The Romans depended on road and sea transport to keep their sprawling empire together.
Their workmanship was so good that we can still see examples of their roads and aqueducts over 2000 years after they were built - a fact which has informed the basic and common-held assumption that virtually any infrastructure spending will be of value in to the future.
Infrastructure creates value when it boosts productivity, specifically by lowering the cost and / or time associated with transporting goods, people, energy. So in order to be of true benefit, the value created by this increased productivity must outweigh the cost.
In today's developed Western economies, however, true opportunities to boost productivity via new infrastructure are scarce.
The main reason infrastructure spending was pro-GDP in days past, was driven by a need to keep up with the rapidly developing and advancing private industrial sector.
So does rebuilding and/or adding infrastructure indeed create economic value? We must consider the hundreds of millions of dollars' worth of proposed infrastructure projects currently in the pipeline in terms of their productivity and cost-benefit.
Replacing existing infrastructure isn't a guaranteed one-way bet. It may be necessary, but unless the replacement generates real gains in productivity, it acts as a tax - diverting capital that could have been invested elsewhere to greater benefit.
Rebuilding an existing bridge might generate immediate value via spending on materials and wages, but if it doesn't create additional productive capacity at least equal to its cost, the benefits of this additional spending run out once the project is complete. On the other hand, the cost of funding (given that these projects are usually funded by debt) continues to grow long into the future.
Taking Auckland as an example, debt constraints on Auckland Council will have a real bearing on the likely investment programme and subsequent direction and quality of infrastructure.
The Council estimated NZ$18.7b would be required for investment in new assets for 2015-2025 as part of their long term plan (LTP) - and while central Government is earmarked to pick up some of this cost, the remainder is a composition of increased borrowings, user charges and rates. With net debt to total revenue at roughly 195 per cent, significantly increasing debt levels will be both costly and problematic.
Funding alternatives could come in the form of private-public partnerships (PPPs), infrastructure bonds or asset sales. While PPPs optically remove debt from the balance sheet, they are most effective when the private operator is responsible for its own revenue generation - with toll roads being a prime example.
Asset sales (particularly in an election year) are a political football. On one hand they can help to reduce debt (and revenue), but this is a move seen by many as 'selling the furniture' and the central Government asset sales back in 2013 have shown the difficulty and political will required for even a partial sell down.
Bridges to nowhere aren't just a waste of money in the present, but burden the economy with costs for years to come.
So, while the general consensus is that we need to invest in more and better infrastructure around the country - roads, rails and ports, we must carefully weigh-up the true costs and benefits of these projects.
Mark Fowler is Head of Fixed Income at Hobson Wealth Partners.