By Andrew Winston

May 14, 2012 - This week, Microsoft is announcing an unusual initiative that it hopes will change how the company operates: an internal fee on carbon.

Starting July 1st - the beginning of the company's fiscal year 2013 - the software giant will charge all of its 100-plus global offices and datacenters a fee for every ton of carbon they produce (mostly from plugging into the electric grid, so-called "indirect" emissions). The money collected will go to purchase Renewable Energy Certificates (RECs) and carbon offsets, allowing Microsoft to declare itself carbon neutral.

Although carbon neutrality is a claim that's gotten less credible over the last few years, Microsoft seems well aware of the challenge and is handling it well. Here's the problem with asserting neutrality: it's hard to ensure that any carbon reductions you're paying for - from, say, capturing methane from a landfill or replacing inefficient cook stoves in Africa - would not have happened anyway (this is the problem of "additionality").

So to navigate this tricky terrain and make the most of its efforts, Microsoft is working with a handful of NGOs and a well-respected partner, Sterling Planet, which will buy the RECs and offsets.

Interestingly, however, carbon neutrality doesn't seem to be the real point of Microsoft's initiative. From my conversations with the company, I take away four major reasons they're doing this:

Behavior change. Microsoft seems more interested in lowering overall carbon emissions and energy use, not just neutrality in and of itself. This focus on actual emissions and outcomes is the right way to go. The offsets become a tool, a last resort to be avoided, and both energy efficiency and using renewable energy (onsite or directly purchased) become the paths of least resistance and cost. As Rob Bernard, Microsoft's Chief Environmental Strategist says, "If you run one of our offices, and you choose to use carbon-based power, we'll charge you more for your energy." And this charge will, in theory, move managers to make greener choices. So the point of this fee, like all "taxes," is to change behavior, discouraging some pathways by making them less palatable.

Accountability throughout the organization. Each division is going to own this issue. Let's say NGOs or customers are asking questions about what kind of energy Microsoft uses to power its datacenters. The executives running that facility, not just centrally located sustainability professionals, will be empowered to address any concerns, drive for greater efficiency, and choose greener power. Pricing carbon is an excellent way to raise awareness internally before the external pressure builds.

Risk reduction. Bernard and his colleague who's running the program, TJ DiCaprio, have encouraged the organization to better understand its profound energy-related risks. As Microsoft takes on more of its customers' operations through cloud-based services, reliance on the utility grid creates real operational and price risk (from outages and volatile prices). Cloud service providers are increasingly proxies for utilities - they require 100% uptime, significant quantities of their own power, and predictable variable cost (which for renewables is nearly zero).

Sales/Becoming the Vendor of Choice. The company knows that its customers are increasingly looking for providers that can offer reliable service at low cost - and low carbon emissions. Driving the organization to use more alternative energy helps land large customers concerned about their value-chain footprint.

So how will Microsoft make this initiative a reality? The execution plan has some interesting elements (see a pithy white paper on the full carbon neutrality plan here). The company will measure carbon footprint in different operational buckets such as plug load (electricity used) and business travel, and then offset each category "like for like" (i.e., buying RECs for electricity and offsets for travel). The fees will vary as well; for example, the company will charge employees for all air travel on a per-mile basis, which raises awareness at the individual level.

The program is smart, but I'm left with one major concern: Will the fees be high enough to change behavior? Right now, the market price of carbon is very low, so Microsoft is charging a small amount per ton. Even so, they will collect north of $10 million, which is enough to buy offsets. Over time, by my calculations, given the growth in cloud services the company is banking on, these carbon fees could rise to a more noticeable $50 million by 2020. But let's be honest, these numbers are clearly rounding errors to a company that netted $23 billion last year.

To be fair, as the behavioral psychology gurus (from books like Nudge) will tell you, sometimes just making people aware of a cost is enough to foment change. So the minor nudge here may be good enough. But in the longer run, the price on carbon needs to be more of a sledgehammer than a nudge. It should reflect the full cost to society of health impacts, national security risks, and price volatility, all of which add up to tens of dollars per ton or more.

Pricing carbon on your own, without a real market in place, is hard, which is why there are so few examples of companies doing it. Going back over a decade, BP put in place an internal carbon trading system that used a "shadow" price to encourage divisions to find the cheapest reduction opportunities (others like Shell have also used this tactic). But executives weren't charged real money. And more recently, athletic apparel company Puma (working with my colleagues at PwC and the UK's TruCost) produced an "environmental P&L" which measured the "real" cost of carbon and other environmental inputs like water. The company is exploring how to include the "price" in operations and give ownership to line managers; it currently says the metrics will "inform operational decisions."

So even with important experiments like these that have gone before, Microsoft's program is perhaps the first actual internal fee at this scale (my research isn't turning up anything exactly like this - please send me examples if you have them!).

It's innovative and committed, but it also points to a massive global failure of leadership on climate policy. We should put a price on carbon across the entire economy. But as Bernard says, "While governments have an important role to play, we hope there's a benefit in us moving faster than the policy world."

He's absolutely right. Companies cannot wait for the government wheels to turn to price and manage carbon - the cost saving, risk reduction, and brand benefits of leading are too high.

This post first appeared at Harvard Business Online and is reprinted here with the kind permission of the author. Andrew Winston, founder of Winston Eco-Strategies, is the author of Green Recovery, a strategic plan for using environmental thinking to survive hard economic times. He is also the co-author of Green to Gold, the best-selling guide to what works - and what doesn't - when companies go green.