Title IX Case Study

Administrative & Regulatory State Outline

Prof. Rick Hills, Spring 2007


Part I  •  Legislative & Administrative Process: Title IX as Case Study


  1. Intro to the Process & Legislative History
    1. Hypo from Title 9: coed teams judged only on merit, leads to 90% male. Doesn’t look like ‘discrimination’ ‘on the basis of sex’ (instead, on strength, etc). So is it compliant?
    2. Methodology
      1. START always w/ the text – dissect carefully [‘on the basis of’]
      2. Ordinary language – dictionaries [‘on the basis of’ means ‘b/c of’]

a)      Will almost always find ambiguities that need resolving

b)      Here, ‘b/c of’ could mean ‘intentionally’ ‘foreseeably’ ‘w/ effect of’ etc

  1. Canons of construction
  2. Purpose of the statute – when ambiguous

a)      Legislative History

Ÿ  What key lawmakers said about it (sponsors are the only ones who read it)

Ÿ  Earlier versions – what they disapproved of, how changed

Ÿ  Committee Reports

Ÿ  Hearing testimony (less reliable, unelected)

Ÿ  BUT be careful – silence doesn’t say much

b)      Broad social purpose generally (what was going on at the time: war, marches, etc)

c)      Some discount purpose or legisl. hist. as not reliable (Textualists)

Ÿ  Polls are not laws even if show exactly what Congress thinks

Ÿ  The Constitution prescribes processes for laws, and only laws have force

  1. Look to similar statutes – precedents interpreting analogous statutes

a)      Congress often Xeroxes from older statutes, so relevant to purposes of Congress

b)      Here, Title 9 is basically Title 6 w/ sex & education added, and Title 6 was xeroxed from Title 7—so what Title 7 said then is relevant, interpreted similarly

  1. Agency actions or interpretations (defer or not, more accountable at least)
  2. Conflicting sources
    1. Sources can conflict, won’t be obvious what the choice is, and the hierarchy isn’t exact
    2. Lawyers must be clever & creative in resolving the conflicts, bring to their side
    3. Silence – How bills can die (see later section on inaction)
      1. Hard to pass bills generally – need more than a majority to do anything

a)      House – need the support of the Rules Committee to get anything done

b)      Senate – need unanimous consent agreements to get anything done

  1. Any committee chair can kill a bill by doing nothing, w/o giving reasons
  2. Filibuster & vetoes
  3. Conference committee: what comes out is what you get, can only approve 100% or reject
  4. So doesn’t usually say much – crap arg
  5. Implementation by the AgencyThe Office of Civil Rights (OCR)
    1. Javits Amendment
      1. Delegated power to OCR to implement Title 9 in athletics
      2. Specifically exempted beauty pagents, scouts, frats, etc (called ‘amending statutes’)
      3. Technically it only gave OCR authority to make regulations, not letters, interps, etc
    2. 1975 Regulation
      1. 10 factors in part c, starts w/ “effectively accom. the interests/abilities of both sexes”
      2. Confusing, and seems to go beyond the statute, since having NO teams would violate
    3. 1979 Policy Interpretation – just an explanation published in the Fed. Reg.
      1. Three-part test – independent ways of complying

a)      Substantially proportionate: 50% women means 50% slots (not $) (or 38%/38%)

b)      History/continuing practice: If prong 1 not met, basically that moving toward it

c)      Full and effective accom.: If you fail prongs 1 and 2, show interests/abilities of the underrepresented sex have been fully and effectively accommodated

  1. Hills says probably the test is in trouble if tried to pull this w/ Title 6 or 7 and race
  2. 1996 Clarification Letter: don’t have to go for prong 1, could just run surveys for prong 3 and you’re fine (but if women ever can support a team and are interested, need to give them a team)
  3. Judicial Decisionsupholding the agency view
    1. Cohen I & II:Brown argued its amounts to quotas and leads to cutting only men’s teams (discrim)
      1. Three-part test doesn’t violate anti-quota provision of 20 USC 1681(b) b/c NOT req’d – prong 1 is a quota system, but can also comply by prong 3 w/o proportionate numbers.
      2. Your BUDGET leads to cutting men’s teams, not req’d, could just add women’s teams as nec.
    2. Kelley: Cutting only men’s teams is fine b/c still disproportionately represented, doesn’t violate 1681(a)
      1. Hills thinks looks like violation just looking at Title 9 itself, they were cut b/c they’re men
    3. Cohen IIREJECTS the analogy w/ Title 7 and statistical imbalances, despite huge similarities
      1. Gender is an actual difference here in qualification, unlike for most jobs
      2. Qualified applicant pool is w/in the school’s control by picking sports (‘we need fast runners’)
    4. Reasons for agency deference
      1. Accountable to Congress, who specifically gave them authority (unlike the courts)
      2. Democratically accountable, represents Pres. who is elected
      3. BUT arg that Congress could have overruled if didn’t like is not great (silence, etc)
  4. Review of Three-Part Test
    1. Deference for 3-part test was controversial b/c it didn’t have to go through notice-and-comment crucible.
      1. Don’t want an agency to bypass 553 to get the same thing w/ an informal stmt.
      2. Some say, don’t give Chevron deference, give Skidmore deference as punishment for bypassing
    2. Argument FOR deference: it’s longstanding as an interp – agency and courts have relied on it, it’s old
      1. So NOW, as a persuasive interp, reasoned decision, published in the register, issued by a single admin agency, on which courts and private parties have relied. That’s the arg at least.
    3. What if they just wanted to change it?
      1. State Farm supports the view that if an agency doctrine is in place, might not be able to just get rid of it, even if not a rule as such. BUT we want agencies to be politically accountable..
      2. B&W position that it’s so longstanding that subsequent legisl. ratifies it is rare, odd, not normal.
      3. Normally just trying to req the agency to explain why it’s reversing. The earlier stmt at least has persuasive force. BUT NOTE that courts might not give the new version deference then.

a)      A lot will depend on WHY they gave the old one deference.

  1. Don’t need to give a concise general stmt unless it’s 553, not ‘proponent’ and all



Part II  •  Statutory Interpretation


  1. Intentionalism and Purposivism
    1. Rector, Holy Trinityà Even clear language can be trumped by contrary intent of the lawmakers
      1. Expressio Unius: Minister is clearly covered by statute – says “labor or service of any kind” and then lists five exceptions (artists, singers, lecturers, etc) & minister isn’t on the list
      2. Ord lang: court says ‘labor or services’ means ‘manual’ (then why have those exceptions?)
      3. INTENT: real driver here – uses Comm. Reports, title of the Act, contemp. events, evils trying to remedy, “we all know” this wasn’t what Congress was trying to prevent
      4. Absurdity – b/c of the country’s general religiosity, blue laws, etc wouldn’t prohibit ministers
    2. Problems w/ Intentionalism
      1. Are they discovering the intent or making it up & then justifying it?
      2. Looks like judges act on their own moralities/interps and answer to no one
      3. Laws are enacted, intentions are not
      4. Gives Congress the wrong incentives, to pass lazy or sloppy laws knowing they’ll be corrected
      5. Pre-empts Congress’ laws if the courts get the intent wrong
    3. Absurdity Canon
      1. Can disregard plain language if it leads to an absurdity, BUT everyone agrees, need the absurdity FIRST – Can only bring in legisl. once absurdity is found, not to find the absurdity
      2. Public Citizen v. DOJ: What is absurd? [Do ABA/GOP/NAACP need to be open, minutes..?]

a)      Broad view: Brennan

Ÿ  Congress is very unlikely to pass or intend

Ÿ  Results that are very likely to be unconstitutional (Savings construction)

b)      Narrow view: Kennedy

Ÿ  Not conceivable for even one sane Congressperson to apply it that way

Ÿ  Unconstitutional does not mean absurd, it means unconst.

  1. Savings Constructions: Should construe statutes to avoid unconstitutionality

a)      CONS:

Ÿ  Disregarding plain text b/c might be unconstit. means can depart from text w/o explaining or justifying that power; creates a kind of shadow constitution

Ÿ  Danger of absurdity canon and savings constr. – can rewrite the statute

Ÿ  Not good judicial reasoning – end up disregarding a lot & relying on little

b)      PROS:

Ÿ  Gives Congress respect, implies it would not have enacted something unconst.

Ÿ  Prudent, since rulings on Const. are sticky, dangerous precedents, just avoid

Ÿ  This method is routine and perfectly fine when the constit. problem is clear and uncontroversial

  1. Plain Meaning/New Textualism
    1. RULES:
      1. Can always use canons to massage the text and make it clear (‘skilled user’ of words)
      2. If the text (+ canons) gives unambiguous result, CAN’T look to purposes or legisl. hist.
      3. EXCEPTIONS: — but note, they are very loath to find these

a)      If the text is actually ambiguous after canons, CAN use legisl. hist. to break the tie

b)      If it leads to absurdity, can reject the text (odd or weird doesn’t count)

  1. à Textualists usu. only use legisl. hist. as a check, to see if Congress meant to be absurd
  2. à Textualists believe ambiguities should be resolved by ord. usage and general principles

a)      Textualists DO use substantive canons, as well as intrinsic aids

b)      In evaluating these sources, test them against the 4 principles below

  1. PROS & Counters
    1. Constitutional Legitimacy: Only statutory text is voted on by Congress & signed by Pres.

a)      Why can they invoke dictionaries and ‘what we mean’ but legisl. hist. is so taboo?

  1. Judicial Restraint: Narrow the sources of interp. to narrow judicial power & discretion

a)      BUT plain meaning still allows weird value judgments, like in Caminetti

b)      When there are ambiguities, something needs to be done that isn’t ridiculous

  1. Predictability/Simplicity: Reduce the costs of researching the law. More sources obfuscates.
  2. Good Ex Ante Effect: Gives Congress good incentives to be clear, write statutes w/ care
  3. Caminetti(1917): Early ex à If text is clear, you may go no further
    1. Mann Act said ‘prostitution…or any other immoral purpose’ (Could mean anything)
    2. Ejusdem Generis: If you have specific examples followed by a general class, you interpret that ‘catch-all’ class to mean only similar items. Implies must involve sex, and that’s it
    3. Non-marital sex is immoral based on cultural norms, so the plain language is clear
    4. Results: Mann Act prosecuted women and lovers even though not illegal to do it
    5. Reconcile w/ Holy Trinity? Said court found ‘absurdity’ (Implies minister not let in is more absurd than prosecuting couples – they say, not absurd, cultural norms)
    6. Green v. Bock Laundry: Rule 609 says ‘to the D’:  Absurd, past convictions let in only against Ps
      1. Scalia finds absurd – looks to legisl. hist., but not helpful. Adds ‘criminal’ in front of ‘D’

a)      Common usage: lots of ppl say D when mean criminal D (less violence than adding P)

b)      Shrink don’t expand: modifier there shrinks meaning, better, means more likely meant

c)      General legal env’t: pro-D bias in protections usu. (not same as legisl. hist.)

d)      Ex Ante Incentives: hatred of Comm. Reports, even tho voted on – b/c ppl could stock it w/ little jewels to sway courts that would never pass if it were in the bill

  1. Mortier: Does use of ‘state’ pre-empt cities from regulating?
    1. Hills’ textual arg: elsewhere it says ‘or political subdivision’ so if ‘state’ doesn’t exclude subdivisions, then you’re rendering that phrase mere surplusage
    2. Scalia simply says, the terms of the statute don’t pre-empt clearly (and it’s a big deal)

a)      He’s right about majority: they dig up mountains of legisl. hist. that doesn’t show their point, really just smokescreen for not liking pre-emption, don’t need it

  1. Majority plug! This is case where maj. says we’re still going to use legisl. hist., it’s the law!
  2. Marshall: LSD ‘mixture’ case. If paper is mixture =  wildly harsh sentences. (What about OJ???)
    1. Expressio Unius: elsewhere distinguishes btw mixture and pure, so if construe ‘mixture’ to include pure, rendering that section mere surplusage. (So then what does mixture mean?)
    2. Ord. usage: chooses a def. of ‘mixture’ based on ord. usage (‘can’t be sep. out’) BUT it isn’t the only one – this is chronic textualist problem, pick a usage among many for no reason
    3. Odd results are NOT a reason for rejecting a given definition when there are competing ones
    4. Ex. of absurd result that isn’t insane or unconstit. enough for textualists
    5. *Frozen into the law. Congress can’t enact a law now that punishes heroin dealers less.
    6. Hinds County School Bd: Ex of convenient use of textualism as a tool to craft a just result
      1. Probably ‘sex’ was a typo, same-sex schools were ok, but court sees the racial motive here
      2. But note that still cited for idea that same-sex schools prohibited, bad precedent
      3. Textualism makes it hard to correct these statute-copying errors






– Fuller picture, more info

– More likely to yield just results

– Avoids irrationality/absurdity

– Intent is hard to determine

– They often don’t have an intent

– Legis. hist. is not the law

– Can pick and choose what supports

– Uncertainty in Congress w/r/t interp.

– Lots of power/discretion to judges




– Constit. legitimacy (only text is law)

– Seems more objective, less discretion

– Better incentives for Congress

– ‘Clear’ language is a value judgment

– Unfair/unforeseen results

– Inflexible and arbitrary at times

– Text can be an accident, no remedy

– Choose btw ord. uses w/o explanation

– Congress can’t always remedy results


  1. Canons of Construction – Intrinsic Aids
    1. Expressio unius: the expression or inclusion of one thing implies the exclusion of others
      1. Surplusage: Words should always be construed to do some work
      2. AND/OR rule: put ‘and’ or ‘or’ for a reason (also note ‘any…or’)
    2. Noscitur a sociis: meaning of unclear words or phrase should be determined by the words around it
    3. Ejusdem generis: in a list of specific items followed by a general class, the general class is deemed to include only items of similar nature as the specific ones.
    4. TIPS:
      1. Always start w/ the text itself & look at ordinary usage or plain meaning (commas, grammar)
      2. Try to think of any possible ways it could be construed, then look at flipside args
      3. For ejusdem & noscitur: what possible meanings could those other words share?
      4. Watch the COMMAS and the way they can divide up clauses
      5. Do any of the words have technical definitions, as well as ord. uses? (Like ‘fraud’ or ‘D’)
  2. Substantive Canons – looks to a value outside the statute as an interpretive aid where ambiguous
    1. Rule of Lenity: In criminal statutes, read an ambiguous statute or provision narrowly, in favor of a D.
      1. McNally v. US: Politician got kickbacks, ‘deprived ppl of right to honest gov’t,’ but not illegal.

a)      Canons all on one side:

Ÿ  ‘Any’ scheme or artifice to defraud ‘OR’ deprive of money or prop.

Ÿ  Similar statute was interpreted to include cheating out of honest services

b)      But majority says ambiguous, and applies rule of lenity

  1. Sup. Ct. said gun in trunk of car was ‘carrying’ and that bartering a gun for drugs counts as ‘using a gun in a drug transaction’ – not using rule against lenity b/c NOT ambiguous, they say

a)      Note that ambiguity is itself ambiguous, makes the rule’s use hard to predict

  1. Advantages:

a)      Always favors a politically unpopular group, so Congress will quickly correct if nec.

b)      Ensures Ds have NOTICE of criminal penalties

c)      Sep. of powers concerns – Congress should decide if something is criminal, not courts

  1. Textualists like this rule, incentives for Congress and all. But won’t use if text is clear enough.
  2. Canon of Constitutional Avoidance:
    1. NLRB v. Catholic Bishop:

a)      Expressio: ‘All’ + 8 exceptions, no others. Looks unambiguous on plain meaning.

b)      Had to go to the legisl. hist. to find ambiguity, and finds it in silence, never addressed.

Ÿ  Strong req’ts: want to see Congress addressing schools AND relig. freedom

Ÿ  Would never do this except for the constit. avoidance need

c)      The avoided issue is, don’t want NLRB ruling on matters of Catholic doctrine, but later, Smith says if the law is neutral, who cares whose beliefs it tramples

Ÿ  Smith may be better rule, since if it runs roughshod over religion, it’s for Congress to decide, better at balancing, accountable, force them to do it

  1. Never get to the q though – does this solve these problems best? (see other pros/cons above)
  2. Canon of Avoidance – avoiding burdens on federalism
    1. Gregory v. Ashcroft: State law forcing judges to retire at 70 looks like discrim. under ADEA

a)      Canons go against majority:

Ÿ  Noscitur: all the exceptions that aren’t covered by ADEA are policy-making execs, like immediate advisors, personal staff, etc. Judges not an exception.

Ÿ  Ejusdem: exceptions construed narrowly after the general class.

Ÿ  Surplusage: if judges are policymakers, who isn’t? Are prosecutors?

b)      Majority puts judges in the exception anyway, b/c of concerns about federalism

Ÿ  Law generally deprives states of ability to make their own stds (like for game wardens) BUT sense that judges are more important than other jobs, burden

  1. Giving a constitutional VALUE a wide berth, b/c don’t want to step on states’ rights

a)      Idea is that Congress shouldn’t take states’ rights lightly, so construe narrowly absent clear evidence that Congress intended to do so

  1. Extrinsic Aids – what else should influence interpretation?
    1. Legislative Inaction or Rejection
      1. General RULE: Silence is a crap argument. (see p.1)
      2. If NOT crap, 3 things tend to be true:

a)      Not a failure to enact, but an actual floor vote

b)      Someone actually discussed the issue on the floor, an actual floor debate on it

c)      The thing rejected very specifically was the opposite of the interp. the court is giving

  1. North Haven Bd. of Ed. v. Bell: à This is IDEAL ex. of when inaction is actually significant

a)      Issue is whether a school janitor has been denied benefits ($) of an educational program under Title 9 – essentially an employment issue, wouldn’t think covered

b)      Failed floor amendments were introduced saying DIDN’T cover – BUT:

Ÿ  Bypassed committee, voted down on the floor. This means the whole House by majority vote rejected the amendment.

Ÿ  And they did accept some amendments to Title 9 then like frats, paid attention

Ÿ  AND Bayh actually got up and said it was orig. meant to cover employment (but note that not wanting ‘no’ emps covered, doesn’t mean want ALL emps)

  1. Reading Statutes in pari materia
    1. Idea is that closely tied statutes should be read as one big statute together
    2. North Haven Bd. of Ed.: Title IX is basically a copy of Title VI, w/ ‘sex’ instead of ‘race’

a)      Doesn’t cover employees:

Ÿ  Title VI doesn’t cover employees, and it’s basically the same thing

Ÿ  Titles VI and VII were passed same time, shows Title VII is only emp. one

Ÿ  Why enact strict procedures for suits in VII if could bypass through VI or IX?

b)      DOES cover:

Ÿ  Title VI had specific exclusion for employment, and Title IX doesn’t, even though it was orig. there in xerox, was cut out specifically in committee

Ÿ  Expressio in VI and absent in IX is exclusio alterius in IX (in pari materia)

  1. Rule Against Implied Repeals
    1. Two canons:

a)      When ambiguous, later statutes should be read NOT to repeal earlier statutes

b)      When ambiguous, narrower statutes should govern over general so both can coexist

  1. General competing rules:

a)      Usually later statutes govern over earlier, if clear (55mph vs 35mph – later governs)

b)      ‘Cardinal’ rule: if at all ambiguous, implied repeals are disfavored, try to keep both

  1. Morton v. Mancari:

a)      The two statutes:

Ÿ  1972 Amend. extended anti-discrimination laws to gov’t employees

Ÿ  1934 statute had given preferences in BIA jobs to Indians. (Was it repealed?)

b)      Plain lang.: discrim based on race? Yes (could try to say ‘nation’ but weak)

c)      BUT unanimous opinion in Congress since 1834 that Indians have self-gov’t rights, special preferences in BIA – no one has ever objected, have re-aff’d for 100+ yrs

d)      Construe ambiguous laws to not be overruling an entire C/L background norm



Part III  •  Political Supervision of Agencies


  1. Intro to Agencies
    1. What are agencies?
      1. APA §551 – Civilian unelected executive officials, who exercise authority of the gov’t
      2. Usually mirror separation of powers w/in them – tasks/branches kept sep., tho all exec.
      3. Big challenge: ~ 30K ppl in Congress and courts are supposed to supervise 2 mil. in agencies.

a)      They hold enormous power, and these two little bodies take ‘watching’ seriously

b)      US courts enforce laws against agencies in civil court (Marbury v. Madison)

  1. Each agency has a governing statute and an area
  2. Why do we have them?
    1. Historically, ballooned in size in 20th century, from collecting taxes and handing out gov’t land to protecting ppl from unsafe products, unfair Ks and work conditions, prohibiting discrim. and protecting the env’t
    2. Took over C/L stuff courts did (torts, Ks) – so WHY BETTER?

a)      Specialization/expertise: Courts are generalists, need Ph.D to say causes cancer, etc.

b)      More accountable: agencies created in the New Deal b/c courts were conservative, distrusted unions, thought you assume the risk of cancer at your job, etc.

Ÿ  Legisl. no longer trusts the courts, gov’t ballooned to replace those fxns

c)      Broader rules: Courts rule case by case, but agencies can look at what remedies and rules are best for all interests, rule more clearly all at once

  1. What do agencies do?
    1. Issue rules
    2. Adjudicate (ALJs)
    3. Other – everything else, massive amount (cleaning things, building highways, being the FBI)
  2. Political Supervision of Agencies by Congress
    1. The Non-Delegation Doctrine – Congress’ constitutional duty to supervise agencies
      1. The idea: Delegation allows Congress to enact a goal and then not support the hard policy choices that are nec. to carry it out. Big policy stuff should be decided by Congress. (Schoen.)
      2. Generally, 3 ways agency can have uncabined power:

a)      Too much power in narrow area (do anything in an area)

b)      Narrow discretion but in a huge area (only wages, but for every job)

c)      Or where power is given to the recipients (regulated do the regulating)

  1. Sources

a)      Textualist: Const. Art. I § 1: “All legisl. powers…shall be vested in a Congress.”

Ÿ  Arg is weak – doesn’t say ‘only’ in a Congress, as long as they supervise

b)      Precedential:

Ÿ  Early stmts: Intelligible Principle Rule that okay to delegate as long as there’s some guidance, something the courts could review

Ÿ  Recovery act cases:

(a)    Yakus upheld, more limited, just prices

(b)   Schechter struck down, whole economy

c)      Structural:

Ÿ  Democratic accountability – Congress does policy, visible, traceable

Ÿ  Notice – need to know the legal std ahead of time

Ÿ  Judicial review – need statutory rule or measure to make review possible

Ÿ  NOTE that the last two are compatible w/ the agency making its own rule, which was rejected in Whitman

  1. General content:

a)      CAN’T delegate ‘legislative power’

b)      à Which means, must have an intelligible principle

c)      à This principle MUST be promulgated by Congress (Whitman v. Amal. Trucking) BUT it doesn’t have to be specific (Whitman)

Ÿ  “fair and equitable” or “requisite to protect health” is fine

  1. Vagueness: There’s no principle to say how much discretion is too much. The RULE is basically anything less than Schechter. It can’t be used when it’s this vague. So courts no longer strike down.
  2. Canon of Construction: Courts today don’t strike down laws, they narrowly construe the law to avoid the vague value. DO still use it like this.

a)      Benzene case: OSHA’s statute says protect health, so w/ Benzene they figured 10ppm was ok, but 1ppm to be safe, no reason why 1 and not 0 or 9. HUGE price difference, not even sure if it will save any lives, seems arbitrary.

Ÿ  Best arg would be to interpret ‘lowest feasible’ to mean cost-benefit-wise

Ÿ  BUT that entails putting a value on human life and weighing against costs, and that’s too much power, so the nondelegation canon means they avoid it

Ÿ  Instead, Stevens’ test is just show “significant risk” at 1ppm, meaning just some benefit from a 1ppm std. (That’s re-writing the statute, ‘signif. risk’)

(a)    Saying pick a number and give a better expl., said one life saved, fine

b)      Destroys the purposes of the nondelegation doctrine

Ÿ  Value of it is to force Congress to take up the qs, debate it, make the std clear

Ÿ  Re-writing the statute to avoid the issue reduces the incentives to do that

  1. PRO-DelegationCongress CAN’T micro-manage
    1. Legislative veto:

a)      Used to retain a legisl. veto over certain actions (deportation) for control reasons

b)      Chadha RULE: Congress cannot veto or enact laws binding outside of Congress w/o vote of 2 houses + Pres. (Internal congressional rules are different.)

c)      They delegated power w/ this process, need to repeat it if they want to step back in

d)      RESULTS:

Ÿ  Single-house resolutions don’t mean anything outside Congress anymore

Ÿ  Congress just makes specific laws, and limits funding (‘provided no money..’)

  1. Congressional removal of executive officers:

a)      Bowsher: Congress can’t delegate power to anyone to implement the laws and still retain control of them (like being able to hire or fire them). Must delegate AWAY.

b)      RESULTS:

Ÿ  Bad: Congress has less control over its own legislation, a LOT harder to police

Ÿ  Good: Congress isn’t very good at implementation, multi-member body

Ÿ  Now they just require the agency heads to come to hearings and yell at them

  1. Political Supervision of Agencies by the President
    1. Presidential power of appointments:
      1. Art. II, § 2, cl. 2: Power to appoint all non-inferior officers. Must be appt’d by the Pres. w/ the advice and consent of the Senate.
      2. Congress can vest appt of inferior officers in ppl other than the Pres. (but Congress can’t)
      3. What does non-inferior mean? Not sure, just top of the food chain, not order-followers.
      4. Big picture: Pres. still has control, by commanding the boss. The inferiors follow orders.
    2. Removal powers
      1. Major Precedents

a)      The Decision of 1789: à Really important precedent, since framers

Ÿ  Pres. has implied power of removal, don’t need ‘removable by the Pres.’ lang.

Ÿ  Congress cannot share in that power (e.g. can’t require A&C of the Senate)

b)      The Tenure of Office Act: Pres. refuses to fire treasonous heads, so Congress enacted the TOA, but Pres. refuses to obey. Wasn’t impeached, b/c TOA was unconst. b/c:

Ÿ  Took Pres.’ implied power away (Principle A) OR

Ÿ  Congress can’t limit the Pres.’ power to needing A&C of Senate (Principle B)

Ÿ  These two are restmts of the two above. Distinction btw taking and limiting.

Ÿ  Ex: A law saying ‘You serve 6 yrs, and cannot be fired by Pres. or Congress’ violates Principle A but not Principle B.

c)      Pendleton Act: Civil service exemptions on Pres. power to hire-and-fire. Congress can insulate inferior officers from Pres. removal.

  1. Myers and Perkins: Congress limited firing, so q is: is a postmaster a non-inferior officer?

a)      Clear rule: If appt’d by Pres., Pres. can fire, for any or no reason. If not appt’d, then Congress can limit. Simple.

b)      Broad rationale: ‘Take-care’ clause: Pres. has a duty to see that the laws are faithfully executed, needs to be able to remove ppl in order to do his job properly.

  1. Humphrey’s Executor: FDR elected, wants to fire head of the FTC. Court strikes down, partly b/c New Deal stuff all getting struck down.

a)      Under Myers, he’s appt’ed, can be removed, should be simple.

b)      Rationale: FTC is quasi-judicial and quasi-legislative. (Holds hearings, issues a report)

c)      Rule: Pres. can fire purely executive officers. Otherwise Congress can limit.

d)      Results:

Ÿ  *Congress can create independent agencies whose chief heads cannot be removed by the Pres.

Ÿ  For when want law impartially enforced or interpreted à applies equally to courts (quasi-jud.) and to agencies dealing w/ economy, like the Fed.

Ÿ  How decide? Weiner says, look to the nature of the function of the agency

  1. Power to Control Agencies through Executive Orders
    1. Youngstown Sheet & Tube v. Sawyer: Pres. exec. order was the last resort, needed to avert the strike b/c wartime, took over the steel factories. Knew it was distasteful, but what to do? Told Congress to reverse him, come up w/ better idea, tell him what to do, and never responded. Then steel industry sued.

a)      Jackson’s more reasoned concurrence is only law insofar as it meshes w/ Black’s short majority opinion. (see diagram)


Non-Delegation Doctrine

Congress consents (Broadest powers)

Zone of Twilight

Inner core: Art. 2 powers

b)      Presidential powers diagram

Ÿ  Outer limit: defined by Art. 1 (non-del.)

Ÿ  Inner core: defined by Art. 2 § 2. Pres. always has these powers regardless of Congress. Outside of this core, Congress can forbid or grant him power.

Ÿ  Congress consents/delegates: then Pres. has power Congress gives him

Ÿ  Zone of Twilight: He has his normal powers, but Congress silent, doesn’t forbid but doesn’t consent either.

c)      Black’s analysis: NOT granted by Congress, and NOT w/in core powers (b/c this is inside the country, not the theater of war). BUT does he recognize the zone of silence? Black sometimes seems like w/o core or consent, no power.

Ÿ  Fight meanwhile btw Jackson/FFurter & Vincent: all recognize this zone, but issue is whether this is zone of silence occasion. The former say no, Congress basically addressed.

Ÿ  Bad arg though, since voted down doesn’t say much. ‘Can’t do x in peacetime’ doesn’t tell you anything about wartime.

  1. Instructions for Exec. orders (coming out of Youngstown as best we can)

a)      If have a clear federal statute that prohibits what Pres. trying to do à need an Art. 2 core power that entitles him to resist Congress, appeal to higher authority for the act.

b)      If Congress authorizes the Pres. to act, and Pres. seems to be outside those limits à then can make a non-delegation arg. (as opposed to an art. 2 arg)

c)      If Congress is silent on it à then need to use all trad. stat. interp. tools to discuss, take sort-of-silence and find implicit consent or implicit prohibition, not silent.

d)      If Congress is really silent and can’t find implied anything à look to traditions of Pres., gloss of history, norms. If long trad., infer Congressional acquiescence, that although Congress could change that, he can act until they do.

e)      BUT w/ takings of private property, even if tradition, suspicious

Ÿ  ‘Contemporary imponderables’ – zone sits uneasy. This section is like Korematsu – court will allow, but ISN’T his ‘power’ or ‘the law’ it’s just that no one will stop him in the circs. Fine, but try to avoid.

Ÿ  RESULT: If Pres. goes outside regulating the gov’t and into lives of private citizens, court will vindicate. But if Congress jumps in and backs the Pres., then courts back off (consent zone). So powers are more likely to be construed narrowly in the zone of twilight.

  1. Exec. Orders & Agency ‘Cost-Benefit’ Analyses

a)      Exec. order (Reagan and all since) req’s ALL agencies to send their plans to OIRA (subdiv. of OMB) w/ findings that the costs are outweighed by benefits, even though the agency’s statute might tell them not to take into acct cost-benefits in plans.

Ÿ  Sanction is not being published in the Fed. Reg. (if fail to do this or found to be faulty) à Under APA, can’t be enforced if not published.


Ÿ  Look to the specific statute (if that helps, I’m not sure it does)

Ÿ  Art. 2 § 2: Pres. can always require agency heads to give opinions as much as want on any subject. So argue only asking for more info on costs/benefits.

Ÿ  Prioritizing vs. not enforcing – fuzzy line

(a)    Pres. has discretion to prioritize, can say something is low priority if doesn’t want to enforce it, but could argue he has duty to enforce laws

(b)   Agencies can disregard what Pres. says if think it’s obeying the law

(c)    Closet veto (not ok) vs. priorities will be based on if it’s AT ODDS w/ the law – so depends on whether looks like priority or veto.

(d)   He always has power to fire, but yet still CAN’T fire for enforcing the law (like refusing to prosecute black ppl more). (Difficult lines.)



Part IV  •  Judicial Review of Agencies’ Actions


  1. Constitutional Review: Procedural Due Process
    1. OVERVIEW:
      1. STEPS always:

a)      Whether DP applies at all (whether int. is type protected by 5th/14th Am: liberty/prop.)

b)      What process is due (weighing of interests)

  1. Sources of proc. DP:

a)      APA is source for agency process

b)      Constit. DP is the safety net in case statutes or APA fail to provide for it properly.

Ÿ  5th Am. and other Bill of Rights apply to states via 14th Am.

Ÿ  But very hard to change these rules, so their scope is limited by the courts

  1. Limitsto procedural DP
    1. The two limits:

a)      Character of the proof and number of the parties

b)      Nature of the interests

  1. Distinction btw rulemaking and adjudication

a)      Londoner: Guy gets assessed (by an agency) for his road being paved w/o a hearing

Ÿ  Character of the proof:

(a)    Decision about road paving is legislative b/c look at city planning etc

(b)   BUT decision about the tax for that paving is involves indiv. facts – how much prop. he owns, amount that abutts, etc. (Adjudicative)

b)      Bi-Metallic: Everyone’s home in Denver will be assessed at 40% higher

Ÿ  Note that state legisl. decided it, not an agency

Ÿ  Number of the parties: large number of ppl are similarly affected

Ÿ  Character of proof: Relevant facts don’t pertain to the specific indiv. cases

(a)    Looks like indiv. facts w/r/t whether your home under-valued—but NO, issue is THIS law, not the assessor, and it doesn’t single you out.

(b)   Can only argue that policy-wise, it wasn’t true that assessor undervalued. Don’t have right to do that. (Check on power is voting.)

c)      SUM:

Ÿ  Any decision by a legisl. to enact a statute is considered legisl. and NOT subject to DP, don’t get a say in it. (BUT do local gov’ts count..?)

Ÿ  Legisl. decisions by an agency (like gen. rules governing large numbers of ppl) are also considered to be legisl. and NOT subject to DP.

Ÿ  Hills says if affected only one party, would prob. be adj. even if looked general, but hard to say (if they published notice though, then maybe..?)

  1. What DP REQ’S:
    1. STEP 1: Whether the int. is protected – liberty or property under the 14th Am.

a)      Roth: To be a prop right, must have a legitimate claim to entitlement (legally-recongized expectation). NOT entitled to a gov’t job/re-hire on a 1-yr K.

b)      Sindermann: BUT need DP whenever there is a legitimate claim to entitlement, regardless of whether this person turns out to be entitled or not

Ÿ  Series of 1-yr Ks, even tho no formal right to re-hire, P could point to manual about informal tenure as creating enough of an expectation for prop. right

  1. STEP 2: How much process is due: Balancing test

a)      Goldberg v. Kelly: Q is whether need hearing before welfare benefits terminated.

Ÿ  Step 1: Welfare benefits are property – an entitlement by statute if qualified, so a taking of them is adjudicative of rights

Ÿ  Step 2: Weighing P’s int/potential harm against gov’t int in not losing money

(a)    The crucial feature here is that welfare is essential to living, outweighs

(b)   But hearing doesn’t have to be full-blown, can be informal, minimal

b)      Mathews v. Eldridge: The balancing test

Ÿ  Weighing three concerns:

(a)    P’s interest in the decision

(b)   Risk of wrongful decision

(c)     int. in providing less process (fiscally, recouping payments, delays)

Ÿ  If P’s int. ´ risk > Gov’t int.  à  then need more process, safeguards


a)      Possibly too general to be useful, say some, should look at history/custom of agency

b)      *Point is, Court is trying to control agencies given that they deal w/ specific qs w/o any guidance from the statute or const. (how much notice here? rely on expertise? etc)

c)      Only imposes outer limits, but not detailed, elaborate limits or guidance

d)      This is the minimum process for agencies, aside from the APA

  1. Statutory Review by Courts: An Overview of the APA
    1. Purpose/history
      1. FDR set agencies up to get around courts’ review – but then wielded unchecked power
      2. APA was result – passed to police the admin state through the use of the courts
    2. Rules and Orders – definitions are in § 551
      1. Rules: the key is future effect. But they can have ‘general or particular applicability’ so in theory could apply to only one person & still be a rule. (Don’t confuse w/ Bi-Metallic.)

a)      Note that licensing is an exception to the ‘future effect’ stuff, see 551.

  1. Orders: This is everything else. There is NO definition of order, it’s a residual category.

a)      Includes hearings about whether x did y (past-focused)

b)      And includes things you don’t think of as adjudicative, like closing down a highway.

  1. PROCESS under the APA & other agency statutes
    1. Rules:

a)      If says ‘hearing’ à 553. Needs to say ‘hearing’ and ‘decision on the record’ for formal rulemaking to apply. 554 tells you when formal applies, and 556-57 tells what it req’s.

Ÿ  Formal rulemaking is too costly, so courts construe 553(c) very narrowly

Ÿ  VAST majority is notice-and-comment (553) rulemaking

  1. Orders:

a)      Trigger words are technically ‘hearing’ and ‘on the record’ too, but court construes very broadly, jealously guards 556-57 process even w/o one or both these words

Ÿ  This is b/c orders are about specific parties, affecting them more than rules do

Ÿ  VAST majority of adjudication is formal

b)      Savings Construction:

Ÿ  Courts are big on 556-57 process as a savings construction for APA for DP

Ÿ  In contrast, don’t even need 553 process for informal rulemaking as a matter of Const. DP, as long as general application. (Bi-Metallic, etc)

  1. NOTE that APA never trumps or repeals other statutes, it’s a default rule. So if the statute doesn’t even say ‘hearing’ then APA doesn’t even apply, no process at all (except Const.).




FORMAL: §§ 554, 556-57









  Must mention BOTH ‘hearing’ AND ‘on the record’ [Narrowly construed]   All other rulemaking (553 notice-and-comment)

[Note: Vast majority of rulemaking]




  Mention EITHER ‘hearing’ OR ‘on the record’ OR strong constit. interest in proc. DP [Broadly construed]   All other orders (ZERO process at all)



  1. Adjudicative Factfinding Review – ‘Substantial Evidence’ in the ‘Whole Record’
      1. APA § 706(2)(e) req’s factfindings to have ‘substantial evidence on the record.’
      2. This is judicial review for formal hearings regarding factfinding.
      3. Applies in 2 circs:

a)      Where 556-57 applies (see above, says ‘hearings’ etc)

b)      OR where the underlying statute says ‘hearing’ or ‘record’ BUT the statute disclaims 556-57 and they don’t apply, b/c they’ve been waived. Can’t shift that waiver to 706.

  1. What does substantial mean?

a)      This phrase comes from directed verdicts, the same as the std to overrule jury verdicts

b)      Admin agencies get the same deference as juries do (and that’s a lot) and more deference than for trial judges (whose std is ‘clear error’).

c)      *It really means ‘the evidence isn’t insubstantial

  1. Whole record?: Note in 706, that it means anything put in briefs or said (if leave out, ok)
  2. Universal Camera: ALJ doesn’t believe Chairman, but NLRB reverses, says didn’t believe other guy.
    1. Hand: Doesn’t treat ALJ as factfinder b/c he cites the APA (first place to go!) at § 557(b) that the agency is the factfinder, reviews de novo, has all the powers of the ALJ at review.
    2. Ffurter reverses: Doesn’t say how much deference, but have to at least look at the ALJ report.

a)      It’s a mandatory part of the record, says multiple parts of the APA

b)      Review is based on the whole record, says 706, so it must be part.

  1. Hand on remand: Says ALJ report must be some considerable deference as to credibility or believability b/c cold record, that stuff is ‘lost in print’

a)      This only applies when credibility is the primary issue. (Don’t have to care if ALJ believed the doctor who said you can get heart attack from black lung, if you can’t)

  1. SUM: Univ. Camera à ALJs should usu. get deference on the issue of credibility

a)      Agencies can still NOT defer, but have to say ‘implausible based on how things work so prob. lying’ rather than ‘we think she’s lying b/c of the words she used’

b)      à BUT point is, agencies and courts can’t just ignore it

  1. Allentown Mack: Ppl re-interviewing for their jobs say no one supports the union, so he takes a poll. NLRB says unfair labor practice, not objectivereason to believe this.
    1. The NLRB’s std is ‘good faith reasonable doubt’ but these stmts are clearly enough à CAN’T actually enforcing a higher std than enunciated, like compelling need, etc.

a)      Can’t use deference of factfinding ability to change the std (‘not met by these facts’)

b)      Oversight: If agency could smuggle in a policy under a different label, Congress can’t hold them accountable. (And ALJs apply the law inconsistently too.)

Ÿ  Non-delegation arg: agencies need to make the std clear and be counted for it

  1. à When looking for subst. evid., it will be w/r/t the policy as officially announced.
  2. SUMMARYof substantial evidence rule:
    1. [Is the agency’s view of the statute correct? Remember, substantial evidence of what?]
    2. Agency must consider the whole record, say something about everything in briefs, can’t ignore.
    3. Hearing officer given substantial weight regarding credibility.
    4. Agency’s decision must support their actual std, not some hidden std it would rather support
    5. Burden in the APA (absent overriding statute) falls on the ‘proponent’ and the agency cannot by regulation change the burdens. Means person/agency/org trying to change the status quo.
    6. Zhen Li Iao v. Gonzales: Posner and the Falun Gong immigration case. ALJ disbelieved her about persecution by picking apart beliefs and testimony. Looks like not deferring on credibility??
      1. ‘Reluctance to make clean determinations of credibility’ – wants ALJ to do it right, say that he doesn’t believe her (demeanor, etc) if he doesn’t. Instead, claims std not met by these facts.

a)      ALJ is actually worried about millions of Falun Gong coming in, or pissing off China.

  1. Like Allentown, ALJ needs to enforce the actual std, not a more strict one.
  2. Exercise of Discretion Review – ‘Arbitrary and Capricious’ Test
      1. APA § 706(2)(a): ‘arbitrary and capricious, an abuse of discretion, or otherwise not in accordance w/ law.’ (AC/AD/NIAWL)
      2. This is the std of judicial review of any informal agency actions – the catch-all residual category of actions that aren’t exactly rules or adj. (giving out Ks, where highways go, etc.)
      3. Applies?: Basically this one for when you don’t have hearings, and subst. evid. if you do
      4. What does ‘abuse of discretion’ mean?

a)      Need to consider alternatives (as statute mentions or req’s)

b)      Need a record to see the basis of decisions (though not formal!)

c)      Post-hoc rationalizations are not enough, suspicious

d)      *It’s really the same std as ‘subst. evid.’ Sounds like more deference, but isn’t.

e)      SEE below for ‘hard look’ informal rulemaking version (Don’ts)

  1. Result: pushing agencies toward more formal factfindings than req’d b/c of need for review
  2. Overton Park: Have to show ‘no feasible and prudent alternative’ before putting a highway through a park. Challenge has to show a prudent/feasible alternative that Sec. failed to rule on/consider.
    1. FIRST: always start w/ statutory interp., need to know what statute is supposed to mean.

a)      Court says Sec. needs to show ‘unique difficulties’ w/ alternative route, and displacing ppl and houses don’t count, b/c always do that when you don’t use a park.

Ÿ  Expressio: no point in giving parks as consideration if everything else is too.

Ÿ  Stat. interp. DRIVES this whole opinion.

b)      Bad arg: Court reads prudence out of the statute. Sec. can only argue not feasible as defined by court as ‘unique diff.s’ but can’t argue it’s not feasible, i.e. not a good idea.

Ÿ  Ignores aspects of the text (prudence)

Ÿ  Gives very little deference to the agency

  1. Review: Not an adjudication (future) but not a rule, just a decision. AC/AD/NIAWL applies.

a)      Consider alternatives – and that there is a record

b)      ‘Clear error of judgment’ is what court looking for.

  1. Post-hoc rationalizations

a)      They are stmts prepared for litigation by DoJ attys, like briefs and pleadings

b)      Can’t ONLY exist in a brief. But really just need the right person to say magic words.

  1. Formal findings before the fact are NOT req’d (b/c informal) BUT:

a)      Judicial review may require some stmt from the agency about which facts were considered. Basically atty tells an aide to draw some up. But trend then to ratchet up the findings beforehand if going to be req’d later w/ suspicions if after-the-fact.

b)      BUT can be risky if before, can’t ever change position if challenged

  1. Aftermath: rich ppl didn’t want highway, and council didn’t want it destroying a black community. Better to defer to council—instead, white flight and decline b/c put highway around the city. AND legally the council could have paved over the park if wanted.
  2. SUM: Need some kind of record of decisionmaking for review (and ‘clear error’ is ambiguous)
  3. ‘Hard look’ version of review for A&C factfinding
    1. State Farm: Revokes old rule about passive restraints (seatbelts/airbags) b/c won’t be used. BUT relied on study where ppl did use them. NO study can’t trump bad study if rescinding.

a)      Proponent has burden of proof, which here is agency, so has to show reg. is ineffective

b)      Higher std to rescind a rule than failure to enact one. Can always fail to act.

  1. Informal rulemaking AC/AD/NIAWL std for factfinding:

a)      DON’T: rely on factors excluded by Congress (illegal, AC)

b)      DON’T: fail to consider important alternatives (if in the statute, illegal not to)

c)      DON’T: ignore evidence. (Goes w/ not considering alternatives)

d)      DON’T: rely on findings that have no plausible basis.

  1. Legal Interpretation Review – The Chevron Steps
      1. Battle btw the court’s way of dealing w/ ambiguity (stat. interp.) and the agency’s (Chevron)
      2. Chevron v. NRDC:

a)      Clean Air Act: What does ‘stationary source’ mean? Def. says ‘bldg, structure, facility, or installation.’ This lang is ambiguous b/c single bldg or single facility? etc.

Ÿ  They’re trying to get around bulky permitting req’ts for each changed bldg

b)      Bubble interp.: other words are surplusage if just meant bldg. Bubble around single complex means covers bldgs when alone, structures when alone, etc.

c)      Agencies’ point: Performance std if one person can choose how to comply. (Not C&C)

d)      Defers to agency – they should make policy decisions to fill gaps in the statute

  1. TWO Justifications for Chevron Deference

a)      Democratic accountability: Agencies are democratically accountable, so if there’s a gap in the statute, they should do the policy decisions to fill it.

Ÿ  This is the ORIGINAL justification for Chevron.

Ÿ  706 is supposed to be gap-filler, review is for courts. But no tension as such.

b)      Implicit delegation: Congress delegated power to the agency to enforce the law, so there is an implicit delegation to interpret the law as well.

Ÿ  This is Scalia’s preferred intent, that has become the ‘REAL’ justification.

Ÿ  BUT ‘implicit’ delegation trumps the explicit delegation to courts in the APA?

  1. The Chevron Two-Step. ASK:

a)      Step One: Is the statute ambiguous? Need a gap first. This is the hard step.

b)      Step Two: Is the interpretation reasonable? Quite deferential for the most part.

  1. Sampler:

a)      Cardoza Fonseca: à Leave only the gap left AFTER massaged by tools of stat. interp.

Ÿ  Agency treats two statutes (‘life is threatened’ and ‘well-founded fear’) the same, objectively by preponderence.

Ÿ  NO deference b/c NOT ambiguous, one is subj., the other is obj., based on traditional principles of statutory interp.

Ÿ  BUT ‘well-founded’? Sounds like RP std, objective.. Not ambiguous?

Ÿ  Scalia is pissed – says if you can use canons, legisl. hist., etc in Step One to see if ambiguous, there will almost never be a Step Two, never any ambig. left

Ÿ  They say there’s still a gap – deference for what subj. fear means, how measure

b)      Young: ‘Shall make regs to the extent necessary’: Does this mean can make NO regs?

Ÿ  Agency says inaction is discretionary

Ÿ  Court: need to make some regs, though number and such is all discretionary

Ÿ  Better arg, since otherwise ‘shall’ etc is unnecessary.

  1. Step Zero: Chevron’s Scope
    1. Intro:

a)      FIRST: Does Chevron even apply? If you don’t have an implied delegation to say what the law is, then agency is NOT entitled to fill the gaps.

Ÿ  Q for Step Zero: Has this power been delegated by Congress here? Is this the right agency, acting through the right behavior, the way Congress meant it to?

b)      How can you tell?

Ÿ  Christensen: Not agency manuals, opinion letters, policy stmts, guidelines, etc.

Ÿ  Mead: Divides up into clear rule + swamp:

(a)    APA: Informal/formal rulemaking or adjudication under 553-557. Like findings of ALJs, etc. These are easy cases, Congress delegated.

(b)   For all other informal actions (where highway goes, customs rulings), look to a general bundle of things:

(i)     Legal effect/who it binds

(ii)   Source of the action (top, or minor officer)

(iii) Made w/ some level of care (10,000 of these, or deliberated?)

Ÿ  Breyer: if low-level decisions (interstitial) are endorsed by the agency, may defer

Ÿ  KEY really is whether the statute has tagged you to make this decision

c)      Hills says: Process and jurisd. will be important in overcoming Step Zero.

Ÿ  Whether agency doing what supposed to be doing, given by Congress, whether using Congressionally-specified procedure.

d)      NOTE that even w/o Chevron deference, still get Skidmore deference

e)      Misc. rules:

Ÿ  NO deference for an agency interpreting multi-agency legislation, like Title VI.

(a)    EVEN IF they do it through 553 notice-and-comment, doesn’t matter.

Ÿ  Criminal law is not a body of law that DoJ lawyers are deferred to as experts in, about what the crim law should mean. They are advocates, and it’s a sanction.

Ÿ  Sometimes they’ll give an agency deference but not one section of the agency

  1. Christensen: à Opinion letters don’t warrant Chevron deference – they get Skidmore deference.

a)      Fair Labor Stds Act: must be ‘reasonable flexibility’ in paying out vacation time.

Ÿ  Harris Cty is worried, wants to tell police when they can take it so not all at once

Ÿ  Asked Wages & Hours (w/in the Dept of Labor) about it, who said that’s okay as long as it was in the K. If not, no. Harris County is unhappy, so sues.

b)      The q is: Does Chevron apply to the opinion the agency actually wrote? Wasn’t done by rulemaking or adjudication. So what’s the legal force of the letter?

c)      Skidmore: The old case about whether the off-time waiting for fires was overtime or not.

Ÿ  Agency opinion wasn’t binding, but will give it power to persuade based on how well-reasoned it is, how expertise plays in. Like ALJ opinions in Univ. Camera.

Ÿ  Basically this is just something less than Chevron.

d)      KEY: ‘Agency manuals, policy stmts, and enforcement guidelines’ lack the force of law and do not warrant Chevron deference.

e)      WHY doesn’t it apply?

Ÿ  Really problematic, as to why the letters don’t have force of law here:

(a)    Congress never bestowed the power.

(i)     BUT they do have power to write these letters, so why not?

(b)   Too informal. Not using 553 at least, that’s EASY deference.

(i)     So what LESS than 553 gets deference? No guidance.

(c)    No legal consequence for violating them.

(i)     Circular, since if they gave it deference it would have force.

Ÿ  Dept. of Labor CAN issue something w/ the force of law on this issue that would get deference, but not THIS way

(a)    Scalia wants to see something from the top, an authoritative stmt

Ÿ  THIS is the first step in the court’s nervousness about Chevron deference

f)       Breyer vs. Scalia

Ÿ  Breyer wants Chevron as canon more, for when ambiguous. Nervous  about deference for some clerk writing a letter in a basement somewhere.

Ÿ  Scalia thinks Congress intended agencies to do the interp. whenever there are gaps, regardless of policy reasons. Need legisl. authority to talk policy.

  1. Mead: à

a)      Customs tariffs for ‘notebooks, memo pads’ etc. Whether covered means big money.

b)      These are QUICK rulings (is this a skirt or pleated wear?), binding w/in the agency. When reviewed, there’s a presumption for them, but weak. Basically de novo review.

c)      WHY doesn’t it apply?

Ÿ  Two concerns: legal force and more/decent procedures

(a)    Very weak, doesn’t bind many ppl, low precedential value

(b)   Casually issued, 10,000 of them, lots of low-level ppl issuing

Ÿ  Test:

(a)    Some stuff clearly entitled (553, etc)

(b)   Some obviously not (low level, binds almost no one)

(c)    But mushy in the middle, like issue agency-wide ruling but no proc.

(i)     NOTE three-part test from OCR is in the middle here.

  1. Breyer’s ‘interstitial’ things:

a)      Breyer: More for low-level decisions (tho not decisionmakers). Tiny things should get deference, but BIG things worried about, like taking over greenhouse gases or cigarettes. Courts as outer-boundary guards. If agency endorses little decisions, maybe defer.

  1. Step One
    1. Is the statute ambiguous? If it’s clear, there’s no cause for Chevron deference.

a)      Tension btw trad. statutory interp. tools and agency control over the interp.

  1. Babbitt v. Sweet Home: Does ‘taking’ animals include destroying habitat? Endangered Species Act. Def. is ‘hurt, harm, injure, harass.’ Agency says acts AND omissions that actually hurt/kill.

a)      Ambiguous?

Ÿ  Majority: YES.

(a)    Surplusage: ‘harm’ does no work if it just means hurt, injure.

(b)   Legisl. hist.: bird watching that affects breeding counts as take.

(c)    Post-enactment hist.: permits assume ‘incidental’ takings prohibited

(d)   If you can shut down a dam in TVA you can shut down Farmer Bob.

Ÿ  Scalia’s dissent: NO ambiguity and wants to squeeze out Chevron.

(a)    Ord. usage is that ‘take’ means take, clear, but extendable to harass, etc

(b)   Noscitur: Other terms in def. make harm mean AFF’V acts

(c)    Absurdity canon: Absurd to penalize ppl for draining your own pond rather than flooding it one yr. Thinks is it a takings issue, poss. unconst.

(d)   Post-enactment hist. sinks his arg about only direct takings, but right that that could mean anything, not usually helpful (HA- B&W)

b)      Ord. interp. à Ambiguity à Deference

Ÿ  When canons go either way, train wreck, cut through w/ Chevron deference.

Ÿ  MCI & B&W say no, can’t do that.

  1. MCI: à Textualism kills ambiguity b/c of narrow view of purpose, damn the torpedos/policy

a)      Context: Statute says ‘every common carrier shall’ file tariffs BUT FCC ‘may, in its discretion, for good cause shown, modify’ the req’ts. This is a hard industry to compete in, so in the 30s req’d AT&T to file to make sure they couldn’t lower prices to zero to drive competitors out of business. FCC exempted MCI, Sprint, etc b/c all the rage, way to compete, lower costs. No point in filing which is HUGE cost, might be squeezed out.

b)      Q is, what does ‘modify’ mean? Scalia writes the maj. this time, bit of revenge-ish.

c)      Dictionary fight: Ord. usage means small change, not getting rid of altogether.

Ÿ  (Very imp. if say ‘can’t mean x’ to show it could mean smthg, avoid surplusage)

Ÿ  1 of 7 dictionaries says basic/imp. change, Scalia rejects, is aberrational


Ÿ  Chevron seems to say that as long as ambig. (meaning Congress didn’t decide it) should allow agencies to make these policy choices.

Ÿ  Why dick around w/ dictionaries? This is about running a system in a new era. If 1 of 7, text is close, args a wash, why not let agency choose if Congress didn’t?

Ÿ  à Chevron machine is turned off when textualism avoids ambiguity. Kills it.

Ÿ  This is JUST like ‘mixture’ debate. You have a SLEW of poss. meanings, but as soon as Scalia decides one is slightly better than another, it all shuts down, and at that point it has nothing to do w/ policy even if there are mjr policy issues.

Ÿ  So the point is, Chevron’s a tie-breaker, but you need a tie first.

e)      Narrow purpose constraint on deference

Ÿ  The key statutory purpose (not legisl. purpose but like Rector, title, etc) of rate-making statutes is filing rates. If take that away, that’s everything.

Ÿ  BUT Stevens says larger purpose is regulating an industry for the benefit of customers and competition.

(a)    Scalia saying narrowest purpose is regulating prices, damn the torpedos.

(b)   Apparently concerned about limiting exec power, sep. of powers (ha..)

  1. Brown & Williamson: à Ignoring text b/c of Congressional adverse possession

a)      FDA tries to regulate cigarettes for >18 in giant rulemaking, 700K comments.

b)      Text says FDA has authority for drugs and delivery devices. Means ‘intend to affect structure or any function of the body.’ Clearly cigarettes are drugs, must have been intended, b/c so well-known. NO—b/c know is not intend. BUT new info – suits, discovery shows they did actually intend the effects. That’s FDA position. Text covers.

Ÿ  Plain language/textualism means it falls under the statute. Text is ignored here.

c)      Plan A (failed):

Ÿ  Ban argument.

(a)    Looking at whole statute, see that if a product is dangerous, have to ban.

(b)   This is like MCI: About ratemaking, can’t not. Dangerous products must be banned, can’t not. (And Congress said they don’t want them banned)

Ÿ  BAD argument.

(a)    First, the text covers it, literally.

(b)   They ‘may’ ban, they don’t have to. FDA can take into acct policy. The statute’s REAL purpose is all about discretion. Round one to Breyer.

d)      Plan B (failed):

Ÿ  Legisl. actions

(a)    Congress has struck down attempts by other agencies to regulate. Also passed 5 or 6 statutes regulating cigarettes, reserving it for themselves.

Ÿ  Silence/expressio  counter

(a)    Congress told FTC etc explicitly that it couldn’t regulate, but didn’t do that w/ FDA, always actively struck down others before.

e)      KEY is Ratification: The FDA position has been prominent and longstanding so Congress has acted under the assumption that it is not covered.

Ÿ  à If a position is longstanding, gets written into the statute, even if wrong.

Ÿ  If Congress doesn’t like it, can change it. It’s like adverse possession.

f)       Nondelegation: Big, important decision, so discretion can be limited. (Benzene arg)

g)      Political issues: An example of how statutory interp. is easily manipulated and political?

Ÿ  When agency is backed by the Pres., it’s at its most legitimate and most entitled to Chevron deference. Since Clinton backed, a lot of ppl thought the court should defer and leave it to the Pres. to fight it out w/ Congress, proper role.

  1. Step Two
    1. Is the interpretation reasonable?

a)      This is a lot like AC/AD/NIAWL review. 706 comes in IF EVER at this step.

Ÿ  So statute is ambiguous, but STILL can’t be:

(a)    poorly reasoned

(b)   inconsistent w/ prior agency opinion in an unexplained way

(c)    Unexplained, esp. failure to consider alternative readings

(d)   LOTS more things here. Is it careful, reasoned, silly..? etc

b)      Generally this is quite deferential. If you can GET to Step Two, usually defer.

  1. Notice and Comment Rulemaking Review – Requiring More ‘Notice’ and ‘Comment’
    1. Statutory Req’ts
      1. This is the premier method of agency action. § 553 req’s notice and comment. Notice means publication in the federal register (NPRMs – notice of proposed rulemakings). Also need to publish in order to make the rule have force when it’s final and ready to go.
      2. Notice

a)      Need to notify ppl of:

Ÿ  Time/place of any hearings (useless, there aren’t any)

Ÿ  The legal authority (just means 25 USC blah)

Ÿ  Terms/substance of the proposed rule, or descriptions of the issues involved

b)      Notice problem: The issues involved part is the real meat of it, but the problem is, the agency’s view of things changes throughout the process, as comments are submitted.

  1. Comment

a)      Submission of written documents but don’t need hearings.

b)      Need to give a ‘concise general stmt’ of ‘basis and purpose’

  1. Remember that certain agency actions aren’t subject to notice and comment, like interp. rules, general stmts of policy, manuals, guidelines, advisory opinions, no force of law b/c no deference
  2. Nova Scotia:
    1. Salt is enough to kill botulism in whitefish. If req heat, it will destroy the fish and industry.
    2. Process & substance: Agency has a big file, sure, but the q is, did they consider this?

a)      Need concise stmt of basis and purpose – so need to show that they considered important alternatives and why they chose the other. (Bankrupting an industry is important.)

Ÿ  Note that it will be a lot less concise and general.

b)      Looking for stmt like ‘This will destroy the industry, but we don’t care, safety first.’

Ÿ  Lose on process if not there, BUT even if it IS there, they will:

Ÿ  Lose on substance if they don’t state a basis for it, like a study saying can’t salt

c)      *Process is a way to get around the substance of confusing issues and Chevron deference

Ÿ  BUT depends on whether have to start all over or just re-publish.

  1. Start all over?

a)      Remedy for inadequate explanations can be just sending it back for a better stmt.

b)      BUT can’t just send it back for inadequate notice, need to start over.

c)      Test is whether logical outgrowth of the original notice, that’s good enough. Not clear rule, but issue is whether ppl had a good sense of what interests were at stake.

Ÿ  Ex.: If they give notice about heating and then publish about salt, ppl might sue, saying no notice, didn’t comment on studies about salt, etc., start over.

Ÿ  BUT ppl keep coming forth w/ new comments, info, changes things…!

  1. MUST highlight important rival stuff and how used their expertise
  2. Weyerhauser: Agency changed rule from 400-300ish based on comments, pretty arbitrarily, and env’talists say no notice, might need to start over.
  3. Vermont Yankee: à Put a stop to any req’ts above written procedures
    1. Don’t want courts bogged down w/ proceedings like discovery, aren’t req’d by 553, stopped it
    2. ONLY opp. for written comment, no hearings or cross-examinations, though many wanted that
    3. BUT note that the completeness of a written comment is just as contested.
    4. Results
      1. All req’d: 1. Have to give consideration 2. state basis and purpose 3. opp. to comment.
      2. Need issues that will be dealt w/ and substance of the rule. LARGER than looks.
      3. Balloons into a huge, really long stmt and process as you req more explanation, support, notice.
      4. Changes throughout the process, have litigation, no opp. to participate as req’d by 553.
      5. Anticipatory Notices: Agencies start saying, here’s 10 issues we’re looking at. Comment please. THEN they send out NPRM, hope not to get anything too surprising.

a)      That might work but it still lengthens the whole process just as much overall.

  1. WHY does all this matter??
    1. Notice and comment is the primo Chevron deference action. It’s premium partly b/c the process is so onerous, assume ALL data is in and everyone had their say.
    2. Really matters b/c Step Zero is a no-brainer, clearly 553 is automatically in.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s