independence.

Cooperative ideas regained momentum in the post-1998 period due to numerous economic stimuli from the newly formed government. Consequently, the number of cooperatives increased dramatically from 1998 to 2001. Nevertheless, it is evident that many cooperatives in this period were mostly inactive and heavily dependent on the central government. Over the period of 2019 to 2024, the government stated that a substantial number of cooperatives, reaching 82.000 units, have been disbanded due to inactivity.

The beginning of Koperasi Merah Putih marks a new chapter for Indonesian cooperatives. In his statements, Prabowo firmly asserts the cooperative’s role as a tool to level the playing field for weaker economic agents. Particularly, he highlighted the significant role that cooperatives will play in increasing value chain efficiency and food sovereignty. Thus, with such an agenda, Prabowo effectively reintegrates the role of cooperatives into the Indonesian development blueprint, a renewed attempt to revive cooperative-based economic thinking.

Despite the second wind, staggering differences in principles between Koperasi Merah Putih and the cooperative idea from 1945 quickly became a source of scrutiny. Koperasi Merah Putih utilizes the Central State Budget as its main source of funding. Funding is also supported by the Village Fund, the Regional State Budget, and People’s Business Credit (KUR). Furthermore, Koperasi Merah Putih involved several state-owned enterprises, including Bank Rakyat Indonesia (BRI), Bank Mandiri, Bank Nasional Indonesia (BNI), Bank Syariah Indonesia (BSI), and the State Treasury Service Office (KPPN) to distribute state funding. The State Treasury Service Office also assumes additional responsibilities, including monitoring, evaluating, and digitalizing transactions.

This evidence highlights the centralized nature of Koperasi Merah Putih’s financing architecture, sharply contrasting the autonomy-based cooperative ideal of 1945. Not to mention, several human capital constraints emerged due to managerial, financial literacy, and monitoring issues amongst its members. In consequence, several human capital development programs were employed by enrolling cross-sectoral organizations.

Thus, with such contrast, many began to question the basis of Koperasi Merah Putih as a substantive development initiative, or whether it is a form of recycling flawed historical cooperative models under renewed political rhetoric. Rather than incentivizing the growth of cooperatives through a bottom-up policy mechanism, the government has chosen a sweeping, centralized program under a single national framework. This places considerable pressure on the state institutions to nurture, monitor, and ensure the effectiveness of Koperasi Merah Putih. Furthermore, this decision reveals the structural differences in the institutional approach taken by autonomous, community-driven cooperatives and the centralized model of cooperatives, while placing a substantially higher administrative burden on responsible ministries and SOEs.

Past studies have scrutinized the form of cooperatives and its conditionalities. Referencing cross-country comparison, Hans Hansmann, in his book “The Ownership of Enterprise,” underlines that in the early 1970s, cooperatives were able to acquire a substantial amount of market share amongst developed countries. Particularly in Europe, cooperatives have accounted for 45 percent of the total agricultural share in France, 48 percent in Germany, 60 percent in the Netherlands, 70 percent in Denmark, and 80 percent in Sweden. Evidently, cooperatives have been successful in incentivizing farmers to effectively negotiate, or to some extent, displace middlemen. By removing monopsonistic agents, cooperatives have also displayed an unambiguous improvement in social welfare. Additionally, Hans Hansmann underlines reasonings that include overcoming asymmetric information and risk bearing as a stimulus for the formation of cooperatives.

A foundational characteristic of cooperatives is a high degree of interest homogeneity amongst its members. This alignment allows collective decision-making without imposing a disproportionate burden on members based on shared decisions. Interest alignment is essential, as it reduces internal conflict and ensures members share similar incentives. In addition, capital is primarily raised from the cooperatives’ members, while external borrowing is utilized as a supplement to collective contributions.

Although external borrowing is acknowledged as an employable form of capital raising for cooperatives, Hansmann underscores in his study that the majority of equity capital in his observed cooperatives is raised from members. For a farmer-owned enterprise, liquidity constraints are evidently much less important than initially expected. In addition, he highlighted that most successful cooperatives have often carefully designed methods to raise capital, adhering to the principle of independence. Therefore, his findings suggests that accumulation of capital is not, and should not, be the most important barrier for effective cooperatives.

Another study by Uma Lele, titled “Co-operatives and the Poor: A Comparative Perspective,” also highlights several additional conditionalities for effective cooperatives. She argues that human capital qualities such as leadership, management capability, experience, and technical competence deserve greater emphasis in building functional cooperative institutions. Furthermore, she contends that external policies related to agricultural pricing, input distribution, technology, infrastructure, and regulatory frameworks must first be improved to ensure meaningful grassroots participation and reduce the dominance of rural political elites.

Ironically, the current institutional design of Koperasi Merah Putih risks repeating the same problem as past cooperative programs. Historically, a centralized intervention, whether in the form of a KUD or extensive subsidy, largely results in dysfunctional cooperatives. That is, an administratively present but functionally dormant institution.

When cooperatives emerged not from the genuine collective interest, but were dominated by a uniform national instruction, they tended to weaken leadership, cripple the internal cohesion, and lack a uniform incentive. Without a strong foundation for local ownership and internal capability, Koperasi Merah Putih will only be the government’s vehicle to channel credit, without adding much value to productivity and farmers’ autonomy. Thus, with such a scheme, the enormous financial and administrative burden in regard to human quality and institutional development will be placed on cross-sectoral organizations, which have shown a disappointing result in the past.

In response, greater emphasis should be placed on enacting incentivizing policies that allow cooperatives to grow organically. Rather than creating thousands of top-down institutions, the government should focus on increasing the attractiveness of cooperatives by removing the structural barriers that prevent them from functioning effectively. Regulatory improvements related to pricing, input distribution, technology, and infrastructure are essential to creating a supportive environment for cooperative activity. In addition, strengthening human capital through leadership training, managerial development, and technical capacity should be prioritized to stimulate genuine grassroots initiative and increase meaningful participation within cooperatives.

All in all, the success of Koperasi Merah Putih should not be measured by sheer numbers, but by whether it aligns itself with the institutional principles that enable cooperatives to flourish. If these lessons remain unaddressed, Koperasi Merah Putih risks becoming yet another well-intentioned program that fails to create substantial change. By prioritizing autonomy, improving regulatory conditions, and strengthening human capital, the government can not only encourage the development of effective cooperatives but also assess whether this initiative is a meaningful path toward rural empowerment. Whether Koperasi Merah Putih ultimately embodies this opportunity or risks becoming a well-intended backfire, remains an open question that the government must urgently confront.