1.3 Million Or Bust!

Yesterday the Conservative Party announced a goal of "creating" 1.3 million net new jobs by 2020



(I put "creating" in quotations, since governments only *create* public-sector jobs, though they have an important role in ensuring a good environment for private sector job creation.)

Is 1.3 million by 2020 a reasonable number? There has been an active twitter discussion on the merits of this number, and the consensus seems to be a cautious "yes". Cautious because while it may be reasonable, it is by no means automatic and depends on how fast the economy grows between now and 2020 or how large immigration flows are between now and then. As I haven't seen more than 140 characters worth of analysis of this at a time, I thought I would put my own thoughts down to explore the numbers. (Please tell me if I've missed someone else's analysis. For detailed tweets that support the figures below, see @kevinmilligan's feed)

First, a rough back of the envelope trend of monthly employment growth, projected out to 2020, is very much in line with the 1.3 million number. Over the past 15 years, the monthly average net new job creation is about 17,700. Projected out 64 months to December 2020 (to be generous), yields just over 1.1 million.

Of course, our population is ageing so history may be of very little guide. In fact, our population is ageing so quickly that basically ALL of statistics Canada's projected population growth for Canada comes from those over the age of 55. Not too surprising, as a good chunk of Canada's population is already over 55:



More than an ageing population, we also likely to have a slightly falling population among the younger age groups. Here are Statistics Canada's projections:



Notice they have various scenarios. I've included the Low, Medium, and High scenarios. This will be useful in the analysis to come.

As employment rates (the fraction of people working) vary dramatically across age groups, this is a very important fact to consider. Let's look:



The 20-24 and 25-54 age groups have a fairly stable employment rate . They are currently 68.6% and 81.5%, respectively. Pre-recession, their employment rates were slightly higher, at around 71% and 82%. Returning to pre-recession employment rates will be an important part of meeting the 1.3 million net new job goal.

So, with all this in mind, we're ready to forecast out future net new jobs created between now and 2020. If we know the change in an age group's population, and it's projected employment rate, we can forecast the projected employment.

Let's start with NO change in employment rates:



Not quite there. Under the high growth scenario, we're looking at 700,000 net new jobs by 2020 -- roughly half the goal.

What if all employment rates returned to their pre-recession rates (with the exception of the 55+ group, which I'll keep at it's current level):

That's getting closer! But, still not quite there. Even the high growth scenario for population is about a quarter million new jobs short of the goal.

So, what will it take? Well, two facts are important:

1) There is very little prospect for much higher employment rates among prime-age workers. We saw in the earlier graph that they've been fairly stable.

2) There are *a lot* of people going to be 55+ by 2020. Changes in their employment rates will therefore be critical.

The increase in 15-19 year old's employment rate won't really matter too much.

By how much must we increase the employment rate among those 55+ in order to have a reasonable chance of meeting the 1.3 million goal? It must rise from it's current 35% to roughly 39%.

Or, perhaps we could dramatically increase the 15-19 year old employment rate to 50%. Then we would only need to increase the employment rate among the 55+ group to 38%.

So, what all this tells me is that 1.3 million new jobs by 2020 is entirely feasible. It would require populations grow at the Medium or High scenarios of Statistics Canada. Immigration can help ensure this is the case. It will also require we increase employment rates among those above the age of 55. What the most effective policies are to achieve this is far beyond the scope of this blog post, so I'll let that one go. But, given that the employment rates among this group have risen so dramatically in recent years (from 20% in the mid-90s to 35% today), a further increase to 39% doesn't seem ridiculous.

The way we think about trade has changed; policy makers should take note

Last week I wrote about the potential gains Canada's auto sector could see from cheaper access to imported parts. The vast majority of auto parts used by producers in Canada are imported. Lower trade barriers will therefore lower costs, lower prices, increase exports, increase employment, and increase output in Canada's auto sector -- to say nothing of the gains for the economy as a whole.

The issue is back in the news yet again. In the Globe and Mail debate on the economy, Stephen Harper committed to successfully concluding the Trans-Pacific Partnership. Saying,

“We’re entering the final stages of a trade discussion in the Asia Pacific that I think frankly is going to conclude successfully, that is going to be the basis of the global trade network in the Asia Pacific for the generation to come,” (source)

In pitching the deal, he had specific words for the auto sector:

"What I say to the auto sector in particular — I am not suggesting they will like everything that is in that — is we simply cannot afford as a country to have our auto sector shut out of global supply chains. That would be a disaster."

The reaction was swift. Ontario's Minister of Economic Development Brad Duguid, for example, came out strongly against the deal. So strongly that he thinks accepting the deal displays a "lack of economic leadership". The impression among media coverage is that the industry is up in arms (for example, here). Those who are fearful are (usually) not making statements about Canada's economy as a whole. (I'm being charitable here). They are concerned with the well being of a particular set of workers or the size of a particular industry.

These concerns are not crazy, but they are based on old ways of thinking about trade. A modern understanding of international trade thinks seriously about global value chains (traded inputs, such as parts).

In all that follows, I'm going to use a standard quantitative model of international trade. All data comes from the World Input-Output Database. If you want details of the model, see here or a forthcoming CJE paper here.

In the old way of thinking about trade: Lowering import costs always harms the import competing sector; lowering its employment and lowering its sales. On average, for each 1% reduction in import costs, sales of the import competing sector fall by just under 1% (but of course that differs by sector).

BUT, these simple models abstract from a very (very!) important aspect of current global trade flows: inputs are traded. Canada's auto sector doesn't just use workers to make cars, it also uses manufactured parts. About 55 cents is spent on motor vehicle parts for each dollar of output of the motor vehicle manufacturing sector as a whole (see Canada's Input-Output Data). Cars need parts and about 80% of all auto parts used by auto manufacturers in Canada are imported from abroad (two-thirds from the US and one-third the rest of the world). See my previous post for further details.

The new way of thinking about trade: think seriously about traded intermediate inputs. Let's incorporate such inputs into our model and see what happens. I'll focus on the transport equipment sector (basically autos). 

This changes everything. Canada's auto sector is so highly integrated globally, with many of its inputs coming from abroad, that lower import costs in that sector are a good thing for that sector!

The key takeaway message here is simple: Concerns about Canada's auto not gaining from lower import costs in that sector are based on old and out-dated ways of thinking about international trade. Most trade around the world is in intermediates, not final goods. We must change the way we think. The gains from trade are so much larger than we previously thought.

It gets even better! Trade agreements don't just force Canada to lower its import costs while other countries do nothing. Exports from our country to theirs also become easier. The gains from lower import and export costs are even larger. Let's take a look by simulating a 1% reduction in export costs and import costs in Canada's auto sector.

The gains are very large! For each 1% reduction in import and export costs (together), employment and production in Canada's transport equipment sector increases by about 0.3%! The overall effect on Canada's real GDP is about 0.05% -- or $1 Billion in increased GDP for each 1% reduction in trade costs in the transport equipment sector. A lot of money to leave on the table if we don't succeed in liberalizing international trade.

Of course, these numbers are only illustrative. While they do represent the newest techniques in international trade research used to quantify the effect of policy changes at the sector level, I don't want to hang my hat on any specific number. The key result to take away from this analysis is that Canada's auto sector likely stands to gain -- not lose -- from lower international trade costs.

The gains from trade liberalization in autos is large, but so are the gains in many other sectors! If I repeat the model exercise as above but for all sectors, I find the following:

I've highlighted the transport and equipment sector in orange. Improving trade in resources (mining, oil, and gas) yields even larger gains. Lowering trade costs in agriculture and processed foods yield also yield sizable gains, highlighted in green, on the order of $100-300 million per 1% reduction in trade costs in those sector. For the sake of Canada's overall economy, I hope TPP negotiations lower agricultural trade costs substantially. They may, but battling the entrenched interests of Supply Management is a daunting challenge that no politician seems yet capable of handling. (As an aside: even agriculture stands to gain, with increased employment and higher sales if trade costs fall in that sector. So, it's really a battle between unproductive producers and productive producers within agriculture.)

Let me end by saying that almost no credible economist thinks trade is harmful to an economy overall. There are small exceptions and nuanced policy positions, of course. But lowering trade barriers is widely and generally found to be a good thing on average. We should have policies that compensate losers or facilitate adjustment, but not policies that block trade liberalization outright. Trade agreements also incorporate clauses that go beyond the mere trade in goods and services. These may or may not be of concern, but it they are a separate issue from the gains from trade. 

Don't let confident political rhetoric fool you, there is only room in reasoned debates for careful analysis, evidence, and data. 

What's a Government Deficit? Does it Matter?

Today the federal government released their Fiscal Reference Tables and reported a small surplus of $1.9 billion last year ("small" since it's only 0.1% of GDP). During an election campaign, this is obviously big news and clearly matters for a party's political fortunes, if today's Conservative Party Banner is any indication:

But, what is a deficit? And, more importantly, why does it matter?

Roughly speaking, it's simple: if a government spends more than it takes in, it must make up the difference by borrowing. That's a deficit. It also matters. We all have a stake in knowing that government programs, its tax policies, and so on, are sustainable into the future. So, we should measure changes in a government's financial position and scrutinize it carefully.

But, as is all too often the case, the devil is in the details. What counts as government revenue? What counts as government spending? If Canada Post records a loss on its worker pension plan, should that count against the federal government budget? If the government sells its shares in General Motors, should that count? Hard questions to which there is no easy answer.

There is more than one way to measure a deficit, and the various measures sometimes tell different stories. Let's look at the 2014-15 year in particular.

Two measures show a deficit, while another shows a surplus. (For some details behind why the measures differ, see page 11 of this.) For those who like dollars:

Notice the government reports a $1.9 billion surplus, but based on change in net debt (another measure of a budget's deficit) the deficit was $4.6 billion!

So what are we to do? What are these measures anyway? Was the budget balanced or not? I am no expert in the pros and cons of the various measures, but interested readers should see this great summary document prepared by the Public Sector Accounting Board (more here). I'll just briefly mention them here.

First, the headline budget balance (the number everyone talks about) is what accountants would recommend (see the PSAB document). Second, there are changes in a government's "net debt". Essentially, this shows whether a government's liabilities is increasing faster than its financial assets. This measure is important economically, but also political. Remember Ralph Klein famously holding up the "Paid in Full" sign?

Alberta didn't actually eliminate is overall debt, it eliminated it's "net debt". So, changes in a government's net debt tells us how far governments are from their own Klein Moment.

The last measure I'll bring up is changes in "accumulated deficits" that also account for non-financial assets (such as physical capital).  This measure is very close to changes in a business equity, but for the government.

Luckily, it really doesn't matter what measure you use most of the time. They usually tell the same story. Below I plot those three measures as a share of Canada's GDP. Notice you can barely even see differences between the lines in most years.

But they don't always agree, as the earlier bar graph demonstrated.

Let's move away from specific measures and think about whether budget balances matter in general. Is the economy harmed by deficits? The CPC and NDP seem to think so, as they both insistent budgets should always be balanced. Justin Trudeau and the Liberal party are taking the other side. They claim we need deficits to "stimulate" the economy (here or here). Are they right?

It's tricky to know for sure -- and providing clean and clear evidence on the effect of government taxes and spending on an economy requires extremely clever empirical strategies. The difficulty is, government spending is generally higher in a recession (more EI cheques, for example) and revenue is generally lower (lower income taxes). So, deficits are correlated with recessions but this doesn't mean deficits cause recessions (at least, not any more than cheese kills people in their sleep). Also, government may ramp up spending in recessions (to "stimulate" economies) and then subsequently economies start to grow. It may look like the stimulus spending helped the economy, but perhaps the economy was going to grow anyway.

Thankfully, there are clever researchers out there and they have some evidence. Interested readers can find a summary here. Most relevant for Canada, the most recent evidence suggests small countries that are open to trade and have flexible currencies (which Canada has) should expect ZERO effect of government spending on GDP. For a variety of reasons (currency movements, interest rate adjustments, consumer spending changes, and so on) the effect of government spending is likely fully crowded-out.

Does a zero effect mean we should always balance the budget each and every year? No. My colleague Ron Kneebone, in a great summary of government budgeting (here), provides a simply visual:

The CPC and NDP seem to currently favour the black line. Economists generally favour the red line. Balance the budget over the business cycle, not every year.

Don't balance the budget each year, but don't run perpetual and permanent deficits either. This consensus doesn't stem from strong evidence of stimulative effects of spending during recessions but more from sensible smoothing of program spending and tax policy over the business cycle. Keep in mind that this has nothing to do with the size of government. You can balance the budget with a higher level of taxation just as easily as with a low level.

Okay, let's review. 

Was Canada's 2014-15 budget balanced? That's in the eye of the beholder. My take is simple: yes, the budget is "balanced" because inflows and outflows (however measured) are roughly the same magnitude. Imbalances on the order of tenths of a percent of GDP just don't matter -- at all. In this election, let's focus less on the binary metric of having a balanced budget or not and talk about policies and programs we do or do not favour.

Will a deficit stimulate Canada's economy? The evidence suggests stimulus probably wouldn't do much to Canada's GDP. Advocates of stimulus spending must present evidence for why they think it will help and why the existing evidence is wrong. They usually don't. We must demand better.

Trade Barriers Harm the Auto Sector

The NDP recently announced that if they form government, they would increase barriers to importing auto parts from abroad. Not surprisingly, Unifor (representing approximately 30,000 motor vehicle manufacturing workers) agreed.

Is this a wise policy position for they and their supporters to take? Will higher barriers to auto parts imports help Canada's auto sector and its workers? No. It's a counterproductive policy that, as I detail below, is actually more likely to harm the auto sector as a whole and its workers (to say nothing of Canada's economy generally). Unifor and the NDP need to seriously rethink their position on this.

How can import barriers harm the sector they are meant to "protect"? By making it more costly to produce and thereby lowering its competitiveness. About 55 cents is spent on motor vehicle parts for each dollar of output of the motor vehicle manufacturing sector as a whole (see Canada's Input-Output Data). Cars need parts, not surprisingly. It turns out that most of these parts are imported. In fact, about 80% of all auto parts used by auto manufacturers are imported from abroad (two-thirds from the US and one-third the rest of the world).

We see this fact manifest itself in the trade data:

Canada's massive trade deficit in parts is surely cause for alarm. Not so. Access to cheap inputs make the end product more competitive. Conveniently, the OECD calculates just how important imported inputs are to the success of each of Canada's sectors. It's huge. Imported inputs are a critical part of for motor vehicle exports.

So, with so much of its imported imported from abroad, what would higher trade costs do to the sector? Trade economists have a wide variety of tools available to estimate the consequences of trade costs. One of the most recent workhorse models (details here or here) allows us to simulate how a change in trade costs will affect output, productivity, employment, exports, and a variety of other metrics for any of Canada's sectors.

The data used to set up this model unfortunately aggregates auto parts and auto manufacturing within a broader Transport Equipment sector. That's okay; presumably the NDP values auto manufacturing workers and auto parts manufacturing workers equally. Workers are also presumably highly mobile between these two sectors. The broad patterns described above for autos holds for this broader sector, with one third of transport equipment inputs used by the transport equipment sector imported from abroad.

So, does cheaper imports of transport equipment help or hurt Canada's transport equipment sector? Let's simulate a 1% decrease in the cost of importing into Canada but let's not change the cost of exporting. This is a unilateral liberalization of Canada's transport equipment trade -- not simulating a joint agreement between us and some other country out there.

What does the model predict? Well, cheaper access to transport equipment inputs abroad does indeed lower the share of those inputs sourced locally -- that is, auto manufacturers will source more of their inputs from abroad. This is not surprising. But, with cheaper access to inputs, production costs fall and productivity rises:

With lower production costs, the sector's output becomes more competitive. It exports more and hires more workers as a consequence! 

So, cheap imports of transport equipment goods into Canada actually helps the transport equipment sector broadly. It may seem counter intuitive at first, but this sector uses so many imported inputs it is in a unique position to itself benefit from free trade. Canada's auto sector is highly and successfully integrated globally. The NDP and Unifor should recognize this and support freer trade. After all, demanding trade barrers is tantamount to admitting Canada's auto sector lacks international competitiveness. It doesn't.