How Pandemics Spread — and What Stops Them

A pandemic is an epidemic that has crossed borders and kept growing, and whether it does so comes down to arithmetic more than to any single dramatic event. If each infected person passes a pathogen to more than one other person, the count climbs exponentially. If each passes it to fewer than one, the outbreak fades. COVID-19 and the 1918 influenza are the two best-documented tests of that rule, and they show what speeds the numbers up, what slows them down, and why the same disease can look mild in one group and lethal in another.

Key takeaways

• R₀ (the basic reproduction number) is the average number of people one infected person infects in a population with no immunity and no countermeasures. Above 1, an outbreak grows; below 1, it shrinks.
• The effective reproduction number (Rₑ) is what is actually happening today, after immunity, behavior change, and interventions are counted. Control means pushing Rₑ below 1, not removing every susceptible person.
• Infection fatality rate (IFR) counts deaths among all infections. Case fatality rate (CFR) counts deaths among confirmed cases, so it runs higher whenever testing misses mild infections.
• Timing matters as much as transmissibility. A disease that spreads before symptoms appear is far harder to contain than one that spreads only when people are visibly sick.
• Vaccines, surveillance, and early reporting are the defenses with the largest measured effects, but none of them works at full strength without the others.

How fast an epidemic grows: the R number

Think of an epidemic as a chain letter. Each recipient forwards it to R₀ new people, and each of those forwards it again. If R₀ is 2, one case becomes two, then four, then eight. After 10 generations of spread, one case has become 1,024. If one generation takes about five days, that is roughly 50 days from a single infection to a thousand, and the next 10 generations add another thousand-fold.

The key feature of exponential growth is that it feels slow and then feels sudden. In the early weeks the daily numbers are small enough to ignore. Then the curve bends upward, and hospitals that were comfortable last week are overwhelmed this week. This is why public-health officials talk about acting before the evidence looks alarming, because by the time it looks alarming, the doubling has been running for weeks.

R₀ is not a fixed property of a germ in the way a boiling point is a property of water. It depends on how the pathogen behaves and also on how people behave: how often they meet, how crowded their homes are, how much time they spend indoors. The same virus can have a different R₀ in a dense city and a sparse rural region.

A few published estimates show how the range looks in practice:

Disease or variant	Estimate	Source and note
Measles	Commonly cited at 12 to 18	CDC says up to 9 in 10 susceptible close contacts of a case will become infected, and the virus can stay infectious in the air for up to two hours after an infected person leaves
SARS-CoV-2, early 2020	2.87 (95% CI 2.39 to 3.34)	Pooled summary reproduction number from Billah, Miah, and Khan, PLOS ONE, 2020
SARS-CoV-2, Omicron	Average 9.5 (range 5.5 to 24)	Liu and Rocklöv, Journal of Travel Medicine, 2022; the authors treat the top of the range as a theoretical ceiling

The Omicron row deserves a caution. That 9.5 is an average of eight published estimates made under different assumptions, and the authors note that the highest value assumes no immune evasion. A single tidy number for R₀ is a modeling summary, not a measurement taken off the virus.

Why a single R₀ is slippery

R₀ is an average, and averages hide the shape of transmission. In many outbreaks, most infected people pass the pathogen to nobody or to one person, while a minority infect many. Epidemiologists call the second group superspreading events, and they are the reason an outbreak with a modest average can still explode in a choir rehearsal, a crowded bar, or a hospital ward.

Estimates also depend on what the data can see. If testing catches only a fraction of infections, the early growth rate is distorted. If people change their behavior as news spreads, the number measured in week six is not the number that applied in week one. This is why careful reports give ranges, and why you will see different R values for the same outbreak in different papers.

One more complication is that the formal definition of R₀ assumes nobody is immune. For a new virus that was nearly true in early 2020. For influenza, measles, or a recurring coronavirus, it is not, so what researchers actually track in the middle of an outbreak is Rₑ.

Herd immunity: what it is and what it is not

If R₀ describes how a pathogen spreads in a population where everyone is vulnerable, Rₑ describes what happens as people become immune. In the simplest model, Rₑ equals R₀ multiplied by the fraction of the population still susceptible. When enough people are immune that Rₑ falls below 1, each case produces fewer than one new case, and the outbreak shrinks even though plenty of unprotected people remain.

That threshold is the herd immunity threshold, and in the simplest model it equals 1 − 1/R₀:

R₀	Herd immunity threshold	Example
1.5	33%	Illustrative
2	50%	Illustrative
2.87	65%	Early COVID-19 summary estimate (Billah et al.)
9.5	89%	Omicron average estimate (Liu and Rocklöv)
12 to 18	92% to 94%	Measles, which is why vaccine programs aim for around 95% coverage

The arithmetic is simple, but the term "herd immunity" is used loosely. A review by Paul Fine, Ken Eames, and David Heymann in Clinical Infectious Diseases, titled "'Herd Immunity': A Rough Guide," notes that the phrase has several meanings, including a threshold proportion of immune people at which infections should decline, and the indirect protection of susceptible people by immune neighbors. That looseness matters because the formula assumes everyone mixes at random and that immunity is permanent and complete. Real populations cluster in schools, towns, and social groups, and immunity to many respiratory viruses wanes or can be evaded by new variants, so the clean threshold is a rough guide, not a guarantee.

The overshoot problem

There is a second, less discussed point. Reaching the threshold does not stop an epidemic instantly. It only means the epidemic has started shrinking, and the momentum of the current wave carries infections past the threshold. In the standard model, an outbreak with an R₀ of 2 that is allowed to burn through an unprotected population ends up infecting about 80% of people, not 50%. With R₀ of 3, the final attack rate is about 94%. This is the arithmetic behind a blunt public-health point: relying on natural infection to reach immunity means accepting infection of most of the population, with the deaths that follow.

Vaccination reaches the same immune fraction without that cost, because immunity is built before exposure rather than after infection. For measles, that is exactly how it works. For a pathogen that evolves quickly, such as influenza or SARS-CoV-2, the target keeps moving, which is why vaccines are updated and why one infection or one vaccination does not settle the question permanently. See how vaccines train the immune system for the biology behind that.

The transmission chain: how a pathogen gets from one person to another

An outbreak needs an unbroken chain: a source, a route, and a susceptible host. Interrupting any link stops the chain, which is why different diseases call for different tools.

Source: where new pandemic pathogens come from

Most new human pathogens begin in animals. The World Health Organization describes zoonoses, diseases that pass from animals to people, as making up a large share of newly identified infectious diseases, and lists HIV and Ebola virus disease among them. Fruit bats are thought to be the natural hosts of Ebola, with spillover to people through contact with the blood or fluids of infected animals such as chimpanzees, gorillas, monkeys, and forest antelope. The mechanics of how a virus crosses species are covered in how viruses jump from animals to humans.

Spillover alone does not create a pandemic. The pathogen must also adapt to spread between people efficiently. Many spillovers end in a dead end after a handful of cases, which is part of why the world sees far more small outbreaks than pandemics.

Route: how it travels

Route	Examples	What controls it
Respiratory droplets and aerosols	COVID-19, influenza, measles, tuberculosis	Ventilation, masks, vaccination, isolation
Fecal-oral (contaminated water and food)	Cholera, typhoid, polio	Clean water, sanitation
Vector-borne	Malaria, dengue, Zika	Mosquito control, nets, vaccines where they exist
Direct contact with body fluids	Ebola, HIV	Safe burial practices, protective equipment, contact tracing

The fecal-oral row has the most famous origin story in epidemiology. In 1854, the London physician John Snow mapped cholera deaths around a water pump on Broad Street and argued that the disease traveled through contaminated water, long before the germ itself had been identified. His reasoning is a model of how to find a transmission route from patterns in the data, and it is told in John Snow and the Broad Street Pump.

The respiratory row is the hardest to control, because breathing is not optional and the air in a shared room is shared by everyone in it. Measles is the extreme case: the CDC notes that the virus can remain infectious in the air for up to two hours after an infected person has left a space.

Timing: why incubation and serial interval decide the outcome

Two time measurements decide how hard an outbreak is to stop. The incubation period is the time between infection and the first symptoms. The serial interval is the time between symptom onset in one person and symptom onset in the person they infected. If the serial interval is short, each generation of infections arrives quickly, and there is less time to find and isolate cases before they infect others.

The most important timing question is whether a person is infectious before they feel sick. A modeling study led by Michael Johansson at the CDC, published in JAMA Network Open in 2021, estimated that under baseline assumptions 59% of SARS-CoV-2 transmission came from people without symptoms at the time: 35% from people who were presymptomatic and 24% from people who never developed symptoms. A strategy built on finding people who are visibly ill cannot catch that share of spread.

Ebola sits at the opposite end of the contrast. The WHO gives an incubation period of 2 to 21 days and describes transmission as requiring direct contact with the blood or body fluids of a sick or deceased person, or with contaminated surfaces. Because transmission follows illness and requires close contact, tools such as isolation, contact tracing, and safe burials can work, though that does not make the disease easy to control in settings with weak health systems.

Feature	COVID-19	Ebola
Main route	Airborne droplets and aerosols	Contact with body fluids
Spread before symptoms	Substantial (59% from people without symptoms in one model)	Not the main driver; contact with sick or deceased persons
Incubation	Days (the median is short; exact values vary by variant)	2 to 21 days (WHO)
Average case fatality	Far lower than Ebola	About 50% (range 25% to 90% across outbreaks, WHO)

Severity: what a pathogen does once it infects you

Transmissibility and lethality are separate properties, and they do not move in lockstep. A virus that spreads easily but rarely kills can still produce a large death toll, because a small percentage of a very large number is a large number. A virus that kills half of those it infects but needs close contact with body fluids may burn out in a small area.

Seasonal influenza illustrates the first case. The WHO estimates that it causes around a billion cases a year, including 3 to 5 million severe illnesses, and between 290,000 and 650,000 respiratory deaths. Ebola illustrates the second: an average case fatality of about 50%, but a transmission route that is hard to sustain without health-system failure.

The dangerous combination is high transmissibility with high lethality. The 1918 influenza came closest in modern history, and smallpox is the historical example of a deadly disease that spread by close, lengthy face-to-face contact and by coughing and sneezing. The CDC says about 3 in 10 people with smallpox died. The World Health Assembly declared smallpox eradicated in 1980, which remains the only human disease eliminated by deliberate vaccination, and the story starts with Jenner's first smallpox vaccination.

IFR versus CFR: why the headline fatality rate is often wrong

Two numbers get confused constantly. The case fatality rate divides deaths by confirmed cases. The infection fatality rate divides deaths by all infections, including the ones nobody counted. Early in an outbreak, testing usually catches the sickest patients first, so the CFR looks alarming and then falls as testing widens.

A hypothetical example shows the gap. Suppose a town has 1,000 infections. Tests confirm only 500 of them, and five people die. The CFR is 5 out of 500, or 1%. The IFR is 5 out of 1,000, or 0.5%. Nothing about the virus changed, only the denominator.

COVID-19 and the age gradient

COVID-19's risk was steeply tied to age. A systematic review and meta-analysis by Andrew Levin and colleagues, published in the European Journal of Epidemiology in December 2020, estimated age-specific infection fatality rates from seroprevalence studies:

Age	Estimated IFR
10	0.002%
25	0.01%
55	0.4%
65	1.4%
75	4.6%
85	15%

The risk rises by orders of magnitude across the age range. A 25-year-old and an 85-year-old were not facing the same disease, in effect, and that gradient drove many of the policy disagreements of 2020 and 2021. It also means any single "COVID fatality rate" depends heavily on the age structure of the population being described. These figures are from the pre-vaccine period and reflect the early variants, so they describe a world that no longer exists in the same form.

Counting the dead

The reported death count and the true death toll are not the same. The WHO's dashboard states that its counts primarily reflect laboratory-confirmed cases and deaths, and it warns that reduced testing and reporting mean its figures do not accurately reflect infection or mortality rates. The agency's reported total is several million deaths, well below the excess-mortality estimate, and it treats excess mortality as the better measure of the full toll.

Excess mortality compares the deaths that occurred with the number expected without the pandemic. In May 2022, the WHO estimated about 14.9 million excess deaths associated with the pandemic from January 1, 2020, through December 31, 2021, with an uncertainty range of 13.3 million to 16.6 million. That figure includes people who died directly of COVID-19 and people who died because health services were disrupted, and it is reduced slightly by deaths that fell, such as from road accidents. The WHO also reported that 57% of the excess deaths were among men and 43% among women, and that about 81% occurred in middle-income countries.

What stops pandemics: interventions that have been measured

Interventions fall into two families. Non-pharmaceutical interventions change behavior or the environment. Pharmaceutical interventions change the body's defenses or the pathogen's ability to replicate. Neither family works alone, and each has costs.

Non-pharmaceutical interventions

Isolation and quarantine. Isolation separates people who are infected. Quarantine separates people who might be. Both break transmission chains when cases are found fast enough, which is why the share of spread that occurs before symptoms matters so much. For a disease like measles or COVID-19, isolation after symptoms appear misses a large slice of transmission.

Contact tracing. Tracing identifies the people a case may have infected so they can be tested, informed, and, if needed, quarantined. It works best when case counts are low enough for health departments to keep up. Once an outbreak outruns tracing capacity, other measures have to do the work.

Masks and respirators. Masks reduce the number of infectious particles that move between people, and respirator-style masks fit more tightly and filter more finely than loose cloth or surgical masks. The size of the benefit depends on the type of mask, how well it fits, how consistently it is worn, and how the study measured the outcome, which is why trials and reviews have produced a wide spread of conclusions. A fair summary is that the evidence is strongest for high-quality respirators in high-risk settings and weaker and more mixed for community mask mandates.

Distancing and lockdowns. Cutting contacts reduces the number of people each case can infect. The simple arithmetic is that if contacts fall by half and everything else stays equal, R falls by half. A lockdown that moves R from 2.5 to 1.25 slows the growth but does not end it. Lockdowns also carry heavy economic, educational, and health costs, which is why many countries used them as short-term tools to buy time for hospitals and vaccines.

Ventilation and air filtration. For airborne diseases, replacing or cleaning indoor air lowers the concentration of infectious particles. Unlike lockdowns, ventilation is a one-time engineering fix that works around the clock without asking anyone to change daily routines, which makes it the least disruptive item on this list.

Vaccines

A vaccine teaches the immune system to recognize a pathogen before the real infection arrives, so the first encounter is not a first encounter. The main designs differ in how they present the target:

Vaccine type	How it works	Examples
Live attenuated	A weakened version of the pathogen	Measles, chickenpox
Inactivated	A killed pathogen	Injected influenza vaccine, polio (IPV)
Subunit or protein	A purified piece of the pathogen	Hepatitis B, HPV
mRNA	Instructions for cells to make one pathogen protein, which the immune system then learns	Pfizer-BioNTech and Moderna COVID-19 vaccines

Live attenuated vaccines often produce strong, durable immunity but are not suitable for everyone, such as some people with weakened immune systems. Inactivated vaccines are safer in that respect but often need multiple doses or boosters. mRNA vaccines contain no live pathogen, and because they only need the genetic sequence of the target protein, they can be designed quickly. Moderna has said it settled on its COVID-19 vaccine sequence within days of the SARS-CoV-2 genome becoming public in January 2020, though the real work of trials, manufacturing, and safety review took the rest of 2020. The technology's longer history is covered in mRNA vaccines and the COVID-19 breakthrough.

How much did vaccination matter? A 2022 modeling study by Oliver Watson and colleagues in The Lancet Infectious Diseases estimated that COVID-19 vaccination prevented 14.4 million deaths in 185 countries and territories during its first year, from December 8, 2020, to December 8, 2021, using reported COVID-19 deaths, and 19.8 million using excess mortality. These are model estimates built on assumptions about what would have happened otherwise, so they are best read as a plausible range and not a precise count. For a plain-language look at how safety is established, see How We Know Vaccines Are Safe.

Vaccines did not stop all transmission, especially after Omicron arrived, because that variant partly evades immunity from earlier infection and vaccination. Their largest effect was on severe disease and death, which is what overwhelms health systems.

Treatments

Antivirals target the virus's replication, and examples include oseltamivir for influenza, antiretroviral therapy for HIV, and direct-acting antivirals for hepatitis C. Antibiotics treat bacterial infections but do nothing against viruses, though they matter in pandemics because bacterial pneumonia can follow a viral infection. The CDC notes that in 1918 there were no antibiotics to treat the secondary bacterial infections that accompanied influenza.

The 1918 influenza: the benchmark pandemic

The 1918 pandemic is the reference point for pandemic planning. The CDC estimates that about 500 million people, one-third of the world's population, were infected, and that the death toll was at least 50 million worldwide, including about 675,000 in the United States.

It also had an unusual age signature. Influenza usually kills infants and the elderly most. The CDC says deaths in 1918 were concentrated in three groups: children under five, adults aged 20 to 40, and adults 65 and older, and it calls the high death rate among healthy 20- to 40-year-olds a unique feature of the pandemic. The reasons for that pattern are still debated, and it is wise to treat any single tidy explanation with suspicion.

Conditions in 1918 amplified the outbreak. Wartime troop movements and crowded barracks carried the virus across continents, and there were no antibiotics for secondary infections, no vaccines for influenza, and no modern intensive care. Control relied on isolation, quarantine, hygiene, disinfectants, and limits on public gatherings, according to the CDC. The CDC describes the pandemic as arriving in three waves, which is one reason a pandemic that seems to be over can come back.

A comparison with COVID-19 should be careful. The two pandemics unfolded in different worlds: 1918 had troop ships and trains, while 2020 had commercial air travel and genome sequencing within weeks, so the geography of spread and the available defenses were very different. For why influenza keeps producing new threats, see how seasonal flu works.

Surveillance and prevention: the case for acting early

If exponential growth is the problem, speed is the solution. Every day of delay in detecting a new pathogen is a day of unchecked doubling. That makes surveillance, the routine collection and sharing of health data, the least glamorous and most valuable tool in the kit.

Sequencing and data sharing. Rapid genome sequencing lets researchers identify a new pathogen, design tests, start vaccine work, and track variants. The SARS-CoV-2 genome was made public in January 2020, which is what allowed vaccine developers to start work within days. Machine learning is now being applied to predicting how viruses evolve, as discussed in AI and viral evolution prediction.

Reporting. Countries depend on one another for early warning. The WHO's International Health Regulations are the legal framework under which countries agree to report events that might constitute a public health emergency of international concern. Reporting is only as good as the local capacity to detect disease and the willingness to report it promptly.

One Health. Because many new pathogens are zoonotic, monitoring animal health, especially where wildlife, livestock, and people mix, is part of prevention. This is an investment in noticing spillover before it becomes human-to-human spread.

New agreements. On May 20, 2025, member states adopted the WHO Pandemic Agreement by consensus at the 78th World Health Assembly, after a committee vote of 124 in favor, none opposed, and 11 abstentions. It sets out principles and tools for coordination on pandemic prevention, preparedness, and response, including equitable access to vaccines, therapeutics, and diagnostics. The text states that it gives the WHO no authority to direct national laws or impose measures such as lockdowns or vaccine mandates. It will open for signature only after the Assembly adopts an annex on pathogen access and benefit sharing, and it takes effect after 60 ratifications. As of its adoption, then, it was an agreement on paper with major details still to negotiate.

The bottlenecks. Even with good detection, a few constraints recur: how fast vaccines can be manufactured at scale, how quickly diagnostic testing can be deployed, how equitably doses are distributed, and whether countries share data and materials in time. Those are political and logistical problems as much as scientific ones.

What people get wrong about pandemics

"If the death rate is low, it is not a big deal." A low rate applied to a very large number of infections still yields large totals. Seasonal influenza's death toll in the hundreds of thousands each year, per the WHO, is the everyday demonstration.

"Herd immunity means everyone is protected." It means transmission is reduced enough that outbreaks shrink. Individuals who are not immune can still get sick, especially in clusters where immunity is low. Measles outbreaks in under-vaccinated communities show how local gaps can sustain spread even when the national average looks high.

"A virus will evolve to be milder over time." This is a common claim, but evolution selects for whatever helps a virus spread, not for mildness. A variant can become more transmissible without becoming milder or more severe. What changed most between 2020 and later years was the immunity of the human population, not an intrinsic softening of the virus.

"The first number we hear is the real number." Early R₀ values, early case counts, and early fatality rates are all provisional. They are revised as testing improves and as researchers see how many infections went uncounted.

"Interventions are all or nothing." Each measure has a partial effect, and the effects combine. A layered strategy with vaccination, ventilation, testing, and targeted isolation can do more than any one tool pushed to its limits, and at lower cost.

Frequently asked questions

Q: What is the difference between an epidemic and a pandemic? A: An epidemic is a rise in cases of a disease above what is expected in a population or region. A pandemic is an epidemic that has spread across multiple countries or continents and affects a large number of people. The labels describe geographic spread, not severity. A pandemic can be mild, and a local epidemic can be deadly.

Q: How does COVID-19 compare with historical pandemics? A: By raw deaths, the WHO estimates about 14.9 million excess deaths in 2020 and 2021, and several million deaths have been reported directly to the WHO. The CDC puts the 1918 influenza at 50 million or more deaths. The world population in 1918 was much smaller than today, so the 1918 pandemic killed a far larger share of people, and its toll fell unusually on young adults. The comparison is imperfect because counting methods, health care, and demographics differ.

Q: Could herd immunity without vaccines have worked for COVID-19? A: In principle, transmission slows once enough people are immune, whether through infection or vaccination. In practice, relying on infection means accepting that most of the population is infected, and the simple model suggests overshoot beyond the threshold. With an age-specific IFR that reaches 15% at age 85 in the pre-vaccine period, that path meant very large numbers of deaths and a heavy burden on hospitals. Vaccination reaches immunity before exposure.

Q: How do new variants emerge? A: Viruses copy their genetic material imperfectly, so every replication cycle can introduce mutations. Most are harmless or harmful to the virus, but occasionally a change helps the virus spread or evade immunity. When that happens, the variant outcompetes others and becomes dominant. Omicron is the clearest example, since its estimated R₀ was far higher than earlier versions and it partly evaded prior immunity.

Q: What is the most likely source of the next pandemic? A: No one can say with confidence. Risk assessments commonly focus on influenza viruses that circulate in animals, coronaviruses, and filoviruses such as Ebola, because these families have produced past outbreaks, and the WHO notes zoonoses account for a large percentage of newly identified infectious diseases. The common ingredients are close contact between people and animals, dense populations, and limited health-system capacity.

Q: Can pandemics be prevented entirely? A: Not entirely. Spillover events from animals will keep occurring. But the damage from a new pathogen depends on how quickly it is detected, how fast countermeasures are deployed, and how equitably they are shared. Surveillance, sequencing, flexible vaccine platforms, and international agreements aim to reduce both the chance that a spillover becomes a pandemic and the harm if it does.

The takeaway

Pandemics are governed by a short list of quantities: how many people one case infects, how long it takes them to do it, how many people are already immune, and what share of infections turn deadly. Change any of those and the outcome changes, which is why the same virus can be a nuisance in one setting and a catastrophe in another. COVID-19 showed both sides of the ledger. A new virus crossed the world in weeks, and vaccines were designed within days of the genome's release and, by one modeling estimate, prevented between 14.4 and 19.8 million deaths in their first year. The lesson that holds across 1918, 2020, and any future outbreak is that the cheapest interventions are the early ones, and they only work if surveillance, funding, and trust are in place before the first case is recognized.

Sources

• World Health Organization, 14.9 million excess deaths were associated with the COVID-19 pandemic in 2020 and 2021 (May 5, 2022)
• WHO, COVID-19 dashboard, reporting notes
• CDC, 1918 Pandemic (H1N1 virus)
• CDC, Measles clinical overview
• CDC, About smallpox
• Levin AT et al., Assessing the age specificity of infection fatality rates for COVID-19, European Journal of Epidemiology, 2020
• Liu Y, Rocklöv J, The effective reproductive number of the Omicron variant of SARS-CoV-2 is several times relative to Delta, Journal of Travel Medicine, 2022
• Billah MA, Miah MM, Khan MN, Reproductive number of coronavirus: a systematic review and meta-analysis, PLOS ONE, 2020
• Johansson MA et al., SARS-CoV-2 transmission from people without COVID-19 symptoms, JAMA Network Open, 2021
• Fine P, Eames K, Heymann DL, "Herd Immunity": A Rough Guide, Clinical Infectious Diseases, 2011
• WHO, Ebola disease fact sheet
• WHO, Influenza (seasonal) fact sheet
• WHO, Zoonoses fact sheet
• WHO, World Health Assembly adopts historic Pandemic Agreement (May 20, 2025)

Related Reading

• How Vaccines Work — and Why Some People Refuse Them
• How Human Immune System Fights Viruses: A Step-by-Step Guide
• How Black Holes Form — and What Happens Inside Them
• How Nuclear Weapons Work — and Why They Haven't Been Used Since 1945
• How Democracy Works — and Why It's Under Threat
• What Is the Stock Market — and How Does It Actually Work?