Governments take another hard look at taxing public clouds



Many that move to cloud computing spend less on taxable hardware and software, and move to services that are more difficult to define as taxable transactions. While most don’t consider tax issues as part of the public cloud computing selection process, the reality is that most state and local governments are now monitoring the growth of cloud computing with an eye toward tax revenue.

Cloud computing is not the only area where the technology outpaces the tax laws. The use of electric cars is giving states fits. States are quickly learning that promoting electric cars comes at a high price. For example, Colorado has joined a growing number of states that impose fees on electric and alternative vehicles to recover “lost” gas tax revenues.

Coloradans who drive electric, alternative fuel, and high-efficiency vehicles will pay a $50 registration fee. This is designed to capture more revenue from vehicles that use “less traditional gasoline.” Other states are looking to offset lost gasoline tax revenue with fees that range from yearly registration fees, to a per-mile tax, or even placing monitoring devices in these cars.

It doesn’t take a crystal ball for states to realize that the growing use of cloud computing impacts tax revenue. Hardware and software sales are avoided in favor of using on-demand services, and that creates problems for state and local governments that want to keep tax revenues stable.

In reaction to the growth of cloud computing, most states have placed a tax moratorium on leveraging public clouds to allow the market to grow. Of course, now that cloud computing is in the midst of dramatic growth, all bets are off.

According to the National Law Review this week:

“On June 30, 2013, the Vermont sales tax moratorium on remote access to software expired. At that time, the Vermont Department of Taxes (Department) reverted to its prior position that interpreted, without any analysis, the Vermont sales tax to apply to prewritten software that was ‘licensed for use and available from a remote server.’ Recently, the Department released draft regulatory language relating to the taxation of remotely accessed software and is currently seeking comments on the draft (due by October 1, 2014).”

The legal jargon is even more confusing.

“The draft regulations provide a great deal of guidance, some good and some bad.  On the positive side, the regulations recognize that a sale cannot occur unless ‘use or control [is] given [to] the purchaser with respect to the software’ such that ‘the purchaser is able to use the software to independently perform tasks.’  This language comports with established legal authorities in the state regarding when sales occur, rather than simply stating that a sale has occurred when software is remotely accessed.”

The draft regulations also provide a list of nontaxable transactions, including.

  • A transaction whose true object is the purchase of a service (to which any transfer of software is merely incidental).
  • Sales of data processing and information services.
  • A transaction where the seller processes the purchaser’s data on the seller’s software.
  • A transaction where the customer runs its own software on the seller’s hardware in a cloud computing environment (such arrangements are commonly referred to as Infrastructure as a Service (IaaS), and the draft regulations refer to them as such).

For those in the cloud computing space, the issues and questions around this new law, and, I suspect, other state laws that are forthcoming, are rather obvious.

First, assuming that the taxable event occurs at the point of consumption, in the case of SaaS, this means that the state is not taxing transactions out of their jurisdiction. So, cloud users need to track point-of-consumption locations, as well as duration of use. That will make those who work via WiFi on flights across the US a complex tax problem.

Second, using cloud-based resources is significantly cheaper than traditional hardware and software. Thus, the tax revenue will be down for most states, even when attempting to tax cloud-based services. This could drive taxes up even higher, and discourage the use of cloud-based resources. This would push the emerging cloud computing market back a few steps.

Third, figure that not all states will tax cloud services, so businesses looking to consume cloud services may figure out a way to do so, in certain areas, and thus avoid taxation. Just as business work in states that have lower or no taxes, they will dial emerging cloud taxes into states that are more cloud computing friendly.

Fourth, why the focus on SaaS and not IaaS? If I deliver infrastructure to support a hosted application or software applications over the Internet, I still deliver value. I’m not sure why one event is taxable (SaaS), and one is not (IaaS). What about hybrid-cloud and multi-cloud computing, where many types of providers are leveraged, and a single application may leverage services from as many as a half-a-dozen public clouds? Good luck tracking taxes on those complex cloud architectures. The lawmakers seem to assume that we always deal with a single cloud computing provider at a time, and that’s almost never the case these days.

Finally, if the burden of tax collection for cloud services rendered is placed on the cloud computing provider, it’s not a huge leap that many providers won’t do business in certain states. Or, more likely, they will fight it out in court as Amazon and other “e-tailers” did in December of 2013.

“The high court, without comment, turned away appeals from LLC and Inc. in their fight against a New York court decision forcing them to remit sales tax the same way in-state businesses do. This could hurt online shopping in that state, since one of the attractions of Internet purchasing is the lack of a state sales tax, which makes some items a little cheaper than they would be inside a store on the corner.”

Figure the same kinds of legal dust-ups will occur once these taxes are rolled out, and better known. By all accounts, the cloud computing industry, as a whole, could grow to well over $100 billion dollars a year. The stakes are getting higher for public cloud consumers and providers, as well as local and state governments.

If these taxes do become more of a burden, I’m sure that cloud brokers and cloud management systems will arise, but also minimize the taxes that have to be paid. Also, more cloud services may move off-shore to avoid emerging tax burdens, and thus the revenue will move off-shore as well.

The right answer is that the cloud industry gets ahead of this by providing guidance to state and local government as to how to approach the issues of taxes on cloud services. Otherwise, we’ll be in for many legal battles, providers will move from state-to-state or off-shore, and there will be some very confused cloud consumers. Hopefully, we’ll avoid that situation.

Comments are closed.