Young enterprises make a disproportionate contribution to job creation, research by the OECD has found.
Even allowing for the jobs extinguished when young firms fail - as a lot of them do - their net impact on employment growth is greater than their share of employment overall.
In New Zealand's case, young firms - five years old or less - employed about one in every five employees in 2001-09.
Between them they generated 53 per cent of the new jobs. They also accounted for 29 per cent of the jobs lost, but even so, their net contribution of 24 per cent to job creation was disproportionately positive.
That pattern holds good across the 18 countries and 10 years or so in the OECD's study.
It finds that young firms' contribution to net employment growth remained positive during the Great Recession, even though it affected them more severely than older firms, both in terms of job creation and job destruction.
"The drop in their growth rate did not change the fact that young firms remained net job creators even during the crisis," says the report. "Most of the job losses during the crisis took place via contractions in surviving mature businesses rather than via exits of older firms, reflecting the greater weight of older businesses in the economy."
Even in good times, the attrition rate among young firms is high.
Statistics New Zealand's business demography statistics tell us that of the enterprises "born" in 2003, only 27 per cent were still going last year.
Survival rates were highest in the primary sector (agriculture, forestry and mining) at 36 per cent, and utilities (electricity, gas and water) at 35 per cent. The chances of going the distance were lowest in the transport, postal and warehousing sectors, where 21 per cent survived.
Among businesses born in 2003 and afterwards, the four-year survival rate has been remarkably consistent at around 50 per cent. Even among those born in the recession year of 2008, 47 per cent were still around four years later.
Larger birth weights (in terms of employee count) increase the chances of survival. "Only one in four non-employing enterprises that started in 2003 survived the full 10 years to 2013," says Statistics NZ. "By comparison almost half the enterprise births in 2003 that had six to 49 employees survived until 2013."
Richard Fabling, one of the New Zealand economists contributing to the OECD study, says there is an "up or out" dynamic.
"There is a baptism of fire. Either you have got a good idea or you don't."
New entrants either record higher than average growth, or they get out of the market.
So why do young firms make such a disproportionate contribution to gross job creation? The reason, the OECD study concludes, is related to their greater experimentation with new technologies, ideas and business models.
The growth of young ventures depends on the ease with which resources, especially labour and capital, can be reallocated, it says, invoking the concept of "creative destruction" associated with the Austrian economist Joseph Schumpeter.
"The restructuring process by which new (more innovative, more productive) production units replace less innovative and less productive ones - known as creative destruction - is reflected in the positive association of job creation and job destruction rates in an economy," it says.
A graph plotting job creation against job destruction rates across countries, years, firm age, firm size and sectors, shows a significant positive correlation between the two.
In New Zealand's case, the enterprise birth rate (the number of births as a percentage of the number of existing enterprises) over the past 10 years has averaged 12.2 per cent. The death rate has been 10.4 per cent. Those numbers are in line with the average rates of 12 and 11 per cent respectively in the OECD study.
But the OECD finds evidence of a clear decline in start-up rates across a wide range of countries.
"During the Great Recession there has been a further decrease in the entry rate of new firms, which was already steadily declining in most countries since the beginning of the last decade," it says.
"Given the importance of firm entry on a host of economic outcomes beyond job creation, for example innovation, competition and productivity, this finding raises particular concerns."
New Zealand's average enterprise birth rate of just over 12 per cent over the past 10 years masks a declining trend: from 16 per cent in 2004 to 9 per cent in 2013, or from 69,000 to 43,000 firms.
Last year was also the fourth successive year in which deaths outnumbered births.
Clearly, the economic cycle has a lot to do with that.
The (February) 2004 year's 16 per cent was the peak for the enterprise birth rate since 2001; two years earlier it had been 12 per cent.
It also coincided with a period when the economy expanded by 5.4 per cent.
The more subdued birth rate over the past four years - 9 or 10 per cent - followed the deepest recession since the 1970s.
But it still leaves open the question whether there are regulatory or other impediments which have a prophylactic effect on the business birthrate, or contribute to a higher infant mortality rate among enterprises than need be.
On the whole, probably not, thinks Business NZ chief executive Phil O'Reilly, who is also the current chair of the OECD's Business and Industry Advisory Committee (BIAC).
He points to the World Bank's league table for ease of doing business, in which New Zealand ranks No 3 overall and top for ease of starting a business.
"There is no particular legal impediment to starting a business, even though it has become somewhat harder with the requirement for at least one New Zealand resident director," he says.
"And our banks are still lending to small business."
In some OECD countries, access to bank funding is still an issue and BIAC is debating with the OECD about getting banking regulation right in order to free up that source of capital, O'Reilly says.
"In New Zealand the banks are open for business. Since the GFC banks have become more disciplined about requiring cashflow forecasts and business plans. That's a good thing," he says.
"But if a business wants to borrow based on its cashflow, its debtors, the interest rate might be a bit high."
It is relatively harder for start-ups to get equity funding in New Zealand, so many debt-fund, with their houses as collateral.
"And that is problematic because it makes them very risk averse about growth."
A lot of businesses, O'Reilly suggests, start small and stay small as a matter of choice.
"They start as a one-man or husband-and-wife sort of business and they stay there because they aren't particularly seeking to expand. That's a perfectly legitimate choice."
But if there is less "up" in New Zealand's up-or-out dynamics compared with larger economies, there is also less "out", he thinks.
"Our survival rates are quite high, I suspect. I don't think it is to do with protectionism or product market regulation or because there is not enough competition in the markets in which they operate."
But the small size of the New Zealand market can tend to limit the size of the competitors they face.
"In New Zealand there is a lot of competition going on with other small businesses but you won't tend to see three or four large businesses eating your lunch if you are a small business. You might, but it is easier to survive as a Mum and Dad business where in the US or UK you would get eaten alive."
O'reilly says one of the things that gets in the way of firms growing is that New Zealand businesspeople tend to want to own 100 per cent of something small rather than 51 per cent of something bigger.
"But we also tend to be able to co-operate well with others. I am particularly alive to that because I deal with trade associations, where I see them doing that."
So it is important competition law fosters competition where it matters, but also does not impede that collaboration where it is useful - a very hard balance to strike.
He gives some of the credit for employment growth among young enterprises to the 90-day rule the trial period when new employees can be fired without the normal recourse against unjustified dismissal.
"Small business just loved that. Prior to that there was an enormous amount of fear among those tiny businesses about taking on staff," he says.
"The perception - not necessarily reality - was that the risk was off the scale, relative to the reward. The 90-day rule has encouraged them to take a punt."
The general problem of a skills mismatch between those job seekers offer and what employers need is particularly acute for young businesses, O'Reilly says.
"They can't afford to train people. So they tend to want to hire quite a precise skill set - often a superman with several skills."
But overall start-ups benefit from a pretty flexible labour market, compared with many OECD countries, O'Reilly says.
And the New Zealand tax system has avoided funding a lot of social expenditure out of payroll taxes, which means the tax "wedge" on wages - the gap between the cost of employing someone and his or her take-home pay - is relatively small.
Richard Fabling says the findings on the implications of firm size and age for employment are only the first fruits of the OECD research programme in this area, which will also explore the links between business demographics and productivity.
Schumpeterian creative destruction is not the only dynamic at work. Older and larger firms can typically afford to invest more capital per worker, undertake more research and development and find it easier to enter export markets.
O'Reilly says business births are important not only for employment, but also for enterprise growth. "The next big company in New Zealand will come from a small company in New Zealand, so the more you create the better."
Bigger businesses tend to have the balance sheets and capabilities that enable them to take on larger-scale risks, he says.
"They will also be the guys who invest in the skill system. They also create management talent that spins out to create new companies, often in the supply chains of the big businesses they come from. And studies tell us we have a lack of management talent in New Zealand," he says.
"So it's important that we don't think that as long as we have lots of start-ups we are good to go."