Now that the Obama administration has finalized its Clean Power Plan regulating greenhouse gas (GHG) emissions from the power sector, the focus of attention turns to the states, which must now find a way to reduce emissions consistent with the Plan. One question states face as they envision a lower carbon future is how much to rely on natural gas-fired generation.

The Environmental Protection Agency Plan encourages states to use existing gas-fired generators more and coal-fired generators less, and to build new zero-emission generators (wind, solar and nuclear). The Plan neither encourages nor discourages construction of new gas-fired generators, but some environmental groups oppose additional natural gas plants, fearing they will slow the advance toward a carbon-free grid. Owners of competing technologies also prefer fewer new gas-fired generators, recognizing that inexpensive natural gas has been a key driver of lower electricity prices that cuts into their profits.

But there are solid reasons why the electric grid needs gas more than ever as more renewable power comes on line.

First, in most electricity markets gas competes most directly with coal, not renewables. The reason is that electricity is dispatched on a marginal cost basis (that is, based on the operating cost of the next available increment of energy): whenever there is a renewable resource available, it will almost always be dispatched to the grid because its zero fuel price will trump the non-zero fuel price of coal- and gas-fired generators. The question, then, is which technologies will power grid operators use to supplement or back up renewable power when the wind is not blowing and the sun is not shining.

Second, gas-fired generators are better able than coal to accommodate more renewable power on the grid, because they can more efficiently adjust their output in response to the variability of renewables’ production. The Texas grid, for example, has been able to integrate large amounts of new wind power recently, in large part because of its complement of gas-fired generation. If Texas had only coal-fired power to back up wind, it would have dispatched less wind power to the grid, because the limited flexibility of coal-fired power would have reduced the ability to respond to variations in wind generation while keeping the lights on.

Third, modern, ultra-efficient gas-fired combined-cycle power plants produce only about half the carbon dioxide and small fractions of the other pollutants emitted by coal-fired power. Reducing carbon dioxide is a multi-decadal task, one we need to accomplish in a cost-effective manner. The U.S. has the oldest coal-fired generation fleet in the world in part because those dirty, old plants produce inexpensive, reliable power. We will need a combination of renewables and new gas-fired generation to replace the retiring coal-fired generators and maintain system reliability.

Indeed, there are technical characteristics of thermal generators (the nuclear-, coal-, and gas-fired generators) that remain essential to the operation of the grid. We currently have no cost-competitive ways for renewables and electricity storage to provide or simulate those technical characteristics.

Fourth, and perhaps most crucially, the enhanced reliability and environmental benefits of gas-fired power become more important with higher levels of renewable penetration, at least until cost-effective electricity storage options become available. While there have been great strides in reducing the cost of battery storage, it remains an extremely expensive solution to the problem of supporting renewable power generation.

In our capitalist system the future energy mix will continue to be determined in large part by price competition. Regulation affects prices, but decisions about which plants to build are made by the private sector, not by regulatory fiat. Gas is currently the most cost-effective complement to renewables and consequently will predominate in new construction of thermal generation.

We expect the costs of electricity storage to continue to fall, and for storage eventually to replace other generation sources as the primary supporter of renewable generation on the grid. But storage is not yet ready for prime time. In the meantime, we need flexible, efficient gas-fired power to ensure that the transition away from much dirtier, higher-carbon coal-fired power continues apace. It would be a dangerous bet to forgo new gas-fired generation now.

David Spence is Professor of Law, Politics & Regulation at the University of Texas at Austin, where he teaches in both the McCombs School of Business and the School of Law. Ross Baldick, is a Professor in UT Austin’s Department of Electrical & Computer Engineering.