Incomes from wages and capital were flat or fell for two-thirds of households in 25 advanced economies between 2005 and 2014—an explosive increase from less than 2 percent in the previous decade.
While it’s broadly assumed that children will grow up to be better off than their parents, the reality is that a new generation of young people in advanced economies risks ending up poorer. In this episode of the McKinsey Podcast, McKinsey senior partner Richard Dobbs and McKinsey Global Institute (MGI) partner Anu Madgavkar talk with Peter Gumbel about the increase in the number of households that experienced flat or falling incomes in the past decade—and the implications for future growth and economic advancement.
Peter Gumbel: Hello, I’m Peter Gumbel, senior editor, based in Paris, at the McKinsey Global Institute. Today I’m delighted to be speaking with Richard Dobbs, a McKinsey senior partner in London, and Anu Madgavkar, an MGI partner based in Mumbai. Richard and Anu have been spearheading new research on income inequality for the McKinsey Global Institute, and they’re the coauthors of a new report called Poorer than their parents? Flat or falling incomes in advanced economies. One of its striking conclusions is that two-thirds of households in 25 advanced economies have been affected by this flat-or-falling-income phenomenon. They’re here today to explain what that is, why it’s happening, and what it means. Richard and Anu, thank you for being with me today.
Richard Dobbs: Hello.
Anu Madgavkar: Hi, happy to be here.
Peter Gumbel: Great. Let me start with you, Anu. What exactly do we mean by flat or falling incomes?
Anu Madgavkar: Our research look at groups of households organized by income segments. We’re then able to track what these typical groups have experienced in terms of income growth over the last decade and the period prior to that. When these typical households, in each income segment, see no advancement in terms of their income, we’re able to say that there is a general lack of economic progress.
By this metric, we see a very dramatic increase in the phenomenon of flat or falling incomes. The proportion of households that have been affected by this trend has virtually exploded, from less than 2 percent to as much as 65 percent to 70 percent of the population, in the past decade—from about 2005 to 2014.
Peter Gumbel: That’s a tremendous increase. Have others already looked at this, or is this research quite unique in terms of the bigger debate out there about income inequality?
Anu Madgavkar: There has been a lot of very influential and important work on income inequality. But much of it, including the work by Thomas Piketty and others, has really focused on the very top of the income distribution—perhaps the top 5 percent of households or the richest 1 percent, even.
We don’t focus just on that segment. We look across the income distribution at multiple segments to understand how inequality has manifested itself in terms of income trends and income growth for all segments, and not just the richest.
There is also another body of work around inequality that has looked much more at poverty, talked about inequality as inequality between countries, and documented the fact that addressing poverty has helped shrink inequality between countries. But we, again, bring a new aspect by focusing on inequality within countries and zooming in on the middle-income group of households, the median household, to understand how inequality is affecting this broad mass of households within the advanced economies. We’re able to then dissect this and present a picture of what’s changed and why it’s changed—and in a relatively new way, which makes this work distinctive.
Peter Gumbel: The countries that you’re focusing on—the 25 countries—mainly include the United States and Western Europe. Given that the recession after the 2008 financial crisis hit so hard, how much of the problem you’ve found is related to the aftermath of the financial crisis and the very slow recovery that we’ve had since then?
Anu Madgavkar: We do find that the global crisis and its slow recovery since 2007 and 2008 had an important role to play. If you think about the period from 1993 to 2005, we found that growth in aggregate demand, in GDP and the employment that came with that, would have boosted disposable incomes for the median household by about 18 percentage points over the period.
But if you look at the period from 2005 to 2014, that 18 percent has come down to just about 4 percent. So there was a big shrinkage in disposable income because of the slow recovery in the aftermath of the crisis. The really interesting point is that notwithstanding the crisis and the recession and the slow recovery, you still did have a positive contribution to disposable income because of overall growth.
Then you also had important long-term structural factors which are demographic in nature, as well as the changing pattern of labor markets that have played in favor of some kinds of workers but not in favor of other kinds of workers. These longer-term structural trends have actually taken away even that 4 percent–positive contribution from aggregate demand over this period.
Peter Gumbel: Turning to you, Richard, obviously these numbers are very startling. But there’s also a bigger political context out there. Could you talk us through what you see as the implications? What are the “so whats” of these findings?
Richard Dobbs: The findings suggest we have a problem in three areas. First of all, on an individual basis, everyone wants their children to be progressing and to be better off than they were. It’s a fairly common, around-the-world feeling.
In the world we are painting here, there’s a significant chance that we’re not going to meet that aspiration. For many individuals, this is going to be deeply disappointing. Second, it will result in lower economic growth. The people who are doing less well typically have a higher propensity to spend [a larger proportion of their incomes]. The fact that they are doing less well means that we’re going to be facing lower economic growth.
Third, this is going to result in rising social tensions. As part of our research, we have surveyed 6,000 people, and we’ve understood their attitudes. The chunks of people who both are not advancing and perceive they’re not advancing—and think that’s going to continue—are much less likely to be supportive of many of the things that have helped drive economic growth over the past few years: free trade, technology, migration. This group of people who are not advancing tends to be very negative about all of those factors. As we see this rising group of people who are not advancing, we’re going to see increasing social tensions. That will come home and cause quite a lot of problems if we don’t address it.
It matters at the individual level—all of us wanting to progress, to have our children progress. It matters at a societal level in terms of the level of growth, and it also matters in terms of social instability.
Peter Gumbel: Is this something that governments have focused on? Or are we coming up with something which is new and adding to the debate, so our message is essentially that it’s important for the government to look at this whole flat-or-falling phenomenon?
Richard Dobbs: Historically, this has not been an issue, and it has not been a high priority for governments. If we look at the numbers prior to 2005, only about 2 percent of people in the developed world were not advancing. It really wasn’t as big an issue for governments in comparison with unemployment and growth in general.
It’s increasingly an area that governments are aware of, and they’re hearing it through their conversations with citizens. But it is an area which hasn’t yet been quantified. Our research was the first big attempt to try and quantify the problem.
Anu Madgavkar: As Richard said, this is not an indicator that was paramount or prime for governments, historically. Therefore, we don’t have a very clean or well-harmonized data set on this idea of not advancing. The way we dealt with it was to use three different methods of trying to gauge the extent of this phenomenon and then to triangulate and validate our results using these different methods.
We did look at country-level data from statistical agencies for six countries that we focused on. There we looked at income distributions of households into deciles and quintiles and compared those trends. That was one method. The second thing we did was a deep dive into worker-level or individual-level microdata spanning about 150,000 households in three countries—France, Italy, and the US.
We were able to understand income trends for different demographic cuts of workers in terms of age, education, gender, and so on. That, again, helped us to build a very consistent picture of where the issue of flat or falling incomes is particularly concentrated and also to validate the extent of the problem.
Then, finally, we did a survey of citizens in three countries. We surveyed about 6,000 people. The intent of this was to gauge people’s experiences or their feelings about this whole phenomenon of not advancing—how they compared their own current financial position with that of their parents and what they thought about their children’s financial position going forward.
Peter Gumbel: I’m going to ask you both what surprised you most about the findings. I don’t know whether you’ll have the same answer.
Anu Madgavkar: The first thing that surprised us was actually the sheer extent of the problem—the fact that a majority of households, particularly at the market-income level, were in this bucket of flat or falling incomes. So the labor markets were not really working for the vast majority, and the extent to which that was consistent across countries was actually interesting and surprising.
Then I think the other surprising factor was that the policies and actions and local interventions by governments, given the labor-market institutions in each country, did vary quite a lot. Therefore, the outcomes, in terms of net disposable income after taxes and transfers, showed quite a great deal of variation between countries. That, to us, was again quite interesting because it did suggest that tailored strategies and interventions can make a difference.
Richard Dobbs: I found it staggering that we have seen such an increase. To go from a world where about 2 percent of the people in the developed world were facing this problem to a world where somewhere between 25 percent and 50 percent of people in the developed world are no longer advancing is such a substantial step up.
Peter Gumbel: What is it about the world’s demographics and about the labor market that is changing and that is having such an impact here?
Richard Dobbs: I’d like to draw out a couple of drivers here that are important. The reason these are important is not just that they worked in the past. We think that as we look at this problem going forward, unless we get the step change in terms of policy, these are going to be bigger drivers—the three factors that are driving the labor market.
First of all, the rise of technology replacing jobs. We’ve seen this today. Remarkably, if you go back 500 years, there were monks who used to write out Bibles, whose jobs disappeared when the printing press came along. But what we’re seeing now is a different speed and scale at which these jobs have been displaced by technology. We’re increasingly seeing more and more sophisticated jobs being displaced.
The second factor is the rise of jobs moving overseas through trade and offshoring and those types of activities. Again, we’ve seen some of this through the offshoring of call centers, but, increasingly, it’s becoming more and more sophisticated.
Then the third factor, and this is a more emotional one because people can see it, is immigration. We have seen a step up in the number of people living outside their countries. Over 200 million people no longer live in their home countries. That has, for some jobs, resulted in greater wage pressure on a cohort of people who are growing up. The immigration numbers are not as big as the other two—technology and trade. But immigration is more apparent to people. That’s why a lot of the anger people feel as a result of not advancing tends to be targeted at immigration.
Peter Gumbel: All these are long-run factors that have been gathering pace. Does this mean that what we’re seeing with incomes is going to continue? Anu, you did some modeling to try and figure out what this could look like in the next ten years.
Anu Madgavkar: It is of course hard to predict what will happen in the future, so we don’t try and do that. But it is quite important to think about the sensitivities of the people who are not advancing to some of these underlying drivers or forces.
We really thought about a couple of drivers. We do think that productivity growth and growth in aggregate demand are going to be important variables that influence what share of people will not be advancing in the decade going forward. If we assume, for example, that the low productivity growth of the 2002–12 period persists for most of these advanced economies into the next decade, we find that an even higher share—70 percent to 80 percent of households—will not advance. So it is really important for us to think about addressing productivity growth as one important lever.
On the other hand, say that productivity growth improves and goes back to the historical high period that the world actually saw over the last 30 years, for example. We find that even then, 10 percent to 20 percent of households across these economies would not advance. That’s because of some of the forces of demographics and some of these technology- and trade-related labor-market shifts, which are not easy to reverse.
Then, finally, we do think that one important variable is the so-called coupling of wages with productivity. There’s been almost a systematic decoupling of wages relative to productivity growth. And therefore we also look at a scenario in which there is much more automation in the workplace and as a result wages don’t go up, although productivity does go up. In that scenario, we would find that 30 percent to 40 percent of households would not advance.
The bottom line on the sensitivities is that there is going to be a sizable proportion of households that do face this issue, even in the most optimistic sort of high-growth, high-productivity type of scenario.
Peter Gumbel: Just to clarify, though, we’re talking here about advanced economies in Europe, the US, and so on. We’re not talking about the developing world, emerging markets—or are we? They’re also facing the same issues of demographics and labor-market disruption.
Anu Madgavkar: That is correct—our research focuses on the advanced economies and not the developing world. We didn’t include it in this framing of the inequality question for a couple of reasons.
The first is that much of the developing world is still grappling with poverty. That’s the central issue. So how do you really get people from a very, very low level of productivity and a subsistence level of consumption up the curve and beat poverty? For much of Africa and parts of South Asia, that’s still an issue for large proportions of households. In terms of not advancing—historically, we haven’t really seen that to be a problem in the developing world.
We looked at data across India, Russia, and Brazil. At least over the 2003–13 period, all segments of households grew on the back of both economic growth, as well as some pro-poor, pro-middle-class, and poverty-alleviation strategies that these countries put into place. That said, if developing economies do slip into recession and slow-growth conditions, this could become an issue there as well. But, historically, we haven’t seen that.
Peter Gumbel: Richard, I’d like to turn to policy issues with you. Is this really a problem that could go away, and what would governments need to do to make it go away?
Richard Dobbs: Governments can play an important role here. The first thing is reminding ourselves that this is uncharted territory. Part of the answer is to take a mind-set of “let’s do a bunch of things, let’s learn from them, and let’s do more.” The thing, though, that’s most important is to do some measurement around this. If the problem is going to be managed, we’d better do a better job of measuring it.
Second, there are a number of interesting experiments around the world going on—a number of things that may have an impact. They may also have negative second-order effects that make them not make sense. But I think we need to be much better at looking at these experiments and understanding them. Let’s measure the impact. Let’s understand the negative effects, if there are any, or the positive effects. These are initiatives around getting growth going through investment, around how you build capabilities through better training and better education.
It’s also about how you get the elderly working, in many parts of the world. The elderly segment is one of the segments that do very badly. How do you get more elderly people into work, including part-time jobs? Then, there’s also the role that things like a higher minimum wage can play.
Government’s ability to do transfer payments and other types of redistribution is becoming harder and harder as government budgets are under more threat. However, many of the levers that I talked about aren’t actually classic transfers. The minimum wage is not a transfer; it doesn’t really cost the government. Even some of the education initiatives are relatively low cost. Some of the investment in the things, like infrastructure, that could be quite positive can be done relatively cheaply, given how cheap debt is. I don’t think it’s impossible for governments to do a greater amount of these things at the moment, even with their tighter fiscal positions and high debt levels.
Peter Gumbel: Anu, just to follow up on that, could you perhaps tell us a bit more about what you saw in terms of effective measures?
Anu Madgavkar: We’ve seen some examples of government measures being effective, and that can happen at two different levels. In the case of Sweden, some of the more labor market–oriented interventions are not about net transfers—this is not about taxes and transfers but about policies and working with employers as well as unions to craft labor-market policies that help preserve jobs, not destroy them.
For example, take a temporary cut in hours worked rather than pushing a lot of people into unemployment. There could be a real focus on job preservation, job creation, and some amount of wage protection. This has actually worked in Sweden. Of course, Sweden did have the financial resources, as well as, I think, the ability for the government to work very constructively with companies and unions to make this happen. What it resulted in is that, in our set of six economies, Sweden is the only one where market incomes were in positive territory for the majority of households. It really stood apart because of that.
Then there are measures that pretty much allow the market to play out wage trends the way they would ordinarily. But then the government steps in and does a set of transfers to boost disposable incomes. Here we found that the US has taken 80 percent of households that had flat or falling market incomes into positive territory at the disposable-income level. It did so simply by targeting different kinds of benefits that these households would actually earn.
Peter Gumbel: Richard, what would you say is the role of business in trying to mitigate some of these effects?
Richard Dobbs: Business can play a very important role in helping to address this; it’s not just governments. First, business can help drive productivity and drive economic growth. There is a benefit to everyone if we have high growth. The rising tide does lift all ships here, and business can play an important role delivering growth through higher productivity.
Second, business can play a role in helping people retrain when jobs disappear. As technology displaces jobs, businesses can run training programs to get people retrained. That requires a slight change of mind-set. There’s a company in the UK called British Gas that actually now runs apprenticeship programs for people up to the age of 50. This reflects the fact that people’s jobs may disappear midcareer and they need to retrain.
Third, businesses can also play a role in helping schools and pre-early–education establishments get people trained with the right capabilities. They can help schools set the agenda for education and encourage more students to be doing business-related subjects at school rather than stuff that’s not directly business related.
Peter Gumbel: Is the future, especially for middle- or lower-income households, pretty bleak?
Anu Madgavkar: There is a message which is more positive and which families and households should embrace. It’s really around resetting expectations about the nature of work itself. It is really important to understand that the traditional or the historical paradigm—“my kids will grow up and go through school and then get a job which is secure and has great benefits, and there’s a company out there that’s going to provide them with employment”—that paradigm is eroding.
The share of workers in part-time, temporary, on-demand work is increasing. That’s not necessarily a bad thing. There will be a lot of opportunities, new kinds of work. But the nature of work might be more oriented toward entrepreneurship and self-employment and looking for opportunities which allow you to do that.
Richard Dobbs: It is a worrying world we’re talking about here. We’re talking about increasing social tensions from a cohort of people who are no longer advancing, no longer buying many of the things that have underpinned historical growth—and I think will be important for future growth—like free trade, like migration, like technology.
But there is good news in that when we look around the world, there are pockets of places where people have done various initiatives, and those initiatives have helped overcome aspects of the problem. They’ve found a way of getting growth to happen, and they’ve found ways of ensuring that growth in the economy gets shared with this cohort of people.
The challenge that I think we face is how we get more systematic at learning from those places and transferring what we learn. This hasn’t been a priority for policy makers, historically. It is increasingly becoming a much more important challenge for policy makers.
Peter Gumbel: Thank you, both Anu and Richard, for your very interesting comments.
Richard Dobbs: Thanks a lot.
Anu Madgavkar: Well, thank you.
Peter Gumbel: I’ve been talking with Anu Madgavkar, a Mumbai-based partner at the McKinsey Global Institute, and Richard Dobbs, a McKinsey senior partner in London. They are the coauthors of a new report on income inequality called Poorer than their parents? Flat or falling incomes in advanced economies. You can find it on McKinsey.com.