> ISTM if that were true, then investment in enforcement would not show such a large immediate profit. Doesn't that by definition demonstrate that there is low-hanging fruit?
You're talking about two different things. One is what you can get directly by auditing people, the other is the deterrent effect from doing more audits. The second has nothing to do with how profitable the first is. It may have no effect at all, if ordinary taxpayers aren't even aware of how many audits are happening. It could easily have a negative effect for people who are paying attention to it, because they respond by spending more on accountants, which increases overhead (negative externality) and is at least as likely to uncover deductions they're not taking but could than deductions they are taking but shouldn't (no increase and potential decrease in government revenue).
As for the first, consider how the "profit making" nature of audits actually works. The IRS spends thousands of dollars auditing many people who haven't done anything wrong, and causes the same people to have to spend thousands of dollars themselves to deal with the audit -- or more, if the principals of a business get bogged down dealing with it, it can easily cost the business more than it costs the IRS. That's all pure overhead and leads to nothing.
Then one of the audits out of many actually finds a mistake, and it's a business with a few million in annual revenue that has unintentionally underpaid its taxes by 2%, but has been making the same mistake for years. That nets the IRS something like half a million dollars, having spent 20% of that to find it, and caused the audited businesses to spend a further 50% of that on accountants and lawyers. So you create $3.5M in overhead to collect $5M, costing you $1M and the taxpayer $2.5M, but that cost is a deduction to the business, which costs you a further $.875M. So you've created $3.5M in overhead to collect $3.125M for actual programs. That is an extremely bad level of efficiency -- more money is going to overhead than programs. But they call this a 5:1 ratio of profit to expense -- which is still pretty bad (20% overhead) -- because they don't count the cost to the business nor the cost to the government when the business takes that cost as a deduction.
Meanwhile all of the businesses that were audited, even the ones that weren't doing anything wrong, had to go out and hire those accountants, who probably found a bunch of deductions some of them weren't taking even though they could. So the next thing that happens is that the treasury takes in a million a year less every year going forward because they induced a bunch of businesses to spend more on tax accounting. That isn't an inefficiency, they're entitled to it under the law, but it's something to keep in mind for the people who only care about government revenue and not the overhead created on the business side.
You're talking about two different things. One is what you can get directly by auditing people, the other is the deterrent effect from doing more audits. The second has nothing to do with how profitable the first is. It may have no effect at all, if ordinary taxpayers aren't even aware of how many audits are happening. It could easily have a negative effect for people who are paying attention to it, because they respond by spending more on accountants, which increases overhead (negative externality) and is at least as likely to uncover deductions they're not taking but could than deductions they are taking but shouldn't (no increase and potential decrease in government revenue).
As for the first, consider how the "profit making" nature of audits actually works. The IRS spends thousands of dollars auditing many people who haven't done anything wrong, and causes the same people to have to spend thousands of dollars themselves to deal with the audit -- or more, if the principals of a business get bogged down dealing with it, it can easily cost the business more than it costs the IRS. That's all pure overhead and leads to nothing.
Then one of the audits out of many actually finds a mistake, and it's a business with a few million in annual revenue that has unintentionally underpaid its taxes by 2%, but has been making the same mistake for years. That nets the IRS something like half a million dollars, having spent 20% of that to find it, and caused the audited businesses to spend a further 50% of that on accountants and lawyers. So you create $3.5M in overhead to collect $5M, costing you $1M and the taxpayer $2.5M, but that cost is a deduction to the business, which costs you a further $.875M. So you've created $3.5M in overhead to collect $3.125M for actual programs. That is an extremely bad level of efficiency -- more money is going to overhead than programs. But they call this a 5:1 ratio of profit to expense -- which is still pretty bad (20% overhead) -- because they don't count the cost to the business nor the cost to the government when the business takes that cost as a deduction.
Meanwhile all of the businesses that were audited, even the ones that weren't doing anything wrong, had to go out and hire those accountants, who probably found a bunch of deductions some of them weren't taking even though they could. So the next thing that happens is that the treasury takes in a million a year less every year going forward because they induced a bunch of businesses to spend more on tax accounting. That isn't an inefficiency, they're entitled to it under the law, but it's something to keep in mind for the people who only care about government revenue and not the overhead created on the business side.