1.1 Introduction to the Core Puzzles of Growth

The study of economic growth stands as the central inquiry of modern macroeconomics. It seeks to answer the most fundamental questions in economics: Why are some countries so much richer than others? What allows some economies to experience sustained improvements in living standards while others stagnate? Before we can construct theoretical models to answer these questions, it is crucial to first understand the key empirical regularities and puzzles that these models must address. These “stylized facts,” derived from decades of data on national incomes and welfare indicators, form the bedrock of our investigation and motivate the entire theoretical endeavor that follows.

A defining feature of the post-war global economy is the immense and growing disparity in incomes per capita. An analysis of the world income distribution, using Purchasing Power Parity (PPP)-adjusted GDP per capita, reveals a significant “spreading out” between 1960 and 2000. As shown in Figure 1.1, the distribution of income, which was relatively clustered in 1960, became far more dispersed by 2000. This pattern, often described as divergence, indicates that the income gaps between rich and poor nations have widened over this period. Examining the logarithm of income per capita (Figure 1.3), which is more appropriate for variables that grow proportionally over time, reinforces this picture of divergence, showing the distribution stretching out at both ends. This picture of divergence between national averages, however, coexisted with a powerful counter-current at the level of the world’s population. The remarkable growth of China and India, particularly in the 1990s, lifted hundreds of millions of people from poverty, creating a significant force toward the relative equalization of income among the inhabitants of the globe. Nevertheless, the widening gap between countries remains a central puzzle for growth theory.

These vast differences in GDP per capita have profound welfare implications. Income is not an end in itself but a means to a higher standard of living, and the data shows a powerful association between national income and other key welfare indicators. As illustrated in Figure 1.6, there is a strong positive correlation between log GDP per capita and life expectancy at birth. Citizens of richer countries can expect to live significantly longer, healthier lives than those in poorer nations. However, the relationship between growth and welfare is not always straightforward. The case of South Africa under apartheid demonstrates that aggregate GDP growth can occur alongside declining real wages and living standards for the majority of the population. This stark example underscores that the distribution of gains from growth is critical and that rising inequality can accompany economic expansion.

How can one country become more than thirty times richer than another? The answer lies in the compounding power of different growth rates over long periods. Table 1.1 provides a compelling illustration of these divergent paths. While the United States experienced steady growth, countries like South Korea and Singapore transformed from relative poverty into high-income economies, representing “growth miracles.” In stark contrast, countries like Nigeria and Guatemala experienced periods of stagnation or even economic malaise, falling further behind. The experience of Brazil, which grew rapidly between 1960 and 1980 only to stagnate thereafter, highlights the fragility of the growth process. These disparate experiences show that small, sustained differences in annual growth rates, when compounded over decades, lead to the enormous income gaps we observe today.

These empirical realities of vast and growing income divergence across nations set the stage for our theoretical inquiry, which begins with a closer examination of the conditions under which incomes might instead converge.

1.2 The Question of Convergence

The observation of income divergence naturally leads to the question of whether there is an opposing force: convergence. The concept of convergence comes in two primary forms. Unconditional convergence is the hypothesis that poorer countries, all else being equal, should grow faster than richer ones. This implies that, over time, income gaps between all countries should automatically narrow, leading to a more equal world income distribution. In contrast, conditional convergence posits a more nuanced relationship: countries grow faster the farther they are from their own long-run potential, or “steady-state,” level of income. This means that income gaps will only shrink between countries that share similar structural characteristics, such as saving rates, population growth, and economic policies.

When we examine the data for the 1960-2000 period, the hypothesis of unconditional convergence finds little support. Figure 1.14 plots the average annual growth rate of GDP per worker against the initial level of GDP per worker for a broad sample of countries. If unconditional convergence were occurring, we would expect to see a clear negative relationship—a downward-sloping cloud of points where countries starting poorer (on the left) exhibit higher growth rates. The scatter plot, however, shows no such relationship, effectively refuting the idea of simple, unconditional convergence on a global scale.

While unconditional convergence fails, the data does suggest a pattern of conditional convergence. The evidence indicates that the postwar income gap between countries that share similar structural characteristics has tended to shrink. This implies that the Solow model’s prediction of transitional dynamics—where an economy grows faster the further it is from its steady state—holds, but only once we account for the fact that different countries are converging to different steady states.

This leads to the next logical question: what are these key structural characteristics that determine a country’s growth potential? Empirical studies have identified several strong correlates of economic growth.

• Physical Capital Investment: As shown in Figure 1.16, there is a strong positive correlation between average growth rates and the rate of physical capital investment. Societies that invest a larger fraction of their output in new capital goods tend to grow more rapidly.
• Human Capital Accumulation: Figure 1.17 reveals a similarly strong positive correlation between growth and measures of human capital, such as average schooling rates. Investment in education and skills appears to be a crucial ingredient for economic success.

Beyond these measurable inputs, empirical analysis consistently finds that there are large differences in the efficiency with which societies use their physical and human capital. We refer to this residual factor as technology. Technology, in this broad sense, captures not only genuine differences in production techniques but also the overall efficiency of the economic environment. Development accounting exercises, which we will explore later, suggest that these cross-country differences in technology are a crucial component in explaining the vast disparities in income per capita.

These factors—physical capital, human capital, and technology—are often termed the “proximate” causes of growth. However, a satisfactory theory must go deeper to uncover the “fundamental” causes that explain why these proximate factors differ so profoundly across societies.