Chinese / English
Home >> Great Minds China Forum >>  3rd Great Minds China Forum >>  Research Frontiers of Macroeconomics: Theories, Methodologies and ToolsResearch Frontiers of Macroeconomics: Theories, Methodologies and Tools
Harald Uhlig:The Five Shanghai Themes


 University of Chicago, CentER, NBER and CEPR 

 First draft: May 25th, 2017

This revision: May 25, 2017


This paper is written for the occasion of the 3rd Great Minds China Forum in Shanghai, May 2017. It discusses five important macroeconomic themes.

1 Introduction

This paper is written for the occasion of the 3rd Great Minds China Forum in Shanghai, May 2017. It discusses five macroeconomic themes and their resulting key questions. The paper is meant to accompany my lecture at that forum, and is essentially a version of the slides. As a result, the explanations and list of references are short. I hope this may be excused, for this occasion.

I was asked to deliver a lecture about the following. The world-wide financial crisis of 2008 has led to a re-orientation in economic research and in macroeconomic research in particular. The research frontier has shifted. What are the new themes, theories, methods and tools that have and that are emerging from that re-orientation? What are big issues going forward, that economic researchers ought to address? Which issues emerge in particular with regards to China?

This is a large task. There is much excellent research out there, and it will not be even remotely feasible to summarize it here. But there is an excellent recent source and portal to the wealth of recent macroeconomic research for anyone interested to pursue a deeper inquiry: the Taylor-Uhlig Handbook of Macroeonomics, Vol. 2 (2016). It contains chapters by many of the leading researchers on many of the central recent topics of macroeconomics, typically summarizing the field and providing many references, along with an assess-ment and outlook. It really is the perfect place to start for anyone seeking to be part of that exciting research frontier or for anyone interested in learning what modern macroeconomic research has to offer in addressing the pressing economic issues of our world.

Given the task and given this reference to the Handbook, let me then focus on five themes in particular. I shall them call “The Five Shanghai Themes”. I will describe each theme briefly. Each theme will then give rise to key questions, which I hope researchers will answer in the future. My aim is to raise questions, not to answer them. Encouraging future thinking, discussions and research on these themes and these questions is the main purpose of this lecture! I will devote most of my space to the first theme: mostly, because it is a theme that is now best understood. The other themes may be more exciting and recent, though. May the reader judge for her- or himself!


2 Technological Progress and Economic Growth


Figure 1 shows GDP per capita in the United States. What is remarkable is how it has been rising at a pretty steady pace over time: even deep recessions seem to be not much more than moderate bumps along that road.

Figure 2 shows the GDP per capita in China relative to the US. I used the Penn World Tables 9.0 or PWT for short as my data source. One can see that ratio of that consumption fell dramatically during the “Great Leap Forward” period, and then remained fairly stable, but has grown dramatically

since about 2000.

Figure 3 includes two other countries: Germany and Japan. They have been catching up to the US level much sooner.

Modern growth theory provides us with an excellent framework for un-derstanding these phenomena. The benchmark is the neoclassical growth model. It allows us to analyze the question, whether growth is due to capi-tal accumulation or technological progress. Let yt be output, γt total factor productivity or TFP, kt capital, nt labor, rt rental rate of capital, θ = 2/3 the labor share. From the production function


yt = γtkt1−θnθt


one can calculate γt, given data for yt, kt, nt and a value for θ. One can further calculate the marginal return to capital and reformulate,


yt r= (1  θ)kt

Data for yt, kt, nt is from the PWT 9.0, though I calculated TFP on my own, given the PWT 9.0 data.

Figure 4 shows the TFP growth across the four countries of focus. Fig-ure 5 shows this for China alone, relative to the US. One now sees a very substantial decline during the “Great Leap Forward” period: TFP in China only now has recovered to the previous relative position.

These TFP calculations now allow me to investigate the sources of growth. Define the TFP-adjusted capital-labor ratio as


1/θ kt

γt nt

Equation (1) relates it to the capital rental rate rt,

As a benchmark, it may be good to consider rt to be roughly constant and the same across countries. Figure 6 shows the result for our four countries. As one can see, that adjusted ratio is reasonably constant for the US and perhaps Germany, while Japan and later China seem to have grown out of a rather capital-scarce phase to a rather capital-intensive phase, possibly accumulating capital faster now than can be justified by TFP growth alone. This might be cause for concern. Capital which is accumulated faster than can be justified on productivity grounds may deliver growth, but will do so at costs that are too high. A slightly slower pace may assure higher living standards and higher happiness or welfare instead.


Modern growth theory has pointed out, that technological progress is the

result of innovations. They in turn require a highly educated and imaginative population of innovators and entrepreneurs. Human capital and education is central to that process. Figure 7 shows the human capital, as provided by the PWT 9.0, largely based on measures of education. The human capital in China is growing, but it is still far behind the other countries. In order to catch up with the living standards in these countries, it is paramount that educational levels in China catch up the educational levels in these countries.

China has made remarkable progress in this regard, as can be seen from the recent rankings of their universities, see tables tab:shanghai and 2. This strikes me as highly important, and a process that deserves continued priority. The German in me cannot help but remark on the distant places that the once fabled German universities now occupy in these rankings. Germany surely still produces leading-edge technologies, among them some of the very

best-engineered cars one can possibly buy. The technological leading firms of Germany naturally have been a source of envy and inspiration for other countries. But if the past is any guide, and if one extrapolates forward the status of German universities vis-a-vis Chinese universities, as indicated by these rankings, then the day may not be far off, when Chinese technological advances will be the role model for Germany rather than the other way around.

Let me conclude this section by framing the resulting big questions, and providing some partial answers, based on the analysis so far. Much more needs to be done here.


1. What is the source of long-run growth? Answer: technology and human capital. Education is key.


2. Is China accumulating too much capital? Tentative answer: yes, re-cently.


3. How can China catch up to the frontier? Tentative answer: further improvements in education and further improvements in the freedom of entrepreneurship and innovation.

House Prices and Financial Markets: Booms and Busts


The connection between housing markets and financial markets is crucial. It is important to understand the boom-bust episodes in both and their inter-connection. This is the central theme in the Handbook chapter by Guerrieri-Uhlig (2016).

Figure 8 shows the house price developments in the US, as calculated by the S&P-Case-Shiller home price indices. Figure 9 shows subprime mortgage originations in the US.

House prices go through booms and busts in many countries, not just the US. Figure 10 shows house price developments in Shanghai, for exam-ple. Clearly, house prices are booming. Will there be a house price bust in Shanghai and elsewhere in China? And what is the connection to the recently popular wealth management products: a development quite similar in flavor to the development of the mortgage backed securities market in the

US? Could a bust in house prices and financial markets be looming in China? China is now a large part of the world economy: what will be the world-wide repercussions of a financial and house-price crisis in China?

The empirical connection between financial booms and busts and housing market booms and busts has been emphasized by Schularick and Taylor in various papers, cited in my Handbook chapter with Guerrieri. Figure 11 provides some graphical insight into their findings, with a further calculation courtesy of Luigi Bocola.

Let me conclude with the resulting big questions. But from here on for-ward, I shall only raise these questions, and leave the answers to subsequent research.


1. Is China/Asia headed for a house price crash?


2. Is China headed for a financial crisis?


3. What will the world-wide repercussions be?

Figure 11: Schularick-Taylor: crisis data for 19 countries. T: Schularick-Taylor crisis dates. All crisis events (blue line) vs 5 events with the highest house price price drop. Calculations: Luigi Bocola. Source: Guerrieri-Uhlig (2016).

4 Aging and the Savings Glut


Many have bemoaned the low returns on savings and the savings glut. Many believe this is the result from an increased demand for savings vehicles by a world population that is living to increasingly higher ages, and where the young need to save for their retirement years.

Indeed, asset returns go through rather dramatic swings, and they are forecastable to a degree. In an influential paper, Lettau and Ludvigson have constructed a variable called “cay”, encapsulating the transitory component in the relationship between income, consumption and wealth. This transitory component is then important for understanding the movements in wealth. With some approximation, that relationship can be stated as


at = const. + ct + 2.5(ct − yt) + trans.compt


where at log-assets, ct log-cons., yt log income: the recent, precise, estimated coefficients are a bit different, see data cay.html. A high transitory component then predicts subsequent de-clines in wealth or low returns in wealth. Figure 12 shows that transitory component. It was high in the late seventies, before a decade of declining stock markets, for example. It is now high again. Is another rather short-term correction overdue, or will we simply see low returns for a long time? Or will instead income and consumption catch up and justify these high levels of wealth? Lettau and Ludvigson have shown that the latter played only a minor role in the past. Low returns on savings then seem to be ahead of us, instead, I suppose. What is the reason? Aging, perhaps?

How can one understand the impact of an aging population on asset returns? A simple model of aging and the savings glut may be useful. I only sketch it here, but for readers familiar with overlapping generations (OLG) models, it should not be hard to fill in the missing details. It is a slightly modified version of the standard neoclassical growth model in a two-period

Figure 12: The Lettau-Ludvigson transitory component of wealth. Is it transitory or is it permanent? And what does it imply for future returns?

OLG model. Consider a two-period OLG with “years” within each period. Agents live 1 − λ “years” when “young” and λ “years” when “old”. Agents have log-preferences, no discounting. Agents supply a total of one unit labor nt = 1, when young, for a wage wt. They save per purchasing capital. Savings earns a return Rt between periods, but not between the “years” within each period. There is linear capital accumulation with depreciation δ. Assume Cobb-Douglas production, no TFP growth:

In particular, one can now see, how r is lower, when the population is older, i.e., when λ is larger. This captures an important relationship: with more retirement savings, returns to savings decrease. This relationship may be at the heart of the recent transitory rise in assets, relative to consumption and income, as shown in figure 12. I believe this model and these empirical findings to be an excellent starting point to dig deeper into these connections.

These then are the big questions, which await answers from subsequent research.


1. What are the macroeconomic implications of a world-wide aging soci-ety?


2. How long can and will we work, as the length of life expands?


3. What is the optimal way to finance pensions?


4. What is the optimal intergenerational risk sharing arrangement? For this issue, allow me to point to my paper with Lans Bovenberg (2008), listed in the references.


5. Have we shifted to a future with permanently lower returns and per-manently higher asset-to-consumption ratios? Or is this temporary? And what are the macroeconomic implications?



5 The Macroeconomics of Health Care


Closely connected to the issue of aging as well as the issue of technological progress is the issue of medical costs and the costs of health care. As we get richer and as we get older, health becomes of increasing concern, and it is only natural that people devote an increasing share of their resources to that end. Figure 13 shows the rising shares in the US. Table 3 shows a comparison across OECD countries, and the rising shares there. I do not have the numbers for China at hand, but I suspect a similar phenomenon there.

What explains these developments? Where do these shares go? Where should they go? I have investigated this question together with my coauthors in Koijen-Philipson-Uhlig, Econometrica (2016), and Koijen-Uhlig (2017). Figure 14 shows our prediction for the long-run share in health spending, while figure 15 shows the prediction for the share of GDP spent on medi-cal R&D. What is optimal for the latter? That depends on patience and curvature of preferences, as figure 16 shows. While I believe that these are interesting answers, much more research is needed to investigate them with sufficient depth, given the importance of these issues.


Let me conclude this section with these big questions:


1. Where will health-expenditure and medical R&D expenditure go, in the long run? US vs China?


2. Where should they go?


3. What will this imply for aging, savings, growth and macroeconomics?

6 The Macroeconomics of Global Warming


The environment we live in is subject to potentially large changes in the next century or so. The theme of global warming is subject to intense dis-cussions world-wide, and a thorny poitical issue. For some well-known facts, figure 17 shows the warming ocean temperatures, while 18 shows the carbon dioxide emissions, which most scientists believe to play a crucial role in the phenomenon of global warming. What are the economic repercussions?

The macroeconomics of global warming has become the focus of some re-cent and fascinating research, surveyed and extended in the Handbook chap-ter by Hassler-Krusell-Smith (2016). I cannot possibly summarize it here and thus simply recommend it most highly for further study. The political dimensions and the distributional consequences deserve much attention too. In particular, a case can be made that the currently leading industrial coun-

tries have already done their share of pollution to get where they are: so how much pollution should be allowed in the countries such as China, to catch up? I do not have a good answer, but it is clearly an important question, which will determine much of our future fate and standards of living.

In sum, the big questions here are


1. Where does Global Warming go and what is the macroeconomic im-pact?


2. What is the optimal amount of pollution?


3. How should international property rights be assigned?


4. How much “catch-up” pollution should be granted to countries like China?

7 Conclusions


I discussed “The Five Shanghai Themes”:


1. Technological Progress and Economic Growth


2. House Prices and Financial Markets: Booms and Busts


3. Aging and the Savings Glut


4. The Macroeconomics of Health Care


5. The Macroeconomics of Global Warming


I raised several “big questions” for each. I do hope that they become the focus of future thinking, discussions and first-rate research. Encouraging such future thinking, discussions and research has been the purpose of this lecture. If you are a researcher embarking on this journey, I highly recommend to take along the Taylor-Handbook as a guide and starting point




[1] Bovenberg, Lans and Harald Uhlig (2008), “Pension Systems and the Allocation of Macroeconomic Risk”, in NBER International Seminar on Macroeconomics 2006, L. Reichlin and K. West, eds., NBER, University of Chicago Press, 241-323.


[2] Guerrieri, Veronica and Harald Uhlig (2016), “Housing and Credit Mar-kets: Booms and Busts,” chapter 17 in Taylor, John and Harald Uhlig, eds., THE HANDBOOK OF MACROECONOMICS, VOL. 2, Elsevier-North Holland, 1427-1496.


[3] J. Hassler, P. Krusell, A.A. Smith Jr. (2016), “Chapter 24: Environ-mental Macroeconomics,” in HANDBOOK OF MACROECONOMICS VOL. 2, John Taylor and Harald Uhlig, eds., Elsevier-North Holland, 1893-2008.


[4] Koijen, Ralph S., Tomas J. Philipson and Harald Uhlig (2016), “Fi-nancial Health Economics, Econometrica, vol. 84, no. 1 (January), pp. 195-242.


[5] Koijen, Ralph S. and Harald Uhlig (2017), “Government Risk and the Macroeconomics of Health,” draft, University of Chicago.


[6] Lettau, Martin and Sydney Ludvigson (2004), “Understanding Trend and Cycle in Asset Values: Reevaluating the Wealth Effect on Con-sumption”, American Economic Review vol. 94, no. 1 (March), 276-299.


[7] Taylor, John and Harald Uhlig (2016), eds., THE HANDBOOK OF MACROE-CONOMICS, VOL. 2, Elsevier-North Holland.

◆please indicate the source if authorized: National Economics Foundation

◆photo:National Economics Foundation