Tag: cost estimate

What if We Ran a Public School System… and No-One Came?

The New Jersey Office of Legislative Services, which estimates the budgetary impact of proposed laws, has just released its analysis of a private school choice bill called the “Opportunity Scholarship Act.” The most remarkable thing about its report is the amount of money it assumes that districts would save for each student they no longer have to teach: $0.

On that assumption, if every student were to leave for the private sector tomorrow, districts would keep right on spending exactly the same amount they spend today. Inefficient though it is, not even state-run monopoly schooling is that bad.

The OLS report does not explain why it assumes that the per pupil savings for students leaving public schools (the “marginal cost”) would be $0. It states that this figure is “indeterminate,” but by not counting it at all is effectively treating it as zero.

In fact, the marginal cost of public schooling is not “indeterminate” at all. Economists “determine” it all the time, and it’s quite easy to do. You simply observe how district spending actually rises and falls with enrollment, using a time-series regression, as I did in 2009 to calculate the marginal cost of public schooling in Nevada (see Appendix A).

Even if the NJ OLS does not conduct a marginal cost estimate specific to New Jersey, they could have done–and should still do–the next best thing: take the marginal cost estimates for other states as a rough guide and estimate the NJ district savings from them. I estimated that Nevada district spending falls by 85% of average per-pupil spending when a student leaves, and Grecu and Lindsay, a couple of years earlier, estimated the figure at 80% for South Carolina.

If they want to be conservative, the NJ OLS could use the lower of these figures, and perhaps also run the numbers for estimates 10% higher and 10% lower.

Any of the above options is preferable to the logical impossibility of their current analysis, which effectively treats the marginal cost of public schooling as $0.

The Senate Bill Would Increase Health Spending

Ezra Klein quotes the Congressional Budget Office’s latest cost estimate of the Senate health care bill when he writes:

“CBO expects that the legislation would generate a reduction in the federal budgetary commitment to health care during the decade following 2019,” which is to say that this bill will cover 30 million people but the cost controls will, within a decade or so, leave us spending less on health care than if we’d done nothing.  That’s a pretty good deal. But it’s not a very well-understood deal.

Indeed, because that’s not what the CBO said.

First, the CBO said the “federal budgetary commitment to health care” would rise by $210 billion between 2010 and 2019 under the Senate bill.  Then, after 2019, it would fall from that higher level.  And it could fall quite a bit before returning to its current level.

Second, the “federal budgetary commitment to health care” is a concept that includes federal spending on health care and the tax revenue that the federal government forgoes due to health-care-related tax breaks, the largest being the exclusion for employer-sponsored insurance premiums.  If Congress creates a new $1 trillion health care entitlement and finances it with deficit spending or an income-tax hike, the “federal budgetary commitment to health care” rises by $1 trillion.  But if Congress funds it by eliminating $1 trillion of health-care-related tax breaks, the “federal budgetary commitment to health care” would be unchanged, even though Congress just increased government spending by $1 trillion.  That’s what the Senate bill’s tax on high-cost health plans does: by revoking part of the tax break for employer-sponsored insurance, it makes the projected growth in the “federal budgetary commitment to health care” appear smaller than the actual growth of government.

Third, the usual caveats about the Senate bill’s Medicare cuts, which the CBO says are questionable and Medicare’s chief actuary calls “doubtful” and “unrealistic,” apply.  If those spending cuts don’t materialize, the “federal budgetary commitment to health care” will be higher than the CBO projects.

Fourth, Medicare’s chief actuary also contradicts Klein’s claim that the Senate bill would “leave us spending less on health care than if we’d done nothing.”  The actuary estimated that national health expenditures would rise by $234 billion under the Senate bill.

And really, Klein’s claim is a little silly.  Even President Obama admits, “You can’t structure a bill where suddenly 30 million people have coverage and it costs nothing.”

Our System of Government Exists to Prevent This Kind of Thing

The Hill’s Congress Blog asks, “Will the Senate pass a health care reform bill before it adjourns for the year?”

I answer:

It’s not looking good – nor should it.

The Reid bill becomes less popular with each passing day.  (So too does President Obama’s handling of health care.)

CBS News is reporting that Reid wants to hold a vote before Christmas because he doesn’t want senators to go home and hear from their constituents.

Reid has been systematically suppressing a complete cost estimate of his bill.

Reid’s manager’s amendment will make unknown, countless, and dramatic changes to that 2,074-page bill – and Reid wants to vote on it before anyone knows what those changes are.

Even Max Baucus admits that not a single senator understands the Reid bill.

Our federalist system, the separation of powers, our bicameral national legislature, six-year terms for Senators, staggered Senate elections, and the Senate’s procedural rules all exist precisely to prevent what Reid is trying to do: ram a sweeping piece of legislation through Congress without due consideration.

ObamaCare Cost Estimate Watch: Day #180

On Day #179 of the ObamaCare Cost Estimate Watch, Sen. Jim Webb (D-Va.) wrote in The Winchester Star of his involvement in the Senate health care debate:

At the start of this debate I was one of eight senators who called on Senate Majority Leader Harry Reid to post the text and complete budget scores of the health-care bill on a public web site for review at least 72 hours prior to both the first vote and final passage. This request was agreed to, affording proper transparency in the process.

On the contrary, as I explain in this Richmond Times-Dispatch oped, Reid did not comply with Webb’s request.

Indeed, a memo recently issued by the Congressional Budget Office suggests that Reid has been working very hard to conceal the legislation’s full cost all along.

Recapping the Costs of the REAL ID Revival Bill

In late July, the Senate Homeland Security and Governmental Affairs Committee passed a new version of PASS ID, the REAL ID revival bill. I’ve posted about various dimensions of it: the national ID question, the politics of PASS ID, whether PASS ID protects privacy, a run-down of the Senate hearing on it, and the inexplicable support of the Center for Democracy and Technology for this national ID law.

Three months later, the committee still has not reported the bill, meaning that the public doesn’t get access to the version the committee passed. (A resolution in the House would require committees there to publish amendments to bills within 24 hours.) But the Congressional Budget Office scored the bill this week. That is often a signal that legislation is on the move.

So it’s a good time to look at costs again. The National Governors Association and the National Conference of State Legislatures both premised their support for PASS ID on the idea that it would reduce costs to states to just $2 billion.

But in July I examined the likely costs of PASS ID and NGA’s cost calculations. To save you a burdensome click, here are some highlights:

But there is reason to doubt [the NGA’s $2 billion] figure. PASS ID is a lot more like REAL ID — the original REAL ID — in the way that most affects costs: the implementation schedule.

Under PASS ID, the DHS would have to come up with regulations in just nine months. States would then have just one year to begin complying. All drivers’ licenses would have to be replaced in the five years after that. That’s a total of six years to review the documents of every driver and ID holder, and issue them new cards.

How did the NGA come up with $2 billion? Maybe they took the extended, watered-down, 75%-over-ten-years estimate and subtracted some for reduced IT costs. (The NGA is free to publish its methodology, of course.)

But the costs of implementing PASS ID to states are more likely to be closer to $11 billion than the $2 billion figure that the NGA puts forward. In just six years, PASS ID would send some 245 million people into DMV offices around the country demanding new cards. States will have to hire and train new employees to handle the workload. They will have to acquire new computer systems, documents scanners, data storage facilities, and so on.

The NGA’s claim of savings from PASS ID is weak. Did the new CBO score change anything?

First, let’s review what a CBO score is. When CBO scores a bill, it reports how a bill will change costs to the federal government. Other CBO reports may include overall costs for federal programs, but when CBO scores a bill, it just reports the difference between current federal spending and spending if a new proposal should pass. CBO sometimes mentions mandates on states and private-sector costs in their bill-scores, but those are rarely if ever thoroughly reported. CBO’s wheelhouse is federal spending, and that’s what it reports.

Now, let’s look at how CBO has done with estimating the costs to states from implementing federal national ID standards.

Its first cut at scoring national ID standards was when it looked at H.R. 10 in the latter stages of the 108th Congress. (This was before REAL ID — H.R. 10 was an early version of the bill that became law as the Intelligence Reform and Terrorism Prevention Act.) When CBO scored H.R. 10 in late 2004, it lumped national ID standards along with several other policies and programs in a category called ”Mandates With no Significant Costs.”

Four months later, in early 2005, CBO scored the REAL ID Act, which had been introduced early in the new Congress. It found then that the national ID standards Congress had put into law in December had changed from a mandate with “no significant costs” to a mandate costing more than $100 million.

CBO thought REAL ID would only cost $20 million more than that, an amount below the reporting threshold of the Unfunded Mandates Reform Act, so CBO did not do a thorough analysis.

Then the folks actually faced with implementing it took a look at REAL ID. More realistic estimates of costs to states in the $10+ billion range came forward, including an estimate from the National Governors Association, as I discussed in my previous post on costs.

With that background we’re ready to look at the CBO score for PASS ID. CBO makes no precise estimate of costs to states. Its specialty, again, is federal spending. But it makes a few observations about such costs:

  • “The bill would require states to issue public notices about their security and privacy policies that include information about how personally identifiable information is used, stored, accessed, and shared.”

This, all should agree, is a complex problem but a small cost.

  • “The bill also would require states to have a process that would allow individuals to access, amend, and correct their information. Information from groups representing state governments [NGA and NCSL, most likely] indicates that most states currently have such policies and procedures, though some may need to be revised.”

No. They. Do. Not.

As I said in my post on the privacy consequences of PASS ID:

This is a new and different security/identity fraud challenge not found in REAL ID, and the states have no idea what they’re getting themselves into if they try to implement such a thing. A May 2000 report from a panel of experts convened by the Federal Trade Commission was bowled over by the complexity of trying to secure information while giving people access to it. Nowhere is that tension more acute than in giving the public access to basic identity information.

No state has opened its driver databases for review and correction by the public. That would be an all-you-can-eat buffet for identity fraudsters. The CBO has been bamboozled about state policies.

But the language of PASS ID finesses this, doesn’t it? It says that opening up identity data and giving the public correction rights would be done “as determined appropriate by the State.” So states wouldn’t really have to do anything, right? Right!

Except that the Department of Homeland Security gets to interpret what that language means, and a court will defer to any reasonable DHS interpretation. That’s the Supreme Court’s Chevron doctrine. (It’s an unfortunate abdication of power to administrative agencies, but it’s the law today.)

If the NGA and NCSL have told their clients that they will have the last word on how PASS IS is implemented, they are wrong. It’s DHS’ call — not states’. There may be huge costs to states — hidden at first, but growing and growing — if they stick their heads into the jaws of the federal lion.

Returning to the CBO’s assessment of state costs:

  • “The bill would repeal the requirements of the REAL ID Act and replace them with more flexible requirements for issuing compliant driver’s licenses and identification cards.”

This is true in some respects, and not in others. As I noted before, PASS ID is on a tighter implementation schedule which is the main driver of costs.

  • “The bill also would authorize appropriations that could be used to pay for those requirements, and it would prohibit the federal government from charging fees to states to access the SAVE and SSOLV data systems.”

Because it’s federal, this is something that CBO actually knows about, and its assessment is that PASS ID would dole out a total of $123 million to states over the next five years. Washington, D.C.’s highest spending year would be fiscal 2013, in which it would spend $39 million, less than $1 million per state.

And those savings when the federal government doesn’t charge states for using its databases? Just $2 million each year in fiscal 2010 and 2011.

Nothing in the CBO estimate changes the conclusion that implementing a national ID would cost states over $10 billion dollars, as they hired new staff, acquired new equipment and systems, and marched 250 million Americans through their DMVs. The federal government is promising to dole out $123 million and offer states a whopping $4 million in savings on data access.

The National Governors Association’s argument that PASS ID reduces costs to states is ludicrous. And the paltry funds Congress might share with states is a drop in the bucket. The homeland security appropriations bill for fiscal 2010 cuts funding for REAL ID by $40 million from its 2009 funding level. PASS ID would fare no better.

State governors and legislatures that have fallen for the PASS ID cost estimates of the National Governors Association and National Conference of State Legislatures should fire these financial advisors. NGA and NCSL are trying to grow federal power at the expense of state coffers.

Baucus Bill Would Cost More than $2 Trillion

Sen. Max Baucus’s (D-MT) health care overhaul would cost more than $2 trillion.  It would expand the deficit.  But he has carefully and methodically hidden those facts – so well that he has completely hoodwinked nearly all the major media.

The media are reporting that the Baucus bill would reduce the deficit by $81 billion over 10 years.  Wrong.

The Baucus bill assumes that Congress will allow the “sustainable growth rate” cuts in Medicare’s physician payments to occur beginning in 2012.  Yet Congress has routinely and repeatedly blocked those cuts, making Baucus’s assumption preposterous.  The CBO handled the issue delicately, but essentially said, “Sure, provided that the sun rises in the west in 2012, then yes, this bill would reduce the deficit.”

That means Baucus will come up at least $200 billion short on the revenue side, making his bill a budget-buster.

The media are reporting that the Baucus bill would cost just $829 billion over 10 years.  Wrong.

As Donald Marron observes, that number omits as much as $75 billion in new federal spending.  It also omits a $33 billion unfunded mandate on state governments.

But the worst part is that the Congressional Budget Office’s preliminary cost estimate omits the cost of the private sector mandates in the Baucus bill.  In Massachusetts, those costs accounted for 60 percent of the total cost of reform.  That suggests the actual cost of the Baucus bill – $829 billion plus $75 billion plus $33 billion, times 2.5 – is well over $2 trillion.

Yet the CBO score pretends those costs aren’t even there.  It’s like a mystery novel that’s missing the last 50 pages.  And the media aren’t even curious.

In the words of Brad DeLong, why, oh why, can’t we have a better press corps?

Cross-posted at Politico’s Health Care Arena.

NYT Nonsense on SAFRA

With the Student Aid and Fiscal Responsibility Act (SAFRA) likely to be voted on by the full House or Representatives today, the media is finally giving some space to debate over the bill. Unfortunately, the New York Times only pays attention to the parts it likes, writing in an editorial today that:

The private lenders and those who do their bidding in Congress have recently taken issue with a Congressional Budget Office analysis that showed that the bill would save about $87 billion over the next 10 years.

They argue, absurdly, for example, that the savings would be smaller if the system were analyzed under accounting rules other than the ones that the federal government is required to use. The aim is to mislead taxpayers and members of Congress into believing that the C.B.O. estimate is dishonest.

Um, excuse me New York Times, but the CBO has never said the bill – not just going from subsidized to direct lending, but the whole bill – would save $87 billion over ten years. Moreover, it has been a series of analyses from the CBO – albeit driven by requests from members of Congress – that have continually increased the cost estimates for SAFRA. (I have linked to all the CBO analyses here.) CBO’s very first estimate of the bill’s likely net cost put it at around $6 billion over ten years, and it only went up from there after incorporating such things as lending risk and potentially higher Pell grant costs.

Of course, the Times isn’t alone in its refusal to talk honestly about SAFRA. Despite all of the CBO estimates, yesterday U.S. Secretary of Education Arne Duncan said SAFRA would give college students and numerous other interests the world without costing taxpayers a dime.  “We’re not asking the taxpayers for one single dollar,” he said. And SAFRA’s sponsor, Rep. George Miller (D-CA), has been touting his bill as a revolutionary money saver since day one.

The truth on this thing is out there, but it’s definitely not in the New York Times.