LONDON – I recently took two trips to China just as the government launched its 12th Five-Year Plan to rebalance the country’s long-term growth model. My visits deepened my view that there is a potentially destabilizing contradiction between China’s short- and medium-term economic performance.

China’s economy is overheating now, but, over time, its current overinvestment will prove deflationary both domestically and globally. Once increasing fixed investment becomes impossible – most likely after 2013 – China is poised for a sharp slowdown. Instead of focusing on securing a soft landing today, Chinese policymakers should be worrying about the brick wall that economic growth may hit in the second half of the quinquennium.

Despite the rhetoric of the new Five-Year Plan – which, like the previous one, aims to increase the share of consumption in GDP – the path of least resistance is the status quo. The new plan’s details reveal continued reliance on investment, including public housing, to support growth, rather than faster currency appreciation, substantial fiscal transfers to households, taxation and/or privatization of state-owned enterprises (SOEs), liberalization of the household registration (hukou) system, or an easing of financial repression.

China has grown for the last few decades on the back of export-led industrialization and a weak currency, which have resulted in high corporate and household savings rates and reliance on net exports and fixed investment (infrastructure, real estate, and industrial capacity for import-competing and export sectors). When net exports collapsed in 2008-2009 from 11% of GDP to 5%, China’s leader reacted by further increasing the fixed-investment share of GDP from 42% to 47%.

Thus, China did not suffer a severe recession – as occurred in Japan, Germany, and elsewhere in emerging Asia in 2009 – only because fixed investment exploded. And the fixed-investment share of GDP has increased further in 2010-2011, to almost 50%.

The problem, of course, is that no country can be productive enough to reinvest 50% of GDP in new capital stock without eventually facing immense overcapacity and a staggering non-performing loan problem. China is rife with overinvestment in physical capital, infrastructure, and property. To a visitor, this is evident in sleek but empty airports and bullet trains (which will reduce the need for the 45 planned airports), highways to nowhere, thousands of colossal new central and provincial government buildings, ghost towns, and brand-new aluminum smelters kept closed to prevent global prices from plunging.

Commercial and high-end residential investment has been excessive, automobile capacity has outstripped even the recent surge in sales, and overcapacity in steel, cement, and other manufacturing sectors is increasing further. In the short run, the investment boom will fuel inflation, owing to the highly resource-intensive character of growth. But overcapacity will lead inevitably to serious deflationary pressures, starting with the manufacturing and real-estate sectors.

Eventually, most likely after 2013, China will suffer a hard landing. All historical episodes of excessive investment – including East Asia in the 1990’s – have ended with a financial crisis and/or a long period of slow growth. To avoid this fate, China needs to save less, reduce fixed investment, cut net exports as a share of GDP, and boost the share of consumption.

The trouble is that the reasons the Chinese save so much and consume so little are structural. It will take two decades of reforms to change the incentive to overinvest.

Traditional explanations for the high savings rate (lack of a social safety net, limited public services, aging of the population, underdevelopment of consumer finance, etc.) are only part of the puzzle. Chinese consumers do not have a greater propensity to save than Chinese in Hong Kong, Singapore, and Taiwan; they all save about 30% of disposable income. The big difference is that the share of China’s GDP going to the household sector is below 50%, leaving little for consumption.

Several Chinese policies have led to a massive transfer of income from politically weak households to politically powerful companies. A weak currency reduces household purchasing power by making imports expensive, thereby protecting import-competing SOEs and boosting exporters’ profits.

Low interest rates on deposits and low lending rates for firms and developers mean that the household sector’s massive savings receive negative rates of return, while the real cost of borrowing for SOEs is also negative. This creates a powerful incentive to overinvest and implies enormous redistribution from households to SOEs, most of which would be losing money if they had to borrow at market-equilibrium interest rates. Moreover, labor repression has caused wages to grow much more slowly than productivity.

To ease the constraints on household income, China needs more rapid exchange-rate appreciation, liberalization of interest rates, and a much sharper increase in wage growth. More importantly, China needs either to privatize its SOEs, so that their profits become income for households, or to tax their profits at a far higher rate and transfer the fiscal gains to households. Instead, on top of household savings, the savings – or retained earnings – of the corporate sector, mostly SOEs, tie up another 25% of GDP.

But boosting the share of income that goes to the household sector could be hugely disruptive, as it could bankrupt a large number of SOEs, export-oriented firms, and provincial governments, all of which are politically powerful. As a result, China will invest even more under the current Five-Year Plan.

Continuing down the investment-led growth path will exacerbate the visible glut of capacity in manufacturing, real estate, and infrastructure, and thus will intensify the coming economic slowdown once further fixed-investment growth becomes impossible. Until the change of political leadership in 2012-2013, China’s policymakers may be able to maintain high growth rates, but at a very high foreseeable cost.