States’ interdependence, globalization, expansion of scope and authority of international organizations, and the growth of the global civil society have increased concerns about the use of power on the world stage. The way authority is defined and understood is linked to the notion of legitimacy. Contrary to power, an authority claims a right to rule and needs to be legitimate. Indeed, subjects of authority must recognize the authority as legitimate and competent to accept and submit to the authority. In Hannah Arendt’s words, authority in this sense is characterized by an “unquestioning recognition by those who are asked to obey; neither coercion nor persuasion is needed.”
Keohane developed six specific criteria to define the legitimacy of global governance institutions and whether they are beyond or below the legitimacy threshold.  First is the minimal moral acceptability of institutions that should not perpetrate (or persist in perpetrating) acts that violate basic universal human rights such as physical security, freedom, and the right to subsistence. The second criterion is inclusiveness: institutions need to be open to all who wish to participate in achieving the goals of the institution. This means that some peoples should not be arbitrarily excluded from global governance discussions. 
The third criterion is epistemic quality, which means that global institutions need to be transparent (provide the necessary information to all and at a reasonable cost) and have an institutional integrity (relationship between performance and the truth). For instance, institutions based on beliefs that are false such as racism become ipso facto illegitimate: “(…) it is that the dispersal of information among a plurality of external epistemic actors provides some counterbalance to informational asymmetries favoring insiders.”
The fourth criterion is the compatibility with national democratic governance mechanisms and that they enhance democracy at national level: they can either make it more difficult for special interests to operate. They can help minority and individual rights. They can foster collective and open deliberations within countries.
Fifth, global governance institutions need to pass the test of comparative benefit and produce better results than in their absence. Benefits can be either substantive (security, welfare, nature conservation) or procedural (ability to cooperate with different cultures and societies).
The final criterion is accountability: it is basically the power of the ruled over the rulers. It includes three elements: standards that must be met by the rulers, information available on the performance of rulers, and the ability to impose sanctions on rulers. All institutions are accountable to another actor or organization: funding States, audit firms, NGO and IO members, individual supporters etc.  In other words, it means that “some actors have the right to hold other actors to a set of standards, to judge whether they have fulfilled their responsibilities in light of these standards, and to impose sanctions if they determine that these responsibilities have not been met yet.”
Accountability implies a relationship between the rule-maker and the rule-taker where the standards for accountability and the authority of each party are recognized as legitimate.  It sanctions unauthorized or illegitimate exercise of power and decisions that are judged as unwise or unjust by accountability holders. Accountability is not relevant only when authority has been delegated. Otherwise it would mean to overlook NGOs, States, subunits of States and TNCs accountability. Accountability should indeed be understood differently at national and global levels.
Grant and Keohane define seven accountability mechanisms in world politics. First, hierarchical accountability describes the bureaucratic organization where superiors can remove subordinates from office when violating a rule. This accountability mechanism takes place within an institution. It is in place within the United Nations, the World Bank group or any other global large bureaucracy.
Second, supervisory accountability designates relations between organizations where one can hold another one accountable. For instance, the IMF or the World Bank are held accountable by States and institutions within States. Third, fiscal accountability describes mechanisms where funding agencies monitor performance and can sanction other organizations such as NGOs or IOs. Fourth, legal accountability refers to the legal requirements power-wielder need to follow. Any public official for instance can be held accountable to for their actions by administrative and criminal courts. Fifth, market accountability designates the organizations that are held accountable to investors and consumers. It is the case of TNCs or countries, which can be sanctioned by investors (or consumers) because of a violation of human rights or the pollution of natural resources. Sixth, peer accountability defines the mutual evaluation of two organizations: NGOs evaluate the content and scientific data from other NGOs for instance. Finally, public reputational accountability is a form of soft power where the general public can hold accountable public or private entities.
When examining the accountability of various non-State actors in global environmental governance, international organizations hold fiscal, supervisory, reputational, and hierarchical accountability: NGOs fiscal, supervisory, reputational, and hierarchical accountability; TNCs domestic, legal, market, hierarchical, and reputational accountability. For States on the other hand, it depends on their power. Weak and dependent States have fiscal and supervisory accountability often to donor States, the IMF or the World Bank. Independent States and great powers have peer and reputational accountability.
Under Keohane’s list of accountability indicators, global actors such as States, NGOs, IOs and TNCs are accountable to multiple entities. As a global public or a global citizenship as such does not exist, global institutions have developed other accountability mechanisms. Indeed, global actors cannot be held accountable in a similar manner as national entities for that particular reason.
As discussed, accountability needs standards, sanctions and information. Thanks to new ICTs, information is becoming widely available in an affordable way, which can definitely improve global accountability toward smaller actors such as NGOs or individuals: “The cost of providing information through web sites are now so low that it is difficult to use cost or inconvenience as an excuse; people around the world are increasingly used to being able to get the information that they want almost instantaneously.”  This largely increases transparency, which improves accountability.
 Grant, Ruth W., Keohane, Robert O. (2005) Accountability and abuses of power in world politics, American political science review, Vol.99, Issue 1, p.29.
 Arendt, Hannah (1970) On violence. Boston, MA: Houghton Mifflin Harcourt, p.45.
 Keohane, Robert O. (2011) Global governance and Legitimacy, Review of International Political Economy, Vol.18, Issue 1, p.99.
 Ibid, p.101.
 Keohane, Robert O. (2011) Op Cit, p.102.
 Ibid, p.103.
 Keohane, Robert O. (2011) Op Cit, p.103.
 Grant, Ruth W., Keohane, Robert O. (2005) Op Cit, p.29.
 Ibid, p.30.
 Ibid, p.36.
 Ibid, p.36.
 Ibid, p.36.
 Ibid, p.37.
 Ibid, p.40.
 Ibid, p.42.