The diplomats have done their job, concluding the Paris climate agreement in December. And political leaders gathered… at the United Nations to sign the new accord. But implementation is surely the tough part. Governments need a new approach to an issue that is highly complex, long term, and global in scale.

At its core, the climate challenge is an energy challenge. About 80% of the world’s primary energy comes from carbon-based sources: coal, oil, and gas. When burned, they emit the carbon dioxide that causes global warming. By 2070, we need a world economy that is nearly 100% carbon-free to prevent global warming from running dangerously out of control.

The Paris agreement recognizes these basic facts. It calls on the world to cut greenhouse-gas emissions (especially CO2) to net-zero levels in the second half of the century. To this end, governments are to prepare plans not only to the year 2030 (the so-called Nationally Determined Contributions, or NDCs), but also to mid-century (the so-called Low-Emission Development Strategies, or LEDS).

The world’s governments have never before attempted to remake a core sector of the world economy on a global scale with such an aggressive timeline. The fossil-fuel energy system was created step by step over two centuries. Now it must be comprehensively overhauled in just 50 years, and not in a few countries, but everywhere. Governments will need new approaches to develop and implement their LEDS.

There are four reasons why politics as usual will not be sufficient. First, the energy system is just that: a system of many interconnected parts and technologies. Power plants, pipelines, ocean transport, transmission lines, dams, land use, rail, highways, buildings, vehicles, appliances, and much more must all fit together into a working whole.

Such a system cannot be overhauled through small incremental steps. A deep overhaul requires system-wide re-engineering to ensure that all parts continue to work effectively together.

Second, there are still many large technological uncertainties in moving to a low-carbon energy system. Should vehicles be decarbonized through battery-electric power, hydrogen fuel cells, or advanced biofuels? Can coal-fired power plants be made safe through carbon capture and storage (CCS)? Will nuclear energy be politically acceptable, safe, and low cost? We must plan investments in research and development to resolve these uncertainties and improve our technological options.

Third, sensible solutions require international energy cooperation. One key fact about low-carbon energy (just like fossil fuels) is that it is not generally located where it will ultimately be used. Just as coal, oil, and gas must be transported long distances, so wind, solar, geothermal, and hydropower must be moved long distances through transmission lines and through synthetic liquid fuels made with wind and solar power.

Fourth, there are of course powerful vested interests in the fossil-fuel industry that are resisting change. This is abundantly clear in the US, for example, where the Republican Party denies climate change for the sole reason that it is heavily funded by the US oil industry. This is certainly a species of intellectual corruption, if not political corruption (it’s probably both).

The fact that the energy system involves so many complex interconnections leads to tremendous inertia. Shifting to a low-carbon energy system will therefore require considerable planning, long lead times, dedicated financing, and coordinated action across many parts of the economy, including energy producers, distributors, and residential, commercial, and industrial consumers. Policy measures such as a tax on carbon emissions can help to address some – but only some – challenges of the energy transition.

Here is another problem. If governments plan only 10-15 years ahead, as is typical in energy policy, rather than 30-50 years, they will tend to make poor system-related choices. For example, energy planners will move from coal to lower-carbon natural gas; but they will tend to underinvest in the much more decisive shift to renewable energy.

Similarly, they may opt to raise fuel standards for internal-combustion automobiles rather than to push the needed shift to electric vehicles. In this sense, planning 30-50 years ahead is vital not only to make the correct long-run choices, but also to inform the correct short-term choices. The UN’s Deep Decarbonization Pathways Project has shown how long-term plans can be designed and evaluated.

None of these challenges sits easily with elected politicians. The decarbonization challenge requires consistent policies over 30-50 years, while politicians’ time horizon is perhaps a tenth of that. Nor are politicians very comfortable with a problem that requires large-scale public and private financing, highly coordinated action across many parts of the economy, and decision-making in the face of ongoing technological uncertainties. Small wonder, then, that most politicians have shied away from this challenge, and that far too little practical progress has been made since the UN Framework Convention on Climate Change was signed in 1992.

One key step, I believe, is to remove these issues from short-term electoral politics. Countries should consider establishing politically independent energy agencies with high technical expertise. Of course, key energy decisions (such as whether to deploy nuclear energy or to build a new transmission grid) will require deep public participation, but planning and implementation should be free of excessive partisan politics and lobbying. Just as governments have successfully given their central banks some political independence, they should give their energy agencies enough leeway to enable them think and act for the long term.

With the Paris climate agreement now in force, we must move urgently to effective implementation.