The Holy Month of Sports Begins
The Holy Month of Sports Begins
One of the most fitting aspects of the phrase “transit deserts” – areas drastically underserved by public transit – is the mirage-like effect you get when, straining your eyes toward the suburban horizon, you trick yourself into thinking that the headlights of an approaching car is actually the bus you desperately want to arrive.
Today, however, as transit apps get more accurate, the era of transit mirages may be nearing an end. You can now simply look at your phone, or in some cases at a television screen at your bus stop, to see when the next bus will arrive.
In many cases, transit apps might also put an end to another desert-like quality of public transit: waiting in uncomfortably cold or hot temperatures. Equipped with accurate information about when the next bus is likely to approach, passengers might have the option of waiting for their bus indoors a short walk away from their bus stop.
This might have significant implications for the public or private sector. The public sector could, perhaps, create transit oases: indoor waiting areas that passengers could enter by using their transit passes to pre-pay their bus fare, or outdoor parkette spaces where people could wait with shade, nice surroundings, picnic tables, etc.* Or the private sector could provide a similar service, maybe combining transit waiting areas with laundromats, convenience stores, cafes, etc.
*[Alternatively, they could have car-sharing cars parked near bus stops, which would double as sheltered bus-waiting areas. That way the cars could also be available for use as “first-mile/last-mile” vehicles, in areas where walking to and from bus stops can be difficult]
Because some transit apps have even begun providing information about how crowded each approaching bus is, having a more comfortable waiting area of this sort could also give passengers the option of waiting a bit longer than they need to, in order to wait for a less crowded bus to arrive. This could be very useful: public buses tend to bunch fairly close together, and the front bus in each grouping tends to be much more crowded than the buses further behind it. With transit apps, people could simply wait to get on a less crowded bus, without taking the risk that the bus further behind will be just as crowded or will not arrive any time soon.
Similarly, transit apps can give passengers real-time information about express buses, which could help allow passengers to let a normal bus go by in order to wait for an express bus approaching after it.
Another common challenge in transit deserts, where waiting times are often long and uncertain, and where bus journeys are often long as well, and often require multiple transfers, and often involve seniors or babies, is the lack of bathrooms at bus stops, and the lack of accessible public bathrooms in general. Transit oases could be a godsend, then, for those who cannot hold their water like a camel.
A somewhat similar, even bolder experiment should perhaps be tried in the NBA playoffs.
Instead of teams saving their best players’ heavy-minute performances for games late in a series – Lebron James, for example, played all 48 minutes in Game 7 of the semi-finals last year; it was the 339th time since 1984 that a player had played an entire game — why not instead have them play big minutes early in the series?
It is true, teams employing this strategy would risk their star players running out of gas late in the series (not unlike how the Tampa Bay Rays risk blowing leads late in games because they have already used up one of their best relievers in the first inning). Yet they would also increase the odds of getting to play in a long series in the first place. Even better, they would increase the odds of their team winning a series quickly, which would actually allow their stars to play far fewer minutes overall than they otherwise would.
An extreme illustration of this: 45 minutes per game in a 4-game sweep = 180 minutes total; 40 minutes per game in 7 games: 280 minutes total (plus two or three more airplanes). As is often the case in life, if you work hard early on you might save yourself work overall.
It’s a simple set of questions: Is it better to concentrate, rather than spread out, your best players’ minutes on the court during a series? The obvious risk to concentrating minutes is that tired players could become less effective or more prone to injury. Nevertheless if it is better to concentrate them, why not do so as early as possible, rather than wait until late in the series to do so?
Usually, the goal of a basketball game is to outscore your opponent. There are certain cases, however, where a second goal is also at play: buying time for your best players to get a much-needed rest on the bench during a playoff game.
Star players tend to play many more minutes per game during the playoffs than they do in the regular season. The games are more important, there are no games played on back-to-back nights, and there are fewer flights between games. In this year’s playoffs, the leaders in minutes per game, Paul George, Damian Lillard, and Nikola Jokic, have been playing 40 minutes per game. (Lillard and Jokic played 58 and 56 minutes in a quadruple overtime game). Lebron James’ career playoff minutes per game is 42. In Lebron’s 21 career playoff elimination games, he averages 45 minutes per game.
It would be ideal if, during those few minutes of game-time when these superstars are resting, their backup players would enter the game and consistently outscore their opponents. But that is not the only way a bench lineup might be effective. They could, perhaps, also be effective by simply slowing down the pace of the game, shooting late in the shot clock, grabbing offensive rebounds to reset the shot clock, or drawing fouls that stop the clock altogether.
Consider two hypothetical bench lineups. One is an above-average bench lineup which usually outscores its opponent, but occasionally gets outscored by its opponents by a lot. The other is a merely average bench lineup which outscores its opponent only fifty percent of the time, but, because it excels at slowing down the game, almost never gets outscored by its opponent by a lot.
On teams which have dominant starting lineups and superstars, the merely average lineup might actually be preferable to the above-average lineup, as it would almost never cough up a big lead created by its dominant starters. There is, moreover, the issue of $$: above-average players are in high demand, so they tend to cost a lot of money. Merely average players who excel only at slowing down the game are not in demand, and so might be purchased on the cheap.
Of course, this is not to say that teams should want their bench lineups to be merely average. But it does mean some teams might benefit from targeting certain skills and strategies for their benches: scoring late in the shot clock, offensive rebounding, drawing fouls, defending in such a way as to make opponents shoot later in the shot clock, etc.
This idea raises some interesting thought-experiment questions. For example:
-how much could you slow the pace of the game without sacrificing too much effectiveness? (The average possession length in the NBA this season ranged from OKC’s 12.9 seconds per possession to Cleveland’s 15.5 seconds per possession. In the playoffs, it has ranged from OKC’s 12.7 seconds to Denver’s 15.5).
-How high could you theoretically get your offensive rebounding percentage? (Denver had the highest offensive rebounding percentage this season, 30.8%. The highest in the playoffs has been Philadelphia, 31.2%).
-How much could you increase your offensive rebounding percentage without giving up too many transition opportunities? (The Clippers and Nuggets tried to grab offensive rebounds on an estimated 70% of their shots in these playoffs; the Raptors only on 46% of shots).
-When and how often is it worthwhile to foul in transition to stop the clock? How often would fighting for offensive rebounds lead to a stopped-clock situation as a result of fouling, being fouled, a jump ball, or an out-of-bounds? (Out-of-bounds stoppages on contested rebounds might be ideal, as they are most likely to lead the refs to waste a lot of time looking at the replay to see which team touched the ball last in a crowd under the net).
Obviously, many of the skills that would allow a lineup to excel at slowing the pace of play would overlap with those that make basketball lineups good in general. But the overlap is far from a perfect one. By focusing not just on creating the best possible bench lineup, but also on creating the bench lineup most capable of wasting time, you might be able to free up roster spending in order to land the big fish superstars you need. All your bench has to do then is waste a handful or two of minutes, to let those superstars catch their breath.
How’s the bathroom business doing, Jim? Well, the profitability ratio stinks, but there’s a lot of liquidity.
Nothing is more common as having to go to the bathroom, yet there are few examples of for-profit bathrooms in our otherwise profit-oriented cities. Even though bathroom access is often bartered indirectly – “customers only” comes to mind – such access is almost never charged for outright.
Outside of sports arenas and shopping malls, bathrooms also tend to be quite small. This is understandable, given the high cost of commercial real estate. Nevertheless it runs somewhat counter to the natural economy of scale that is bathroom installation. Once you have already ripped up the floor and sorted out the plumbing, the cost of each additional toilet or sink may be comparatively low.
There are, perhaps, reasons to think that for-profit bathrooms will become more common soon:
Transaction costs are the scourge of any high-liquidity, low-margin market, and the bathroom business is no exception. If you are planning to charge a dollar (say) per person, then obviously you are not going to turn a profit if you have to hire somebody to man the till or pay a transaction fee to a payment systems company. Only with an automated, extremely low-cost payment system might for-profit bathrooms actually work.
By telling you where nearby bathrooms are, apps could allow bathrooms to be findable even if they are in relatively out-of-the-way locations where real estate is not so pricey: in basements, second floors, side streets, etc. Apps could also let people reserve the use of a bathroom immediately ahead of time, or at least check to see if there is a bathroom available.
As transit systems introduce features like, for example, two-hour transfer passes, their passengers will be able to get off a bus or train to use bathroom without being charged a double fare. Transit apps can also help in this regard: passengers can now check to see when the next bus is coming, and so can estimate whether or not they will have time to use the bathroom without missing it. This will also allow passengers to wait indoors for the bus when the weather is bad; and in some cases, perhaps, do so near a for-profit bathroom.
It might also be possible to automate much, or at least part of, the process of cleaning a bathroom, which could go a long way towards making a public bathroom profitable. Expensive bathroom cleaning robots might also make public bathrooms more of an economy of scale; the toilet-to-robot ratio would increase the more toilets a bathroom has.
Of course, a profitable bathroom does not need to be a for-profit bathroom. It could instead be that bathrooms will become more common, cleaner, or less costly for the public sector to provide.
In recent years there have been two excellent new technologies, car apps and transit apps, which have nevertheless been unable to successfully solve traffic problems. Transit apps, which tell you when buses or trains are coming and, in some cases, tell you how crowded each bus is*, are useful but are still no antidote to challenges such as getting to and from bus stops or waiting for buses in bad weather. Car apps, which can summon lifts or carpools or make it easy to rent a car, tend to do little or nothing to alleviate traffic jams, and can also be relatively costly or inconvenient, especially in bad weather when there are no vehicles nearby, or when demand for lifts outstrips supply.
*This is very useful because buses tend to bunch fairly close together, and the front bus in each grouping tends to be much more crowded than the buses behind them. With this information, people can simply wait a tiny bit longer to get on a less crowded bus, without taking the risk that the bus further behind will be just as crowded or will not arrive any time soon.
What might be needed, therefore, is a way to use these new technologies to get to and from transit, especially in bad weather, and make it easier to wait for buses in bad weather, and do so without adding to traffic jams that block more efficient modes of transportation such as buses or streetcars or (in good weather) bikes.
Here, then, is a possible solution: have car-sharing cars double as sheltered bus stops in bad weather.
Such a system could have a number of advantages:
It is not just car-sharing technologies that could make this idea viable, but also transit apps.
By having the cars equipped with these apps, people will able to use them as sheltered bus stops without needing to have a clear view of the horizon to see if a bus is approaching. Even though these cars would ideally be located immediately next to bus stops, the ability to know in advance when a bus is coming means that if necessary they could be located a bit further away from the bus stops, on an adjacent side street, without the risk of people missing their bus.
That might not even be needed though, since the car-sharing bus stops might not need to take up much more space than the current public bicycle-sharing systems often do, particularly if one-seaters or pedal-cars are used. (In Toronto, there are already 360 public bicycle-sharing stations, even though few people use them in the winter or during heat waves). People would then be able to use their transit passes to unlock the car’s door by prepaying their bus fare, so that the cars would not be misused too much.
Most of us have heard the crude joke that ends in the punchline “We’ve already established the principle that you are a whore. Now we’re just haggling over the price”. Well, in hockey, I would like to argue for the principle that the best time to pull your goalie is when your opponents are physically exhausted from being stuck on the ice during an especially long shift. Though obviously it is difficult to know how much earlier than usual a team would be wise to pull its goalie in order to take advantage of facing exhausted opponents, still I believe they should consider doing so at least somewhat earlier than usual. There are three reasons why:
This sounds like circular reasoning, but really it might just be common sense: you want to pounce on your enemy when your enemy is weakest. When players are exhausted, it is harder for them to get the puck and score a cross-ice empty netter. (And, if they attempt to do so and miss, it will lead to an icing that will allow your team to bring on a fresh line against their exhausted one). Similarly, when they are exhausted it will be easier for your fresh-legged extra attacker to help your team get a high-quality scoring attempt.
Coaches generally try to pull their goalies when their teams are already in the offensive zone, but they often fail to do so simply because holding onto the zone is so difficult in hockey that the opposing team is frequently able to clear the puck out before the extra attacker has time to get there himself. As a result, teams with their goalie pulled often waste precious time or give up an empty net goal trying to regain entry and get solidly set up within the offensive zone again. Against an exhausted line, in contrast, it is much easier to hold on to possession, so your extra attacker will more likely have time to join the play while your team is still set up in the offensive zone.
Exhausted players are usually bailed out by their goalie, who freezes the puck to let them get a line change or call a timeout, or else they are bailed out by so-called puck luck: a favourable bounce of one sort or another, which allows the exhausted players to clear the zone and start a line change. But if you bring on your fresh extra attacker, the exhausted opponent will become much less likely to be bailed out by their goalie or by puck luck. Their goalie will have a harder time freezing the puck as he is more likely to be screened during every shot and outmanned in every scramble in front of the net. Puck luck too is less likely to be helpful to the exhausted team because, of course, puck luck is not mostly about actual luck, rather it is about open space – which there will be less of – and about effort and skill – which exhausted players have much less of.
Thus, you may trigger a spiral of death: exhausted players will be much less able to get a line change in, and so will become even more tired, and so will become even less able to get a line change in, and on and on until finally the spiral reaches a point of conclusion: ideally, the game-tying goal.
So: Do you, reader, believe this principle I am trying to establish? Good. Then let the haggling begin.
– (This is an unpleasant article, sorry) –
On the anniversary of the Yonge street van attack which killed ten people a year ago, Canadian media has been debating whether or not city streets should put in place more barriers to help keep pedestrians safe in the event of any future van attacks. This seems to be missing the point: Toronto had not one but two terrorist attacks in the past year, the Yonge van attack and the Danforth gun attack. Luckily – in the relative sense only – no more than thirteen people were killed. But one does not need to be a neurotic to consider what might have occurred if, instead of Yonge and Danforth, they had attacked Yonge-and-Danforth: Canada’s most crowded subway station.
If we have learned anything about terror attacks in the 21stcentury, it is that they are simple to imagine ahead of time (9-11, for instance, was neither the first quadruple airplane hijacking nor the first significant attempt to destroy the World Trade towers nor the first attempt to crash a hijacked airplane into a major Western landmark nor even the first major attack carried out by Al Qaeda that week), possible to prevent (hardening cockpit doors, for example), and can have secondary consequences nearly as damaging as, or plausibly even far more damaging than, the terrorist attack itself.
Now think back to Toronto. An attack at Yonge-Bloor station is frighteningly easy to imagine: a gunman, or gunmen, at rush hour, just as a train or trains are pulling into the station, videos of the result circulating online immediately afterwards. Aside from any damage such an attack might cause directly, it would potentially have even graver secondary impacts in the form of copycat attacks, car accidents, air pollution, political consequences, etc., not just in Toronto but also nationwide and beyond.
The point here is not to ask Canadian media to debate how to make areas like Yonge-Bloor safe from gunmen – that is not a discussion we would ideally need or want to carry out publicly. Rather, Canada’s intelligence agencies and governments should perhaps immediately address the problem. Waiting until a decade from now, when a relief subway line will hopefully be built and Yonge-Bloor will be expanded, may be taking a very big risk. Indeed, the station expansion may even increase crowding during construction for several years.
There could, maybe*, be relatively simple ways of reducing this threat. For example:
*I will admit: despite what I have written above, I know basically nothing about the risk of an attack of this kind, or how that risk should be addressed. [I will even cop to the fact that I may to some degree be using the fear of such an attack manipulatively, as a way to promote transit-friendly or pedestrian-friendly policies that I would want even if no such risk existed…]. I would, therefore, be very interested to learn what people who are actually knowledgeable think about this issue.
In any event, these two suggestions seem unlikely to occur; many would see them as (to quote Doug Ford) “a war on the car”. Yet maybe seeing them as a war on terror would be much more accurate.
One-seater electric automobiles may be on the rise, thanks to vehicle-sharing apps and range improvements in batteries. It is possible that these one-seaters will be a slothful and anti-social mode of transportation, like they are in Wall-E. Perhaps more likely, however, is that one-seaters will facilitate transportation that is actually more healthy and social than our existing four-wheel-dominated system. One-seaters may promote cycling, either via velomobiles (one-seater lightweight pedal-cars with an electric motor so that cyclists don’t have to break a sweat going uphill) or via the creation of more bike lanes in cities with relatively extreme climates (lanes which one-seaters would have mostly to themselves whenever the weather is not friendly to bicycles). One-seaters may also promote transit ridership, by making it easier to get to or from stations.
How then should cities prepare for one-seaters? While it would be nice if streets had three separate lanes for buses, bicycles, and one-seaters, this is not likely to be politically viable any time soon. Cities may need to be more creative, and more compromising, in seeking ways to promote one-seaters as an alternative for normal cars without unduly limiting transit or bicycle usage.
One possible way of doing this would be a sort-of seasonal three card monty, using a bus lane and an adjacent bike lane. When the weather is bad – say, during winter or a long heat wave – one-seaters would share the bus lane with buses and use the bike lane as a parking lane. This would allow a senior citizen to park close to his or her destination (or his or her bus stop), to avoid slipping on ice. It would also allow one-seaters to overtake buses at bus stops: if the bus pulls up to the curb at stops, a one-seater could pass the bus on its left. Similarly, it would give one-seaters the option of pulling over to let a bus overtake them.
When the winter or heat wave ends, the one-seater-parking lane would become a bike lane, usable by bicycles and one-seaters both. This would create less traffic for buses, at a time when the weather is nice enough to make walking to and waiting at bus stops a convenience. One-seaters and cyclists would however also be able to use the bus lane in order to overtake slower cyclists ahead of them, and to overtake buses at bus stops. Indeed, at times when buses do not run too frequently (nights, weekends), the bus lane would be left more or less open for one-seaters and the bike lane more or less open for bicycles.
It is true that bicycles would, comparatively, be the losers in this relationship. They would have to give up their bike lanes during the winter (or, in very hot climates, during the summer), and share them with one-seaters during the rest of the year. Nevertheless, given how few streets currently have bike lanes, cyclists might still benefit hugely from this relationship, particularly in climates in which bicycling tends to be a seasonal pursuit.
Let’s use a real-world example of where such a system could maybe be effective: Scarborough, Ontario. Scarborough is a suburb of Toronto, where the weather and distances people travel can be difficult for cyclists in the winter – and often in summer as well.
Unlike Toronto’s other suburbs, the shape of Scarborough’s coastline is slanted, giving the city’s main cycling path, the Gatineau hydro corridor (image above, map below), a diagonal shortcut route through the fairly strict grid pattern of the city’s streets. In a few years this corridor will be linked with the city’s newest half-subway line, the Eglinton Crosstown. The hydro/cycling corridor already connects, more or less, to a another subway line (at Kennedy Station), a hospital, a university, and, via the Don Valley, to downtown Toronto.
Another, narrower diagonal hydro corridor, meanwhile, which nearly links up perpendicularly with the Gatineau around Lawrence East SRT station (which may eventually become a SmartTrack GO Train station, even if the SRT rail line is closed and is not replaced with an LRT along the same route), could become a shortcut route in another direction.
If one-seater lanes were built in these diagonal hydro corridors, and if a three card monty bus/bike/one-seater system were implemented on a number of Scarborough’s main streets intersecting these corridors, then the suburb’s transit, cycling, and driving might be significantly improved. Ideally, it will one day become easy to ride a bicycle or a velomobile from U of T Scarborough campus all the way to U of T in the heart of downtown T.
2008 was as significant a year from a demographic perspective as it was from a financial one. In 2008 the world’s age dependency ratio — the number of people who are either younger than 15 or older than 65, relative to the number of people aged 15-65 — reached its lowest point. From a peak of approximately 77 in 1967, the ratio fell to a floor of 54 in 2008, a level it has remained at every year in the decade since. This low is not likely to be surpassed. The UN predicts that the ratio will rise gradually during the generation ahead, as more Baby Boomers turn 65 and birthrates keep falling worldwide.
The age dependency ratio is a useful, though obviously imperfect, measure of economic potential. The larger a country’s dependency ratio, the heavier the economic burden (to put it crudely) its working-age population may need to bear. The country with the highest such ratio in the world, Niger, with a ratio of 112, has a burden 1.12 times as heavy as those who bear it. The country with the lowest dependency, South Korea, with a ratio of 38, has a burden that is only about a third as heavy as those who carry it. The Gulf Arab kingdoms have even lower ratios than that (the UAE’s is just 18!), but only because they have so many temporary foreign workers.
It is not surprising that a lower dependency ratio tends to correlate somewhat with economic success. Not only is a country with fewer dependents more able to invest its time and money in increasing its productivity, but productive countries also tend to have low fertility rates, which keep dependency levels low in the short-term (though not in the long term, when low fertility rates lead to small working-age populations). As such, a low dependency ratio can be both a cause and an effect of economic growth. Even the oldest country in the world, Japan, only has a dependency ratio of 66.5, much lower than those of the young countries within Sub-Saharan Africa.
In recent history, the correlation between economic growth and age dependency can be seen most clearly in East Asia. China’s rapid economic growth has tracked its dependency ratio’s steep fall, while Japan’s stalled economic growth has tracked its own dependency ratio’s rise. China’s dependency ratio, which is today the lowest in the world apart from South Korea (not counting city-states or the Gulf Arab monarchies), was almost twice as high a generation ago, and only fell below the US’s in 1990. That same year, Japan’s ratio fell below Germany’s to become the world’s lowest other than Singapore or Hong Kong. A rapidly aging population has since made Japan’s become by far the highest in the developed world, however. Japan’s ratio has also risen higher than those of many developing nations in recent years, even than some of the world’s poorest nations, such as Haiti.
Outside Japan, East Asia now has the lowest dependency ratios of any region, by far. Not only China and South Korea but also Thailand, Taiwan, Singapore, Hong Kong, Vietnam, Malaysia, and even North Korea all have ratios between 38-44, the lowest in the world anywhere outside of the Persian Gulf. Indonesia’s too, at 48.5, is now lower than those of most countries in the world, while the Philippines, the major outlier in the region with a dependency ratio of 57.5, no longer has a high ratio by global standards either. This trend, however, is finally beginning to change. China’s ratio has begun to rise since 2010, prompting many to worry that the country “will become old before it becomes rich”. The dependency ratios of Vietnam, Thailand, and South Korea have also begun rising during the past several years. And Japan’s already high ratio will continue to rise quickly unless it finally decides to raise its extremely low immigration rate.
The years 2008-2010, in addition to being when the global dependency ratio and the Chinese dependency ratio both reached their lowest levels, was also when the EU’s dependency ratio rose higher than that of the US, for the first time since 1984. The EU’s dependency burden has continued to rise relative to the US in the decade since, a fact that has perhaps contributed, at least to a minor extent, to the US’s stronger economic performance during this period. Indeed, at the risk of attributing far more significance to the age dependency ratio than is justified, I will also point out the fact that countries in Central Europe have enjoyed a much lower ratio and a much stronger economic performance than has the EU as a whole. Similarly, Canada has had the lowest dependency ratio and one of the strongest economies among rich Western nations in recent years. Ratios in Canada and Central Europe were particularly low during the financial crisis:
Another intriguing case is Italy, which has a ratio that has been rising at fast pace since 2010, reaching the highest level in its modern history in 2017, at the same time as its economy has become perhaps the primary point of concern in European politics. A similar trend has existed throughout Southern Europe, with the ratios of Greece, Spain, and France reaching high levels in the years after 2010. Although it is actually France which has the highest dependency ratio of these countries, a result of its having a relatively large population of children, it is Italy which has their highest old age dependency ratio (population older than 65, relative to population 15-65):
If we look at Europe as a whole, including countries in its surrounding region, we can see there is a divergence occurring between northern and southern countries. Northern countries such as Germany, Russia, and Poland, which have had some of the lowest dependency burdens in the world in recent decades, will see sharp increases in the years ahead because their largest population cohorts are approaching 65 years old and they have few teenagers approaching 15 years old. (An exception to this is Ireland, which has had a fairly high ratio because of relatively high birth rates, but is not likely to have this increase much going forward, as it has few people approaching 65). Mediterranean countries, in contrast, will have their dependency ratios rise more slowly, because they have more children or because (particularly in Spain) their largest age cohorts are now only in their forties rather than their fifties. Within the EU this is especially true of France, but it is even more true of non-EU Mediterranean countries such as Turkey and Tunisia. These countries used to have far higher ratios than the EU or Russia, but no longer do today.
This fall in dependency in places like Turkey and North Africa is part of a greater trend, in which countries in the “global south”, particularly those outside of Sub-Saharan Africa, have recently seen their ratios fall much more quickly than countries in Europe, North America, or Northeast Asia. India’s dependency ratio, for example, fell below both the US’s and Germany’s in 2016. So did Bangladesh’s. (Pakistan’s ratio is falling too, but still remains high, around the level of Japan’s). Latin America’s is even lower; it recently became the lowest of any region, excepting East Asia. The major country that has had the most significant fall in dependency, however, is Iran:
Of course, age dependency ratios are simplistic. They treat all people above the age of 65 and below the age of 15 as if they were the same, and all people between 15 and 65 years old as if they were the same. Yet if (for example) we were to change the upper limit of working age from 65 to 70, Japan’s dependency ratio would fall substantially as a result, because Japan’s largest age cohort today is 65-70 years old. If, on the other hand, we were to change the lowerlimit of working age from 15 to 20, many middle-income countries’ ratios would rise substantially. To address these obvious shortcomings, alternative measures of dependency have been created. Examples of these include the economic dependency ratio, health care cost age dependency ratio, pension cost dependency ratio, and prospective old age dependency ratio. For each of these measures, Canada is forecast to have the biggest increase in the decade ahead among significant OECD countries, while Italy and Britain are expected to have among the smallest increases.
A primary lesson that can be learned from the analysis of age dependency ratios is that the common “young population good, old population bad” view of countries’ economic prospects is a misleading one. In reality countries with young populations tend to remain poor, in part because the youngest countries in the world (in Sub-Saharan Africa) are much younger than the oldest countries in the world are old. It will still be a number of decades before aging populations lead Europe or North America to have a higher age dependency ratio than Sub-Saharan Africa. And even that assumes that no unexpected shifts in migration or fertility will occur.
What age dependency ratios do show is two big trends, both of which have to do with middle-income economies. The first trend is the emergence of what we might call a goldilocks belt, located between the aging populations of North America, Europe, and Northeast Asia and the youthful populations of Sub-Saharan Africa. South Asia, North Africa, and Latin America all now appear to be in the process of supplanting high-income countries in terms of having the demographic trends that are arguably most conducive to (or at least, indicative of) economic growth.
The second trend is that Northeast Asia’s dependency ratio, which has been the lowest in the world for a generation and probably played a significant role in helping the region emerge from a low-income to middle-income level, bottomed out almost a decade ago and is continuing to rise.
Taken together, these trends suggest opportunities for middle-income countries, particularly those countries located in or near to the Mediterranean and Caribbean regions, to increase their exports to developed economies, given the aging labour forces of developed economies and traditional exporters in East Asia. In contrast, these trends also suggest that there should perhaps be a greater level of caution regarding the younger, high-growth economies in East Africa, such as Ethiopia or Kenya, which have recently been among the favorites of some emerging market investors.